Digital flowmeter

[0002] This application claims priority from U.S. Provisional Application No. 60/066,554, filed Nov. 26, 1997, and entitled DIGITAL FLOWMETER, which is incorporated by reference.

[0003] The invention relates to flowmeters.

[0004] Flowmeters provide information about materials being transferred through a conduit. For example, mass flowmeters provide a direct indication of the mass of material being transferred through a conduit. Similarly, density flowmeters, or densitometers, provide an indication of the density of material flowing through a conduit. Mass flowmeters also may provide an indication of the density of the material.

[0005] Coriolis-type mass flowmeters are based on the well-known Coriolis effect, in which material flowing through a rotating conduit becomes a radially travelling mass that is affected by a Coriolis force and therefore experiences an acceleration. Many Coriolis-type mass flowmeters induce a Coriolis force by sinusoidally oscillating a conduit about a pivot axis orthogonal to the length of the conduit. In such mass flowmeters, the Coriolis reaction force experienced by the travelling fluid mass is transferred to the conduit itself and is manifested as a deflection or offset of the conduit in the direction of the Coriolis force vector in the plane of rotation.

[0006] Energy is supplied to the conduit by a driving mechanism that applies a periodic force to oscillate the conduit. One type of driving mechanism is an electromechanical driver that imparts a force proportional to an applied voltage. In an oscillating flowmeter, the applied voltage is periodic, and is generally sinusoidal. The period of the input voltage is chosen so that the motion of the conduit matches a resonant mode of vibration of the conduit. This reduces the energy needed to sustain oscillation. An oscillating flowmeter may use a feedback loop in which a sensor signal that carries instantaneous frequency and phase information related to oscillation of the conduit is amplified and fed back to the conduit using the electromechanical driver.

[0007] The invention provides a digital flowmeter, such as a digital mass flowmeter, that uses a control and measurement system to control oscillation of the conduit and to generate mass flow and density measurements. Sensors connected to the conduit supply signals to the control and measurement system. The control and measurement system processes the signals to produce a measurement of mass flow and uses digital signal processing to generate a signal for driving the conduit. The drive signal then is converted to a force that induces oscillation of the conduit.

[0008] The digital mass flowmeter provides a number of advantages over traditional, analog approaches. From a control perspective, use of digital processing techniques permits the application of precise, sophisticated control algorithms that, relative to traditional analog approaches, provide greater responsiveness, accuracy and adaptability.

[0009] The digital control system also permits the use of negative gain in controlling oscillation of the conduit. Thus, drive signals that are 180° out of phase with conduit oscillation may be used to reduce the amplitude of oscillation. The practical implications of this are important, particularly in high and variable damping situations where a sudden drop in damping can cause an undesirable increase in the amplitude of oscillation. One example of a variable damping situation is when aeration occurs in the material flowing through the conduit.

[0010] The ability to provide negative feedback also is important when the amplitude of oscillation is controlled to a fixed setpoint that can be changed under user control. With negative feedback, reductions in the oscillation setpoint can be implemented as quickly as increases in the setpoint. By contrast, an analog meter that relies solely on positive feedback must set the gain to zero and wait for system damping to reduce the amplitude to the reduced setpoint.

[0011] From a measurement perspective, the digital mass flowmeter can provide high information bandwidth. For example, a digital measurement system may use analog-to-digital converters operating at eighteen bits of precision and sampling rates of 55 kHz. The digital measurement system also may use sophisticated algorithms to filter and process the data, and may do so starting with the raw data from the sensors and continuing to the final measurement data. This permits extremely high precision, such as, for example, phase precision to five nanoseconds per cycle. Digital processing starting with the raw sensor data also allows for extensions to existing measurement techniques to improve performance in non-ideal situations, such as by detecting and compensating for time-varying amplitude, frequency, and zero offset.

[0012] The control and measurement improvements interact to provide further improvements. For example, control of oscillation amplitude is dependent upon the quality of amplitude measurement. Under normal conditions, the digital mass flowmeter may maintain oscillation to within twenty parts per million of the desired setpoint. Similarly, improved control has a positive benefit on measurement. Increasing the stability of oscillation will improve measurement quality even for meters that do not require a fixed amplitude of oscillation (i.e., a fixed setpoint). For example, with improved stability, assumptions used for the measurement calculations are likely to be valid over a wider range of conditions.

[0013] The digital mass flowmeter also permits the integration of entirely new functionality (e.g., diagnostics) with the measurement and control processes. For example, algorithms for detecting the presence of process aeration can be implemented with compensatory action occurring for both measurement and control if aeration is detected.

[0014] Other advantages of the digital mass flowmeter result from the limited amount of hardware employed, which makes the meter simple to construct, debug, and repair in production and in the field. Quick repairs in the field for improved performance and to compensate for wear of the mechanical components (e.g, loops, flanges, sensors and drivers) are possible because the meter uses standardized hardware components that may be replaced with little difficulty, and because software modifications may be made with relative ease. In addition, integration of diagnostics, measurement, and control is simplified by the simplicity of the hardware and the level of functionality implemented in software. New functionality, such as low power components or components with improved performance, can be integrated without a major redesign of the overall control system.

[0015] In one general aspect, a digital flowmeter includes a vibratable conduit, a driver connected to the conduit and operable to impart motion to the conduit, and a sensor connected to the conduit and operable to sense the motion of the conduit. A control and measurement system connected between the driver and the sensor includes circuitry that receives a sensor signal from the sensor, generates a drive signal based on the sensor signal using digital signal processing, supplies the drive signal to the driver, and generates a measurement of a property of material flowing through the conduit based on the signal from the sensor.

[0016] Embodiments may include one or more of the following features. The meter may include a second sensor connected to the conduit and operable to sense the motion of the conduit. In this case, the control and measurement system is connected to the second sensor and receives a second sensor signal from the second sensor, generates the drive signal based on the first and second sensor signals, and generates the measurement of the property of material flowing through the conduit based on the first and second sensor signals. The control and measurement system may digitally combine the first and second sensor signals and generate the drive signal based on the combination of the sensor signals.

[0017] The control and measurement system may generate different drive signals for the two drivers. The drive signals may have, for example, different frequencies or amplitudes.

[0018] The digital flowmeter also may include circuitry for measuring current supplied to the driver. The circuitry may include a resistor in series with the driver and an analog-to-digital converter in parallel with the resistor and configured to measure a voltage across the resistor, to convert the measured voltage to a digital value, and to supply the digital value to the control and measurement system.

[0019] The digital flowmeter also may include a first pressure sensor connected to measure a first pressure at an inlet to the conduit and a second pressure sensor connected to measure a second pressure at an outlet of the conduit. Analog-to-digital converters may be connected and configured to convert signals produced by the first pressure sensor and the second pressure sensor to digital values and to supply the digital values to the control and measurement system. Temperature sensors may be connected to measure temperatures at the inlet and outlet of the conduit.

[0020] The control and measurement system may generate the measurement of the property by estimating a frequency of the first sensor signal, calculating a phase difference using the first sensor signal, and generating the measurement using the calculated phase difference. The control and measurement system may compensate for amplitude differences in the sensor signals by adjusting the amplitude of one of the sensor signals. For example, the control and measurement system may multiply the amplitude of one of the sensor signals by a ratio of the amplitudes of the sensor signals.

[0021] When the sensor signal is generally periodic, the control and measurement system may process the sensor signal in sets. Each set may include data for a complete cycle of the periodic sensor signal, and consecutive sets may include data for overlapping cycles of the periodic sensor signal. The control and measurement system may estimate an end point of a cycle using a frequency of a previous cycle.

[0022] The control and measurement system may analyze data for a cycle to determine whether the cycle merits further processing. For example, the system may determine that a cycle does not merit further processing when data for the cycle does not conform to expected behavior for the data, where the expected behavior may be based on one or more parameters of a previous cycle. In one implementation, the system determines that a cycle does not merit further processing when a frequency for the cycle differs from a frequency for the previous cycle by more than a threshold amount. The system may determine whether the frequencies differ by comparing values at certain points in the cycle to values that would occur if the frequency for the cycle equalled the frequency for the previous cycle.

[0023] The control and measurement system may determine a frequency of the sensor signal by detecting zero-crossings of the sensor signal and counting samples between zero crossings. The system also may determine a frequency of the sensor signal using an iterative curve fitting technique.

[0024] The control and measurement system may determine an amplitude of the sensor signal using Fourier analysis, and may use the determined amplitude in generating the drive signal.

[0025] The control and measurement system may determine a phase offset for each sensor signal and may determine the phase difference by comparing the phase offsets. The system also may determine the phase difference using Fourier analysis. The control and measurement system may determine a frequency, amplitude and phase offset for each sensor signal, and may scale the phase offsets to an average of the frequencies of the sensor signals. The control and measurement system may calculate the phase difference using multiple approaches and may select a result of one of the approaches as the calculated phase difference.

[0026] The control and measurement system may combine the sensor signals to produce a combined signal and may generate the drive signal based on the combined signal. For example, the control and measurement system may sum the sensor signals to produce the combined signal and may generate the drive signal by applying a gain to the combined signal.

[0027] In general, the control and measurement system may initiate motion of the conduit by using a first mode of signal generation to generate the drive signal, and may sustain motion of the conduit using a second mode of signal generation to generate the drive signal. The first mode of signal generation may be synthesis of a periodic signal having a desired property, such as a desired initial frequency of conduit vibration, and the second mode of signal generation may use a feedback loop including the sensor signal.

[0028] In other instances, the first mode of signal generation may include use of a feedback loop including the sensor signal and the second mode of signal generation may include synthesis of a periodic signal having a desired property. For example, the control and measurement system may generate the drive signal by applying a large gain to the combined signal to initiate motion of the conduit and generating a periodic signal having a phase and frequency based on a phase and frequency of a sensor signal as the drive signal after motion has been initiated. The desired property of the synthesized signal may be a frequency and a phase corresponding to a frequency and a phase of the sensor signal.

[0029] The control and measurement system generates an adaptable, periodic drive signal. For example, the meter may include positive and negative direct current sources connected between the control and measurement system and the driver, and the control and measurement system may generate the drive signal by switching the current sources on and off at intervals having a phase and frequency based on the sensor signal. The control and measurement system may generate the drive signal by synthesizing a sine wave having a property corresponding to a property of the sensor signal, such as a phase and a frequency corresponding to a phase and a frequency of the sensor signal.

[0030] The control and measurement system may digitally generate a gain for use in generating the drive signal based on one or more properties of the sensor signal. For example, the control and measurement system may digitally generate the gain based on an amplitude of the sensor signal.

[0031] The driver may be operable to impart an oscillating motion to the conduit. The control and measurement system also may digitally implement a PI control algorithm to regulate the amplitude of conduit oscillation. The control and measurement system also may digitally generate the drive signal based on the sensor signal so as to maintain an amplitude of oscillation of the conduit at a user-controlled value. In support of this, the control and measurement system may generate a negative drive signal that causes the driver to resist motion of the conduit when the amplitude of oscillation exceeds the user-controlled value and a positive drive signal that causes the driver to impart motion to the conduit when the amplitude of oscillation is less than the user-controlled value.

[0032] The control and measurement system may include a controller that generates a gain signal based on the sensor signal and a multiplying digital-to-analog converter connected to the controller to receive the gain signal and generate the drive signal based on the gain signal.

[0033] When the digital flowmeter includes a second sensor connected to the conduit and operable to sense the motion of the conduit, the control and measurement system may include a controller that generates the measurement, a first analog-to-digital converter connected between the first sensor and the controller to provide a first digital sensor signal to the controller, and a second analog-to-digital converter connected between the second sensor and the controller to provide a second digital sensor signal to the controller. The controller may combine the digital sensor signals to produce a combined signal and to generate a gain signal based on the first and second digital sensor signals. The control and measurement system also may include a multiplying digital-to-analog converter connected to receive the combined signal and the gain signal from the controller to generate the drive signal as a product of the combined signal and the gain signal.

[0034] The control and measurement system may selectively apply a negative gain to the sensor signal to reduce motion of the conduit.

[0035] The control and measurement system also may compensate for zero offset in the sensor signal. The zero offset may include a component attributable to gain variation and a component attributable to gain nonlinearity, and the control and measurement system may separately compensate for the two components. The control and measurement system may compensate for zero offset by generating one or more correction factors and modifying the sensor signal using the correction factors.

[0036] The control and measurement system may calculate phase offsets for the first and second sensor signals. The phase offset may be defined as a difference between a zero-crossing point of a sensor signal and a point of zero phase for a component of the sensor signal corresponding to a fundamental frequency of the sensor signal. The control and measurement system may combine the calculated phase offsets to produce a phase difference.

[0037] The control and measurement system may generate the measurement of the property by estimating a frequency of the first sensor signal, estimating a frequency of the second sensor signal, with the frequency of the second sensor signal being different from the frequency of the first sensor signal, and calculating a phase difference between the sensor signals using the estimated frequencies.

[0038] When the sensor is a velocity sensor, the control and measurement system may estimate a frequency, amplitude, and phase of the sensor signal, and may correct the estimated frequency, amplitude, and phase to account for performance differences between a velocity sensor and an absolute position sensor. Instead of controlling the apparent amplitude of oscillation (i.e., the velocity of oscillation when the sensor is a velocity sensor), the system may control the true amplitude by dividing the sensor signal by the estimated frequency. This correction should provide improved amplitude control and noise reduction.

[0039] The control and measurement system may estimate a first parameter of the sensor signal, determine a rate of change of a second parameter, and correct the estimated first parameter based on the determined rate of change. For example, the system may correct an estimated frequency, amplitude, or phase of the sensor signal based on a determined rate of change of the frequency or amplitude of oscillation of the conduit. The system may perform separate corrections for each sensor signal.

[0040] The digital flowmeter may be a mass flowmeter and the property of material flowing through the conduit may be a mass flow rate. The digital flowmeter also may be a densitometer and the property of material flowing through the conduit may be a density of the material.

[0041] The control and measurement system may account for effects of aeration in the conduit by determining an initial mass flow rate, determining an apparent density of material flowing through the conduit, comparing the apparent density to a known density of the material to determine a density difference, and adjusting the initial mass flow rate based on the density difference to produce an adjusted mass flow rate. The system may further account for effects of aeration in the conduit by adjusting the adjusted mass flow rate to account for effects of damping. To further account for effects of aeration in the conduit, the system may adjust the adjusted mass flow rate based on differences between amplitudes of the first and second sensor signals.

[0042] The vibratable conduit may include two parallel planar loops. The sensor and driver may be connected between the loops.

[0043] The meter may include a power circuit that receives power on only a single pair of wires. The power circuit provides power to the digital control and measurement system and to the driver, and the digital control and measurement system is operable to transmit the measurement of the property of material flowing through the conduit on the single pair of wires. The power circuit may include a constant output circuit that provides power to the digital control and measurement system and drive capacitor that is charged by excess power from the two wires. The digital control and measurement system may discharge the drive capacitor to power the driver, and may monitor a charge level of the drive capacitor and discharge the drive capacitor after a charge level of the capacitor reaches a threshold level. The digital control and measurement system also may discharge the drive capacitor periodically and to perform bidirectional communications on the pair of wires.

[0044] The control and measurement system may collect a first data set for a period of the periodic signal and process the first data set to generate the drive signal and the measurement. The system may collect a second data set for a subsequent period of the sensor signal simultaneously with processing the first data set. The period corresponding to the first data set may overlap the period corresponding to the second data set.

[0045] The control and measurement system may control the drive signal to maintain an amplitude of the sensor signal at a fixed setpoint, reduce the fixed setpoint when the drive signal exceeds a first threshold level, and increase the fixed setpoint when the drive signal is less than a second threshold level and the fixed setpoint is less than a maximum permitted value for the setpoint. The first threshold level may be 95% or less of a maximum permitted drive signal.

[0046] The control and measurement system may perform an uncertainty analysis on the measurement. In this case, the control and measurement system may transmit the measurement and results of the uncertainty analysis to the control system.

[0047] The control and measurement system may use digital processing to adjust a phase of the drive signal to compensate for a time delay associated with the sensor and components connected between the sensor and the driver.

[0048] Other features and advantages of the invention will be apparent from the following description, including the drawings, and from the claims.

[0049] FIG. 1 is a block diagram of a digital mass flowmeter.

[0050] FIGS. 2A and 2B are perspective and side views of mechanical components of a mass flowmeter.

[0051] FIGS. 3-3C are schematic representations of three modes of motion of the flowmeter of FIG. 1.

[0052] FIG. 4 is a block diagram of an analog control and measurement circuit.

[0053] FIG. 5 is a block diagram of a digital mass flowmeter.

[0054] FIG. 6 is a flow chart showing operation of the meter of FIG. 5.

[0055] FIGS. 7A and 7B are graphs of sensor data.

[0056] FIGS. 8A and 8B are graphs of sensor voltage relative to time.

[0057] FIG. 9 is a flow chart of a curve fitting procedure.

[0058] FIG. 10 is a flow chart of a procedure for generating phase differences.

[0059] FIGS. 11A-11D, 12A-12D, and 13A-13D illustrate drive and sensor voltages at system startup.

[0060] FIG. 14 is a flow chart of a procedure for measuring frequency, amplitude, and phase of sensor data using a synchronous modulation technique.

[0061] FIGS. 15A and 15B are block diagrams of a mass flowmeter.

[0062] FIG. 16 is a flow chart of a procedure implemented by the meter of FIGS. 15A and 15B.

[0063] FIG. 17 illustrates log-amplitude control of a transfer function.

[0064] FIG. 18 is a root locus diagram.

[0065] FIGS. 19A-19D are graphs of analog-to-digital converter performance relative to temperature.

[0066] FIGS. 20A-20C are graphs of phase measurements.

[0067] FIGS. 21A and 21B are graphs of phase measurements.

[0068] FIG. 22 is a flow chart of a zero offset compensation procedure.

[0069] FIGS. 23A-23C, 24A, and 24B are graphs of phase measurements.

[0070] FIG. 25 is a graph of sensor voltage.

[0071] FIG. 26 is a flow chart of a procedure for compensating for dynamic effects.

[0072] FIGS. 27A-35E are graphs illustrating application of the procedure of FIG. 29.

[0073] FIGS. 36A-36L are graphs illustrating phase measurement.

[0074] FIG. 37A is a graph of sensor voltages.

[0075] FIGS. 37B and 37C are graphs of phase and frequency measurements corresponding to the sensor voltages of FIG. 37A.

[0076] FIGS. 37D and 37E are graphs of correction parameters for the phase and frequency measurements of FIGS. 37B and 37C.

[0077] FIGS. 38A-38H are graphs of raw measurements.

[0078] FIGS. 39A-39H are graphs of corrected measurements.

[0079] FIGS. 40A-40H are graphs illustrating correction for aeration.

[0080] FIG. 41 is a block diagram illustrating the effect of aeration in a conduit.

[0081] FIG. 42 is a flow chart of a setpoint control procedure.

[0082] FIGS. 43A-43C are graphs illustrating application of the procedure of FIG. 41.

[0083] FIG. 44 is a graph comparing the performance of digital and analog flowmeters.

[0084] FIG. 45 is a flow chart showing operation of a self-validating meter.

[0085] FIG. 46 is a block diagram of a two-wire digital mass flowmeter.

[0086] Referring to FIG. 1, a digital mass flowmeter 100 includes a digital controller 105, one or more motion sensors 110, one or more drivers 115, a conduit 120 (also referred to as a flowtube), and a temperature sensor 125. The digital controller 105 may be implemented using one or more of, for example, a processor, a field-programmable gate array, an ASIC, other programmable logic or gate arrays, or programmable logic with a processor core. The digital controller generates a measurement of mass flow through the conduit 120 based at least on signals received from the motion sensors 110. The digital controller also controls the drivers 115 to induce motion in the conduit 120. This motion is sensed by the motion sensors 110.

[0087] Mass flow through the conduit 120 is related to the motion induced in the conduit in response to a driving force supplied by the drivers 115. In particular, mass flow is related to the phase and frequency of the motion, as well as to the temperature of the conduit. The digital mass flowmeter also may provide a measurement of the density of material flowing through the conduit. The density is related to the frequency of the motion and the temperature of the conduit. Many of the described techniques are applicable to a densitometer that provides a measure of density rather than a measure of mass flow.

[0088] The temperature in the conduit, which is measured using the temperature sensor 125, affects certain properties of the conduit, such as its stiffness and dimensions. The digital controller compensates for these temperature effects. The temperature of the digital controller 105 affects, for example, the operating frequency of the digital controller. In general, the effects of controller temperature are sufficiently small to be considered negligible. However, in some instances, the digital controller may measure the controller temperature using a solid state device and may compensate for effects of the controller temperature.

[0089] A. Mechanical Design

[0090] In one implementation, as illustrated in FIGS. 2A and 2B, the conduit 120 is designed to be inserted in a pipeline (not shown) having a small section removed or reserved to make room for the conduit. The conduit 120 includes mounting flanges 12 for connection to the pipeline, and a central manifold block 16 supporting two parallel planar loops 18 and 20 that are oriented perpendicularly to the pipeline. An electromagnetic driver 46 and a sensor 48 are attached between each end of loops 18 and 20. Each of the two drivers 46 corresponds to a driver 115 of FIG. 1, which each of the two sensors 48 corresponds to a sensor 110 of FIG. 1.

[0091] The drivers 46 on opposite ends of the loops are energized with current of equal magnitude but opposite sign (i.e., currents that are 180° out-of-phase) to cause straight sections 26 of the loops 18, 20 to rotate about their co-planar perpendicular bisector 56, which intersects the tube at point P (FIG. 2B). Repeatedly reversing (e.g., controlling sinusoidally) the energizing current supplied to the drivers causes each straight section 26 to undergo oscillatory motion that sweeps out a bow tie shape in the horizontal plane about line 56—56, the axis of symmetry of the loop. The entire lateral excursion of the loops at the lower rounded turns 38 and 40 is small, on the order of {fraction (1/16)} of an inch for a two foot long straight section 26 of a pipe having a one inch diameter. The frequency of oscillation is typically about 80 to 90 Hertz.

[0092] B. Conduit Motion

[0093] The motion of the straight sections of loops 18 and 20 are shown in three modes in FIGS. 3A, 3B and 3C. In the drive mode shown in FIG. 3B, the loops are driven 180° out-of-phase about their respective points P so that the two loops rotate synchronously but in the opposite sense. Consequently, respective ends such as A and C periodically come together and go apart.

[0094] The drive motion shown in FIG. 3B induces the Coriolis mode motion shown in FIG. 3A, which is in opposite directions between the loops and moves the straight sections 26 slightly toward (or away) from each other. The Coriolis effect is directly related to mνW, where m is the mass of material in a cross section of a loop, ν is the velocity at which the mass is moving (the volumetric flow rate), W is the angular velocity of the loop (W=W_osin ωt), and mν is the mass flow rate. The Coriolis effect is greatest when the two straight sections are driven sinusoidally and have a sinusoidally varying angular velocity. Under these conditions, the Coriolis effect is 90° out-of-phase with the drive signal.

[0095] FIG. 3C shows an undesirable common mode motion that deflects the loops in the same direction. This type of motion might be produced by an axial vibration in the pipeline in the embodiment of FIGS. 2A and 2B because the loops are perpendicular to the pipeline.

[0096] The type of oscillation shown in FIG. 3B is called the antisymmetrical mode, and the Coriolis mode of FIG. 3A is called the symmetrical mode. The natural frequency of oscillation in the antisymmetrical mode is a function of the torsional resilience of the legs. Ordinarily the resonant frequency of the antisymmetrical mode for conduits of the shape shown in FIGS. 2A and 2B is higher than the resonant frequency of the symmetrical mode. To reduce the noise sensitivity of the mass flow measurement, it is desirable to maximize the Coriolis force for a given mass flow rate. As noted above, the loops are driven at their resonant frequency, and the Coriolis force is directly related to the frequency at which the loops are oscillating (i.e., the angular velocity of the loops). Accordingly, the loops are driven in the antisymmetrical mode, which tends to have the higher resonant frequency.

[0097] Other implementations may include different conduit designs. For example, a single loop or a straight tube section may be employed as the conduit.

[0098] C. Electronic Design

[0099] The digital controller 105 determines the mass flow rate by processing signals produced by the sensors 48 (i.e., the motion sensors 110) located at opposite ends of the loops. The signal produced by each sensor includes a component corresponding to the relative velocity at which the loops are driven by a driver positioned next to the sensor and a component corresponding to the relative velocity of the loops due to Coriolis forces induced in the loops. The loops are driven in the antisymmetrical mode, so that the components of the sensor signals corresponding to drive velocity are equal in magnitude but opposite in sign. The resulting Coriolis force is in the symmetrical mode so that the components of the sensor signals corresponding to Coriolis velocity are equal in magnitude and sign. Thus, differencing the signals cancels out the Coriolis velocity components and results in a difference that is proportional to the drive velocity. Similarly, summing the signals cancels out the drive velocity components and results in a sum that is proportional to the Coriolis velocity, which, in turn, is proportional to the Coriolis force. This sum then may be used to determine the mass flow rate.

[0100] 1. Analog Control System

[0101] The digital mass flowmeter 100 provides considerable advantages over traditional, analog mass flowmeters. For use in later discussion, FIG. 4 illustrates an analog control system 400 of a traditional mass flowmeter. The sensors 48 each produce a voltage signal, with signal V_A0being produced by sensor 48aand signal V_B0being produced by sensor 48b. V_A0and V_B0correspond to the velocity of the loops relative to each other at the positions of the sensors. Prior to processing, signals V_B0and V_B0are amplified at respective input amplifiers 405 and 410 to produce signals V_A1and V_B1. To correct for imbalances in the amplifiers and the sensors, input amplifier 410 has a variable gain that is controlled by a balance signal coming from a feedback loop that contains a synchronous demodulator 415 and an integrator 420.

[0102] At the output of amplifier 405, signal V_A1is of the form:

[0103] V_A1=V_Dsin ωt+V_Ccos ωt,

[0104] and, at the output of amplifier 410, signal V_B1is of the form:

[0105] V_B1=−V_Dsin ωt+V_Ccos ωt,

[0106] where V_Dand V_Care, respectively, the drive voltage and the Coriolis voltage, and ω is the drive mode angular frequency.

[0107] Voltages V_A1and V_B1are differenced by operational amplifier 425 to produce:

[0108] V_DRV=V_A1−V_B1=2V_Dsin ωt,

[0109] where V_DRVcorresponds to the drive motion and is used to power the drivers. In addition to powering the drivers, V_DRVis supplied to a positive going zero crossing detector 430 that produces an output square wave F_DRVhaving a frequency corresponding to that of V_DRV(ω=2πF_DRV). F_DRVis used as the input to a digital phase locked loop circuit 435. F_DRValso is supplied to a processor 440.

[0110] Voltages V_A1and V_B1are summed by operational amplifier 445 to produce:

[0111] V_COR=V_A1+V_B1=2V_Ccos ωt,

[0112] where V_CORis related to the induced Coriolis motion.

[0113] V_CORis supplied to a synchronous demodulator 450 that produces an output voltage V_Mthat is directly proportional to mass by rejecting the components of V_CORthat do not have the same frequency as, and are not in phase with, a gating signal Q. The phase locked loop circuit 435 produces Q, which is a quiadrature reference signal that has the same frequency (ω) as V_DRVand is 90° out of phase with V_DRV(i.e., in phase with V_COR). Accordingly, synchronous demodulator 450 rejects frequencies other than ω so that V_Mcorresponds to the amplitude of V_CORat ω. This amplitude is directly proportional to the mass in the conduit.

[0114] V_Mis supplied to a voltage-to-frequency converter 455 that produces a square wave signal F_Mhaving a frequency that corresponds to the amplitude of V_M. The processor 440 then divides F_Mby F_DRVto produce a measurement of the mass flow rate.

[0115] Digital phase locked loop circuit 435 also produces a reference signal I that is in phase with V_DRVand is used to gate the synchronous demodulator 415 in the feedback loop controlling amplifier 410. When the gains of the input amplifiers 405 and 410 multiplied by the drive components of the corresponding input signals are equal, the summing operation at operational amplifier 445 produces zero drive component (i.e., no signal in phase with V_DRV) in the signal V_COR. When the gains of the input amplifiers 405 and 410 are not equal, a drive component exists in V_COR. This drive component is extracted by synchronous demodulator 415 and integrated by integrator 420 to generate an error voltage that corrects the gain of input amplifier 410. When the gain is too high or too low, the synchronous demodulator 415 produces an output voltage that causes the integrator to change the error voltage that modifies the gain. When the gain reaches the desired value, the output of the synchronous modulator goes to zero and the error voltage stops changing to maintain the gain at the desired value.

[0116] 2. Digital Control System

[0117] FIG. 5 provides a block diagram of an implementation 500 of the digital mass flowmeter 100 that includes the conduit 120, drivers 46, and sensors 48 of FIGS. 2A and 2B, along with a digital controller 505. Analog signals from the sensors 48 are converted to digital signals by analog-to-digital (“A/D”) converters 510 and supplied to the controller 505. The A/D converters may be implemented as separate converters, or as separate channels of a single converter.

[0118] Digital-to-analog (“D/A”) converters 515 convert digital control signals from the controller 505 to analog signals for driving the drivers 46. The use of a separate drive signal for each driver has a number of advantages. For example, the system may easily switch between symmetrical and antisymmetrical drive modes for diagnostic purposes. In other implementations, the signals produced by converters 515 may be amplified by amplifiers prior to being supplied to the drivers 46. In still other implementations, a single D/A converter may be used to produce a drive signal applied to both drivers, with the drive signal being inverted prior to being provided to one of the drivers to drive the conduit 120 in the antisymmetrical mode.

[0119] High precision resistors 520 and amplifiers 525 are used to measure the current supplied to each driver 46. A/D converters 530 convert the measured current to digital signals and supply the digital signals to controller 505. The controller 505 uses the measured currents in generating the driving signals.

[0120] Temperature sensors 535 and pressure sensors 540 measure, respectively, the temperature and the pressure at the inlet 545 and the outlet 550 of the conduit. A/D converters 555 convert the measured values to digital signals and supply the digital signals to the controller 505. The controller 505 uses the measured values in a number of ways. For example, the difference between the pressure measurements may be used to determine a back pressure in the conduit. Since the stiffness of the conduit varies with the back pressure, the controller may account for conduit stiffness based on the determined back pressure.

[0121] An additional temperature sensor 560 measures the temperature of the crystal oscillator 565 used by the A/D converters. An A/D converter 570 converts this temperature measurement to a digital signal for use by the controller 505. The input/output relationship of the A/D converters varies with the operating frequency of the converters, and the operating frequency varies with the temperature of the crystal oscillator. Accordingly, the controller uses the temperature measurement to adjust the data provided by the A/D converters, or in system calibration.

[0122] In the implementation of FIG. 5, the digital controller 505 processes the digitized sensor signals produced by the A/D converters 510 according to the procedure 600 illustrated in FIG. 6 to generate the mass flow measurement and the drive signal supplied to the drivers 46. Initially, the controller collects data from the sensors (step 605). Using this data, the controller determines the frequency of the sensor signals (step 610), eliminates zero offset from the sensor signals (step 615), and determines the amplitude (step 620) and phase (step 625) of the sensor signals. The controller uses these calculated values to generate the drive signal (step 630) and to generate the mass flow and other measurements (step 635). After generating the drive signals and measurements, the controller collects a new set of data and repeats the procedure. The steps of the procedure 600 may be performed serially or in parallel, and may be performed in varying order.

[0123] Because of the relationships between frequency, zero offset, amplitude, and phase, an estimate of one may be used in calculating another. This leads to repeated calculations to improve accuracy. For example, an initial frequency determination used in determining the zero offset in the sensor signals may be revised using offset-eliminated sensor signals. In addition, where appropriate, values generated for a cycle may be used as starting estimates for a following cycle.

[0124] FIG. 5 provides a general description of the hardware included in a digital flowmeter. A more detailed description of a specific hardware implementation is provided in the attached appendix, which is incorporated by reference.

[0125] a. Data Collection

[0126] For ease of discussion, the digitized signals from the two sensors will be referred to as signals SV₁and SV₂, with signal SV₁coming from sensor 48aand signal SV₂coming from sensor 48b. Although new data is generated constantly, it is assumed that calculations are based upon data corresponding to one complete cycle of both sensors. With sufficient data buffering, this condition will be true so long as the average time to process data is less than the time taken to collect the data. Tasks to be carried out for a cycle include deciding that the cycle has been completed, calculating the frequency of the cycle (or the frequencies of SV₁and SV₂), calculating the amplitudes of SV₁and SV₂, and calculating the phase difference between SV₁and SV₂. In some implementations, these calculations are repeated for each cycle using the end point of the previous cycle as the start for the next. In other implementations, the cycles overlap by 180° or other amounts (e.g., 90°) so that a cycle is subsumed within the cycles that precede and follow it.

[0127] FIGS. 7A and 7B illustrate two vectors of sampled data from signals SV₁and SV₂, which are named, respectively, sv1_in and sv2_in. The first sampling point of each vector is known, and corresponds to a zero crossing of the sine wave represented by the vector. For sv1_in, the first sampling point is the zero crossing from a negative value to a positive value, while for sv2_in the first sampling point is the zero crossing from a positive value to a negative value.

[0128] An actual starting point for a cycle (i.e., the actual zero crossing) will rarely coincide exactly with a sampling point. For this reason, the initial sampling points (start_sample_SV1 and start_sample_SV2) are the sampling points occurring just before the start of the cycle. To account for the difference between the first sampling point and the actual start of the cycle, the approach also uses the position (start_offset_SV1 or start_offset_SV2) between the starting sample and the next sample at which the cycle actually begins.

[0129] Since there is a phase offset between signals SV₁and SV₂, sv1_in and sv2_in may start at different sampling points. If both the sample rate and the phase difference are high, there may be a difference of several samples between the start of sv1_in and the start of sv2_in. This difference provides a crude estimate of the phase offset, and may be used as a check on the calculated phase offset, which is discussed below. For example, when sampling at 55 kHz, one sample corresponds to approximately 0.5 degrees of phase shift, and one cycle corresponds to about 800 sample points.

[0130] When the controller employs functions such as the sum (A+B) and difference (A−B), with B weighted to have the same amplitude as A, additional variables (e.g., start_sample_sum and start_offset_sum) track the start of the period for each function. The sum and difference functions have a phase offset halfway between SV₁and SV₂.

[0131] In one implementation, the data structure employed to store the data from the sensors is a circular list for each sensor, with a capacity of at least twice the maximum number of samples in a cycle. With this data structure, processing may be carried out on data for a current cycle while interrupts or other techniques are used to add data for a following cycle to the lists.

[0132] Processing is performed on data corresponding to a full cycle to avoid errors when employing approaches based on sine-waves. Accordingly, the first task in assembling data for a cycle is to determine where the cycle begins and ends. When nonoverlapping cycles are employed, the beginning of the cycle may be identified as the end of the previous cycle. When overlapping cycles are employed, and the cycles overlap by 180°, the beginning of the cycle may be identified as the midpoint of the previous cycle, or as the endpoint of the cycle preceding the previous cycle.

[0133] The end of the cycle may be first estimated based on the parameters of the previous cycle and under the assumption that the parameters will not change by more than a predetermined amount from cycle to cycle. For example, five percent may be used as the maximum permitted change from the last cycle's value, which is reasonable since, at sampling rates of 55 kHz, repeated increases or decreases of five percent in amplitude or frequency over consecutive cycles would result in changes of close to 5,000 percent in one second.

[0134] By designating five percent as the maximum permissible increase in amplitude and frequency, and allowing for a maximum phase change of 5° in consecutive cycles, a conservative estimate for the upper limit on the end of the cycle for signal SV₁may be determined as: $\mathrm{end\_sample}\mathrm{\_SV1}\le \mathrm{start\_sample}\mathrm{\_SV1}+\frac{365}{360}*\frac{\mathrm{sample\_rate}}{\mathrm{est\_freq}*0.95}$

[0135] where start_sample_SV1 is the first sample of sv1_in, sample_rate is the sampling rate, and est_freq is the frequency from the previous cycle. The upper limit on the end of the cycle for signal SV₂(end_sample_SV2) may be determined similarly.

[0136] After the end of a cycle is identified, simple checks may be made as to whether the cycle is worth processing. A cycle may not be worth processing when, for example, the conduit has stalled or the sensor waveforms are severely distorted. Processing only suitable cycles provides considerable reductions in computation.

[0137] One way to determine cycle suitability is to examine certain points of a cycle to confirm expected behavior. As noted above, the amplitudes and frequency of the last cycle give useful starting estimates of the corresponding values for the current cycle. Using these values, the points corresponding to 30°, 150°, 210° and 330° of the cycle may be examined. If the amplitude and frequency were to match exactly the amplitude and frequency for the previous cycle, these points should have values corresponding to est_amp/2, est_amp/2, −est_amp/2, and −est_amp/2, respectively, where est_amp is the estimated amplitude of a signal (i.e., the amplitude from the previous cycle). Allowing for a five percent change in both amplitude and frequency, inequalities may be generated for each quarter cycle. For the 30° point, the inequality is $\mathrm{sv1\_in}\left(\mathrm{start\_sample}\mathrm{\_SV1}+\frac{30}{360}*\frac{\mathrm{sample\_}\mathrm{rate}}{\mathrm{est\_freq}*1.05}\right)>0.475*\mathrm{est\_amp}\mathrm{\_SV}$

[0138] The inequalities for the other points have the same form, with the degree offset term (x/360) and the sign of the est_amp_SV1 term having appropriate values. These inequalities can be used to check that the conduit has vibrated in a reasonable manner.

[0139] Measurement processing takes place on the vectors sv1_in(start:end) and sv2_in(start:end) where:

[0140] start=min(start_sample _SV1, start_sample_SV2), and end=max(end_sample_SV1, end_sample_SV2).

[0141] The difference between the start and end points for a signal is indicative of the frequency of the signal.

[0142] b. Frequency Determination

[0143] The frequency of a discretely-sampled pure sine wave may be calculated by detecting the transition between periods (i.e., by detecting positive or negative zero-crossings) and counting the number of samples in each period. Using this method, sampling, for example, an 82.2 Hz sine wave at 55 kHz will provide an estimate of frequency with a maximum error of 0.15 percent. Greater accuracy may be achieved by estimating the fractional part of a sample at which the zero-crossing actually occurred using, for example, start_offset_SV1 and start_offset_SV2. Random noise and zero offset may reduce the accuracy of this approach.

[0144] As illustrated in FIGS. 8A and 8B, a more refined method of frequency determination uses quadratic interpolation of the square of the sine wave. With this method, the square of the sine wave is calculated, a quadratic function is fitted to match the minimum point of the squared sine wave, and the zeros of the quadratic function are used to determine the frequency. If

[0145] sv_t=A sin x_t+δ+σε_t,

[0146] where sv_tis the sensor voltage at time t, A is the amplitude of oscillation, x_tis the radian angle at time t (i.e., x_t=2πft), δ is the zero offset, ε_tis a random variable with distribution N(0,1), and σ is the variance of the noise, then the squared function is given by:

[0147] sv_t²=A²sin²x_t+2A(δ+σε_t) sin x_t+2δσε_t+δ²+σ²ε_t².

[0148] When x_tis close to 2π, sin x_tand sin²x_tcan be approximated as x_0t=x_t−2π and x_0t², respectively. Accordingly, for values of x_tclose to 2π, a_tcan be approximated as: $\begin{array}{c}{a}_t^2\approx A^2{x}_{0t}^2+2A\left(\delta +{\mathrm{\sigma ϵ}}_t\right)x_{0t}+2{\mathrm{\delta \sigma ϵ}}_t+{\delta }^2+{\sigma }^2{ϵ}_t^2\\ \approx \left(A^2{x}_{0t}^2+2A\delta x_{0t}+{\delta }^2\right)+{\mathrm{\sigma ϵ}}_t\left(2A{x}_{0t}+2\delta +{\mathrm{\sigma ϵ}}_t\right).\end{array}$

[0149] This is a pure quadratic (with a non-zero minimum, assuming δ≠0) plus noise, with the amplitude of the noise being dependent upon both σ and δ. Linear interpolation also could be used.

[0150] Error sources associated with this curve fitting technique are random noise, zero offset, and deviation from a true quadratic. Curve fitting is highly sensitive to the level of random noise. Zero offset in the sensor voltage increases the amplitude of noise in the sine-squared function, and illustrates the importance of zero offset elimination (discussed below). Moving away from the minimum, the square of even a pure sine wave is not entirely quadratic. The most significant extra term is of fourth order. By contrast, the most significant extra term for linear interpolation is of third order.

[0151] Degrees of freedom associated with this curve fitting technique are related to how many, and which, data points are used. The minimum is three, but more may be used (at greater computational expense) by using least-squares fitting. Such a fit is less susceptible to random noise. FIG. 8A illustrates that a quadratic approximation is good up to some 20° away from the minimum point. Using data points further away from the minimum will reduce the influence of random noise, but will increase the errors due to the non-quadratic tennis (i.e., fourth order and higher) in the sine-squared function.

[0152] FIG. 9 illustrates a procedure 900 for performing the curve fitting technique. As a first step, the controller initializes variables (step 905). These variables include end_point, the best estimate of the zero crossing point; ep_int, the integer value nearest to end_point; s[0 . . . i], the set of all sample points; z[k], the square of the sample point closest to end_point; z[0 . . . n−1], a set of squared sample points used to calculate end_point; n, the number of sample points used to calculate end_point (n=2k+1); step_length, the number of samples in s between consecutive values in z; and iteration_count, a count of the iterations that the controller has performed.

[0153] The controller then generates a first estimate of end_point (step 910). The controller generates this estimate by calculating an estimated zero-crossing point based on the estimated frequency from the previous cycle and searching around the estimated crossing point (forwards and backwards) to find the nearest true crossing point (i.e., the occurrence of consecutive samples with different signs). The controller then sets end_point equal to the sample point having the smaller magnitude of the samples surrounding the true crossing point.

[0154] Next, the controller sets n, the number of points for curve fitting (step 915). The controller sets n equal to 5 for a sample rate of 11 kHz, and to 21 for a sample rate of 44 kHz. The controller then sets iteration_count to 0 (step 920) and increments iteration_count (step 925) to begin the iterative portion of the procedure.

[0155] As a first step in the iterative portion of the procedure, the controller selects step_length (step 930) based on the value of iteration_count. The controller sets step_length equal to 6, 3, or 1 depending on whether iteration_count equals, respectively, 1, 2 or 3.

[0156] Next, the controller determines ep_int as the integer portion of the sum of end_point and 0.5 (step 935) and fills the z array (step 940). For example, when n equals 5, z[0]=s[ep_int-2*step_length]², z[1]=s[ep_int-step_length]², z[2]=s[ep_int]², z[3]=s[ep_int+step_length]², and z[4]=s[ep_int+2*step_length]².

[0157] Next, the controller uses a filter, such as a Savitzky-Golay filter, to calculate smoothed values of z[k−1], z[k] and z[k+1] (step 945). Savitzky-Golay smoothing filters are discussed by Press et al. in Numerical Recipes in C, pp. 650-655 (2nd ed., Cambridge University Press, 1995), which is incorporated by reference. The controller then fits a quadratic to z[k−1], z[k] and z[k+1] (step 950), and calculates the minimum value of the quadratic (z*) and the corresponding position (x*) (step 955).

[0158] If x* is between the points corresponding to k−1 and k+1 (step 960), then the controller sets end_point equal to x* (step 965). Thereafter, if iteration_count is less than 3 (step 970), the controller increments iteration_count (step 925) and repeats the iterative portion of the procedure.

[0159] If x* is not between the points corresponding to k−1 and k+1 (step 960), or if iteration_count equals 3 (step 970), the controller exits the iterative portion of the procedure. The controller then calculates the frequency based on the difference between end_point and the starting point for the cycle, which is known (step 975).

[0160] In essence, the procedure 900 causes the controller to make three attempts to home in on end_point, using smaller step_lengths in each attempt. If the resulting minimum for any attempt falls outside of the points used to fit the curve (i.e., there has been extrapolation rather than interpolation), this indicates that either the previous or new estimate is poor, and that a reduction in step size is unwarranted.

[0161] The procedure 900 may be applied to at least three different sine waves produced by the sensors. These include signals SV₁and SV₂and the weighted sum of the two. Moreover, assuming that zero offset is eliminated, the frequency estimates produced for these signals are independent. This is clearly true for signals SV₁and SV₂, as the errors on each are independent. It is also true, however, for the weighted sum, as long as the mass flow and the corresponding phase difference between signals SV₁and SV₂are large enough for the calculation of frequency to be based on different samples in each case. When this is true, the random errors in the frequency estimates also should be independent.

[0162] The three independent estimates of frequency can be combined to provide an improved estimate. This combined estimate is simply the mean of the three frequency estimates.

[0163] c. Zero Offset Compensation

[0164] An important error source in a Coriolis transmitter is zero offset in each of the sensor voltages. Zero offset is introduced into a sensor voltage signal by drift in the pre-amplification circuitry and the analog-to-digital converter. The zero offset effect may be worsened by slight differences in the pre-amplification gains for positive and negative voltages due to the use of differential circuitry. Each error source varies between transmitters, and will vary with transmitter temperature and more generally over time with component wear.

[0165] An example of the zero offset compensation technique employed by the controller is discussed in detail below. In general, the controller uses the frequency estimate and an integration technique to determine the zero offset in each of the sensor signals. The controller then eliminates the zero offset from those signals. After eliminating zero offset from signals SV₁and SV₂, the controller may recalculate the frequency of those signals to provide an improved estimate of the frequency.

[0166] d. Amplitude Determination

[0167] The amplitude of oscillation has a variety of potential uses. These include regulating conduit oscillation via feedback, balancing contributions of sensor voltages when synthesizing driver waveforms, calculating sums and differences for phase measurement, and calculating an amplitude rate of change for measurement correction purposes.

[0168] In one implementation, the controller uses the estimated amplitudes of signals SV₁and SV₂to calculate the sum and difference of signals SV₁and SV₂, and the product of the sum and difference. Prior to determining the sum and difference, the controller compensates one of the signals to account for differences between the gains of the two sensors. For example, the controller may compensate the data for signal SV₂based on the ratio of the amplitude of signal SV₁to the amplitude of signal SV₂so that both signals have the same amplitude.

[0169] The controller may produce an additional estimate of the frequency based on the calculated sum. This estimate may be averaged with previous frequency estimates to produce a refined estimate of the frequency of the signals, or may replace the previous estimates.

[0170] The controller may calculate the amplitude according to a Fourier-based technique to eliminate the effects of higher harmonics. A sensor voltage x(t) over a period T (as identified using zero crossing techniques) can be represented by an offset and a series of harmonic terms as:

[0171] x(t)=a₀/2+a₁cos(ωt)+a₂cos(2ωt)+a₃cos(3ωt)+ . . . +b₁sin(ωt)+b₂sin(2ωt)+ . . .

[0172] With this representation, a non-zero offset a₀will result in non-zero cosine terms a_n. Though the amplitude of interest is the amplitude of the fundamental component (i.e., the amplitude at frequency ω), monitoring the amplitudes of higher harmonic components (i.e., at frequencies kω, where k is greater than 1) may be of value for diagnostic purposes. The values of a_nand b_nmay be calculated as: $a_n=\frac{2}{T}{\int }_0^{T}x\left(t\right)\cos n\omega dt,\mathrm{and}$ $b_n=\frac{2}{T}{\int }_0^{T}x\left(t\right)\sin n\omega dt.$

[0173] The amplitude, A_n, of each harmonic is given by:

[0174] A_n={square root over (a_n²+L +b_n²+L )}.

[0175] The integrals are calculated using Simpson's method with quadratic correction (described below). The chief computational expense of the method is calculating the pure sine and cosine functions.

[0176] e. Phase Determination

[0177] The controller may use a number of approaches to calculate the phase difference between signals SV₁and SV₂. For example, the controller may determine the phase offset of each harmonic, relative to the starting time at t=0, as: ${\varphi }_n={\tan }^{-1}\frac{a_n}{b_n}.$

[0178] The phase offset is interpreted in the context of a single waveform as being the difference between the start of the cycle (i.e., the zero-crossing point) and the point of zero phase for the component of SV(t) of frequency ω. Since the phase offset is an average over the entire waveform, it may be used as the phase offset from the midpoint of the cycle. Ideally, with no zero offset and constant amplitude of oscillation, the phase offset should be zero every cycle. The controller may determine the phase difference by comparing the phase offset of each sensor voltage over the same time period.

[0179] The amplitude and phase may be generated using a Fourier method that eliminates the effects of higher harmonics. This method has the advantage that it does not assume that both ends of the conduits are oscillating at the same frequency. As a first step in the method, a frequency estimate is produced using the zero crossings to measure the time between the start and end of the cycle. If linear variation in frequency is assumed, this estimate equals the time-averaged frequency over the period. Using the estimated, and assumed time-invariant, frequency ω of the cycle, the controller calculates: $I_1=\frac{2\omega }{\pi }{\int }_0^{\frac{2\pi }{\omega }}SV\left(t\right)\sin \left(\omega t\right)dt,\mathrm{and}$ $I_2=\frac{2\omega }{\pi }{\int }_0^{\frac{2\pi }{\omega }}SV\left(t\right)\cos \left(\omega t\right)dt,$

[0180] where SV(t) is the sensor voltage waveform (i.e., SV₁(t) or SV₂(t)). The controller then determines the estimates of the amplitude and phase: $\begin{array}{c}\mathrm{Amp}=\sqrt{I_1^2+I_2^2},\mathrm{and}\\ \mathrm{Phase}={\tan }^{-1}\frac{I_2}{I_1}.\end{array}$

[0181] The controller then calculates a phase difference, assuming that the average phase and frequency of each sensor signal is representative of the entire waveform. Since these frequencies are different for SV₁and SV₂, the corresponding phases are scaled to the average frequency. In addition, the phases are shifted to the same starting point (i.e., the midpoint of the cycle on SV₁). After scaling, they are subtracted to provide the phase difference: $\begin{array}{c}\mathrm{scaled\_phase}{\mathrm{\_SV}}_1={\mathrm{phase\_SV}}_1\frac{\mathrm{av\_freq}}{{\mathrm{freq\_SV}}_1},\\ \mathrm{scale\_shift}{\mathrm{\_SV}}_2=\frac{\left({\mathrm{midpoint\_SV}}_2-{\mathrm{midpointt\_SV}}_1\right)h{\mathrm{freq\_SV}}_2}{360},\mathrm{and}\\ \mathrm{scaled\_phase}{\mathrm{\_SV}}_2=\left({\mathrm{phase\_SV}}_2+\mathrm{scale\_shift}{\mathrm{\_SV}}_2\right)\frac{\mathrm{av\_freq}}{{\mathrm{freq\_SV}}_2},\end{array}$

[0182] where h is the sample length and the midpoints are defined in terms of samples: ${\mathrm{midpoint\_SV}}_x=\frac{\left({\mathrm{startpoint\_SV}}_x+{\mathrm{endpoint\_SV}}_x\right)}{2}.$

[0183] In general, phase and amplitude are not calculated over the same time-frame for the two sensors. When the flow rate is zero, the two cycle mid-points are coincident. However, they diverge at high flow rates so that the calculations are based on sample sets that are not coincident in time. This leads to increased phase noise in conditions of changing mass flow. At full flow rate, a phase shift of 4° (out of 360°) means that only 99% of the samples in the SV₁and SV₂data sets are coincident. Far greater phase shifts may be observed under aerated conditions, which may lead to even lower rates of overlap.

[0184] FIG. 10 illustrates a modified approach 1000 that addresses this issue. First, the controller finds the frequencies (f₁, f₂) and the mid-points (m₁, m₂) of the SV₁and SV₂data sets (d₁, d₂) (step 1005). Assuming linear shift in frequency from the last cycle, the controller calculates the frequency of SV₂at the midpoint of SV₁(f_2m1) and the frequency of SV₁at the midpoint of SV₂(f_1m2) (step 1010).

[0185] The controller then calculates the starting and ending points of new data sets (d_1m2and d_2m1) with mid-points m₂and m₁respectively, and assuming frequencies of f_1m2and f_2m1(step 1015). These end points do not necessarily coincide with zero crossing points. However, this is not a requirement for Fourier-based calculations.

[0186] The controller then carries out the Fourier calculations of phase and amplitude on the sets d₁and d_2m1, and the phase difference calculations outlined above (step 1020). Since the mid-points of d₁and d_2m1are identical, scale-shift_SV₂is always zero and can be ignored. The controller repeats these calculations for the data sets d₂and d_1m2(step 1025). The controller then generates averages of the calculated amplitude and phase difference for use in measurement generation (step 1030). When there is sufficient separation between the mid points m₁and m₂, the controller also may use the two sets of results to provide local estimates of the rates of change of phase and amplitude.

[0187] The controller also may use a difference-amplitude method that involves calculating a difference between SV₁and SV₂, squaring the calculated difference, and integrating the result. According to another approach, the controller synthesizes a sine wave, multiplies the sine wave by the difference between signals SV₁and SV₂, and integrates the result. The controller also may integrate the product of signals SV₁and SV₂, which is a sine wave having a frequency 2f (where f is the average frequency of signals SV₁and SV₂), or may square the product and integrate the result. The controller also may synthesize a cosine wave comparable to the product sine wave and multiply the synthesized cosine wave by the product sine wave to produce a sine wave of frequency 4f that the controller then integrates. The controller also may use multiple ones of these approaches to produce separate phase measurements, and then may calculate a mean value of the separate measurements as the final phase measurement.

[0188] The difference-amplitude method starts with: ${\mathrm{SV}}_1\left(t\right)=A_1\sin \left(2\pi ft+\frac{\varphi }{2}\right)\mathrm{and}{\mathrm{SV}}_2\left(t\right)=A_2\sin \left(2\pi ft-\frac{\varphi }{2}\right),$

[0189] where φ is the phase difference between the sensors. Basic trigonometric identities may be used to define the sum (Sum) and difference (Diff) between the signals as: $\begin{array}{c}\mathrm{Sum}\equiv {\mathrm{SV}}_1\left(t\right)+\frac{A_1}{A_2}{\mathrm{SV}}_2\left(t\right)=2A_1\cos \frac{\varphi }{2}·\sin 2\pi ft,\mathrm{and}\\ \mathrm{Diff}\equiv {\mathrm{SV}}_1\left(t\right)-\frac{A_1}{A_2}{\mathrm{SV}}_2\left(t\right)=2A_1\sin \frac{\varphi }{2}·\cos 2\pi ft.\end{array}$

[0190] These functions have amplitudes of 2A₁cos(φ/2) and 2A₁sin(φ/2), respectively. The controller calculates data sets for Sum and Diff from the data for SV₁and SV₂, and then uses one or more of the methods described above to calculate the amplitude of the signals represented by those data sets. The controller then uses the calculated amplitudes to calculate the phase difference, φ.

[0191] As an alternative, the phase difference may be calculated using the function Prod, defined as: $\begin{array}{c}\mathrm{Prod}\equiv \mathrm{Sum}\times \mathrm{Diff}=4A_1^2\cos \frac{\varphi }{2}·\sin \frac{\varphi }{2}·\cos 2\pi ft·\sin 2\pi ft\\ =A_1^2\sin \varphi ·\sin 4\pi ft,\end{array}$

[0192] which is a function with amplitude A²sin φ and frequency 2f. Prod can be generated sample by sample, and φ may be calculated from the amplitude of the resulting sine wave.

[0193] The calculation of phase is particularly dependent upon the accuracy of previous calculations (i.e., the calculation of the frequencies and amplitudes of SV₁and SV₂). The controller may use multiple methods to provide separate (if not entirely independent) estimates of the phase, which may be combined to give an improved estimate.

[0194] f. Drive Signal Generation

[0195] The controller generates the drive signal by applying a gain to the difference between signals SV₁and SV₂. The controller may apply either a positive gain (resulting in positive feedback) or a negative gain (resulting in negative feedback).

[0196] In general, the Q of the conduit is high enough that the conduit will resonate only at certain discrete frequencies. For example, the lowest resonant frequency for some conduits is between 65 Hz and 95 Hz, depending on the density of the process fluid, and irrespective of the drive frequency. As such, it is desirable to drive the conduit at the resonant frequency to minimize cycle-to-cycle energy loss. Feeding back the sensor voltage to the drivers permits the drive frequency to migrate to the resonant frequency.

[0197] As an alternative to using feedback to generate the drive signal, pure sine waves having phases and frequencies determined as described above may be synthesized and sent to the drivers. This approach offers the advantage of eliminating undesirable high frequency components, such as harmonics of the resonant frequency. This approach also permits compensation for time delays introduced by the analog-to-digital converters, processing, and digital-to-analog converters to ensure that the phase of the drive signal corresponds to the mid-point of the phases of the sensor signals. This compensation may be provided by determining the time delay of the system components and introducing a phase shift corresponding to the time delay.

[0198] Another approach to driving the conduit is to use square wave pulses. This is another synthesis method, with fixed (positive and negative) direct current sources being switched on and off at timed intervals to provide the required energy. The switching is synchronized with the sensor voltage phase. Advantageously, this approach does not require digital-to-analog converters.

[0199] In general, the amplitude of vibration of the conduit should rapidly achieve a desired value at startup, so as to quickly provide the measurement function, but should do so without significant overshoot, which may damage the meter. The desired rapid startup may be achieved by setting a very high gain so that the presence of random noise and the high Q of the conduit are sufficient to initiate motion of the conduit. In one implementation, high gain and positive feedback are used to initiate motion of the conduit. Once stable operation is attained, the system switches to a synthesis approach for generating the drive signals.

[0200] Referring to FIGS. 11A-13D, synthesis methods also may be used to initiate conduit motion when high gain is unable to do so. For example, if the DC voltage offset of the sensor voltages is significantly larger than random noise, the application of a high gain will not induce oscillatory motion. This condition is shown in FIGS. 11A-11D, in which a high gain is applied at approximately 0.3 seconds. As shown in FIGS. 11A and 11B, application of the high gain causes one of the drive signals to assume a large positive value (FIG. 11A) and the other to assume a large negative value (FIG. 11B). The magnitudes of the drive signals vary with noise in the sensor signals (FIGS. 11C and 11D). However, the amplified noise is insufficient to vary the sign of the drive signals so as to induce oscillation.

[0201] FIGS. 12A-12D illustrate that imposition of a square wave over several cycles can reliably cause a rapid startup of oscillation. Oscillation of a conduit having a two inch diameter may be established in approximately two seconds. The establishment of conduit oscillation is indicated by the reduction in the amplitude of the drive signals, as shown in FIGS. 12A and 12B. FIGS. 13A-13D illustrate that oscillation of a one inch conduit may be established in approximately half a second.

[0202] A square wave also may be used during operation to correct conduit oscillation problems. For example, in some circumstances, flow meter conduits have been known to begin oscillating at harmonics of the resonant frequency of the conduit, such as frequencies on the order of 1.5 kHz. When such high frequency oscillations are detected, a square wave having a more desirable frequency may be used to return the conduit oscillation to the resonant frequency.

[0203] g. Measurement Generation

[0204] The controller digitally generates the mass flow measurement in a manner similar to the approach used by the analog controller. The controller also may generate other measurements, such as density.

[0205] In one implementation, the controller calculates the mass flow based on the phase difference in degrees between the two sensor signals (phase_diff), the frequency of oscillation of the conduit (freq), and the process temperature (temp):

[0206] T_z=temp−T_c,

[0207] noneu_mf=tan(π*phase_diff/180), and

[0208] massflow=16(MF₁*T_z²+MF₂*T_z+MF_3)*noneu_mf/freq,

[0209] where T_cis a calibration temperature, MF₁-MF₃are calibration constants calculated during a calibration procedure, and noneu_mf is the mass flow in non-engineering units.

[0210] The controller calculates the density based on the frequency of oscillation of the conduit and the process temperature:

[0211] T_z=temp−T_c,

[0212] c₂=freq², and

[0213] density=(D₁*T_z²+D₂*T_z+D₃)/c₂+D₄*T_z²,

[0214] where D₁-D₄are calibration constants generated during a calibration procedure.

[0215] D. Integration Techniques

[0216] Many integration techniques are available, with different techniques requiring different levels of computational effort and providing different levels of accuracy. In the described implementation, variants of Simpson's method are used. The basic technique may be expressed as: ${\int }_{t_n}^{t_{n+2}}ydt\approx \frac{h}{3}\left(y_n+4y_{n+1}+y_{n+2}\right),$

[0217] where t_kis the time at sample k, y_kis the corresponding function value, and h is the step length. This rule can be applied repeatedly to any data vector with an odd number of data points (i.e., three or more points), and is equivalent to fitting and integrating a cubic spline to the data points. If the number of data points happens to be even, then the so-called ⅜ths rule can be applied at one end of the interval: ${\int }_{t_n}^{t_{n+3}}ydt\approx \frac{3h}{8}\left(y_n+3y_{n+1}+3y_{n+2}+y_{n+3}\right).$

[0218] As stated earlier, each cycle begins and ends at some offset (e.g., start_offset_SV1) from a sampling point. The accuracy of the integration techniques are improved considerably by taking these offsets into account. For example, in an integration of a half cycle sine wave, the areas corresponding to partial samples must be included in the calculations to avoid a consistent underestimate in the result.

[0219] Two types of function are integrated in the described calculations: either sine or sine-squared functions. Both are easily approximated close to zero where the end points occur. At the end points, the sine wave is approximately linear and the sine-squared function is approximately quadratic.

[0220] In view of these two types of functions, three different integration methods have been evaluated. These are Simpson's method with no end correction, Simpson's method with linear end correction, and Simpson's method with quadratic correction.

[0221] The integration methods were tested by generating and sampling pure sine and sine-squared functions, without simulating any analog-to-digital truncation error. Integrals were calculated and the results were compared to the true amplitudes of the signals. The only source of error in these calculations was due to the integration techniques. The results obtained are illustrated in tables A and B.

[0222] For sine functions, Simpson's method with linear correction was unbiased with the smallest standard deviation, while Simpson's method without correction was biased to a negative error and Simpson's method with quadratic correction had a relatively high standard deviation. For sine-squared functions, the errors were generally reduced, with the quadratic correction providing the best result. Based on these evaluations, linear correction is used when integrating sine functions and quadratic correction is used when integrating sine-squared functions.

[0223] E. Synchronous Modulation Technique

[0224] FIG. 14 illustrates an alternative procedure 1400 for processing the sensor signals. Procedure 1400 is based on synchronous modulation, such as is described by Denys et al., in “Measurement of Voltage Phase for the French Future Defence Plan Against Losses of Synchronism”, IEEE Transactions on Power Delivery, 7(1), 62-69, 1992 and by Begovic et al. in “Frequency Tracking in Power Networks in the Presence of Harmonics”, IEEE Transactions on Power Delivery, 8(2), 480-486, 1993, both of which are incorporated by reference.

[0225] First, the controller generates an initial estimate of the nominal operating frequency of the system (step 1405). The controller then attempts to measure the deviation of the frequency of a signal x[k] (e.g., SV₁) from this nominal frequency:

[0226] x[k]=A sin [(ω₀+Δω)kh+φ]+ε(k),

[0227] where A is the amplitude of the sine wave portion of the signal, ω₀is the nominal frequency (e.g., 88 Hz), Δω is the deviation from the nominal frequency, h is the sampling interval, φ is the phase shift, and ε(k) corresponds to the added noise and harmonics.

[0228] To generate this measurement, the controller synthesizes two signals that oscillate at the nominal frequency (step 1410). The signals are phase shifted by 0 and π/2 and have amplitude of unity. The controller multiplies each of these signals by the original signal to produce signals y₁and y₂(step 1415): $\begin{array}{c}{y}_1=x\left[k\right]\cos \left({\omega }_0\mathrm{kh}\right)=\frac{A}{2}\sin \left[\left(2{\omega }_0+\mathrm{\Delta \omega }\right)\mathrm{kh}+\phi \right]+\frac{A}{2}\sin \left(\mathrm{\Delta \omega }\mathrm{kh}+\phi \right),\mathrm{and}\\ y_2=x\left[k\right]\sin \left({\omega }_0\mathrm{kh}\right)=\frac{A}{2}\cos \left[\left(2{\omega }_0+\mathrm{\Delta \omega }\right)\mathrm{kh}+\phi \right]+\frac{A}{2}\cos \left(\mathrm{\Delta \omega }\mathrm{kh}+\phi \right),\end{array}$

[0229] where the first terms of y₁and y₂are high frequency (e.g., 176 Hz) components and the second terms are low frequency (e.g., 0 Hz) components. The controller then eliminates the high frequency components using a low pass filter (step 1420): $y_1^{\prime }=\frac{A}{2}\sin \left(\mathrm{\Delta \omega }\mathrm{kh}+\phi \right)+{ϵ}_1\left[k\right],\mathrm{and}$ $y_2^{\prime }=\frac{A}{2}\cos \left(\mathrm{\Delta \omega }\mathrm{kh}+\phi \right)+{ϵ}_2\left[k\right],$

[0230] where ε₁[k] and ε₂[k] represent the filtered noise from the original signals. The controller combines these signals to produce u[k] (step 1425): $\begin{array}{c}u\left[k\right]=\left(y_1^{\prime }\left[k\right]+{\mathrm{jy}}_2^{\prime }\left[k\right]\right)\left(y_1^{\prime }\left[k-1\right]+{\mathrm{jy}}_2^{\prime }\left[k-1\right]\right)\\ =u_1\left[k\right]+{\mathrm{ju}}_2\left[k\right]\\ =\frac{A^2}{4}\cos \left(\mathrm{\Delta \omega }h\right)+j\frac{A^2}{4}\sin \left(\mathrm{\Delta \omega }h\right),\end{array}$

[0231] which carries the essential information about the frequency deviation. As shown, u₁[k] represents the real component of u[k], while u₂[k] represents the imaginary component.

[0232] The controller uses the real and imaginary components of u[k] to calculate the frequency deviation, Δf (step 1430): $\Delta f=\frac{1}{h}\arctan \frac{u_2\left[k\right]}{u_1\left[k\right]}.$

[0233] The controller then adds the frequency deviation to the nominal frequency (step 1435) to give the actual frequency:

[0234] f=Δf+f₀.

[0235] The controller also uses the real and imaginary components of u[k] to determine the amplitude of the original signal. In particular, the controller determines the amplitude as (step 1440):

[0236] A²=4{square root over (u₁²+L [k]+u₂²+L [k])}.

[0237] Next, the controller determines the phase difference between the two sensor signals (step 1445). Assuming that any noise (ε₁[k] and ε₂[k]) remaining after application of the low pass filter described below will be negligible, noise free versions of y₁′[k] and y₂′[k] (y₁*[k] and y₂*[k]) may be expressed as: $y_1^{*}\left[k\right]=\frac{A}{2}\sin \left(\Delta \omega kh+\phi \right),\mathrm{and}$ $y_2^{*}\left[k\right]=\frac{A}{2}\cos \left(\Delta \omega kh-\phi \right).$

[0238] Multiplying these signals together gives: $v=y_1^{*}{y}_2^{*}=\frac{A^2}{8}\left[\sin \left(2\phi \right)+\sin \left(2\mathrm{\Delta \omega }kh\right)\right].$

[0239] Filtering this signal by a low pass filter having a cutoff frequency near 0 Hz removes the unwanted component and leaves: $v^{\prime }=\frac{A^2}{8}\sin \left(2\phi \right),$

[0240] from which the phase difference can be calculated as: $\phi =\frac{1}{2}\arcsin \frac{8v^{\prime }}{A^2}.$

[0241] This procedure relies on the accuracy with which the operating frequency is initially estimated, as the procedure measures only the deviation from this frequency. If a good estimate is given, a very low filter can be used, which makes the procedure very accurate. For typical flowmeters, the operating frequencies are around 95 Hz (empty) and 82 Hz (full). A first approximation of half range (88 Hz) is used, which allows a low-pass filter cut-off of 13 Hz. Care must be taken in selecting the cut-off frequency as a very small cut-off frequency can attenuate the amplitude of the sine wave.

[0242] The accuracy of measurement also depends on the filtering characteristics employed. The attenuation of the filter in the dead-band determines the amount of harmonics rejection, while a smaller cutoff frequency improves the noise rejection.

[0243] F. Meter with PI Control

[0244] FIGS. 15A and 15B illustrate a meter 1500 having a controller 1505 that uses another technique to generate the signals supplied to the drivers. Analog-to-digital converters 1510 digitize signals from the sensors 48 and provide the digitized signals to the controller 1505. The controller 1505 uses the digitized signals to calculate gains for each driver, with the gains being suitable for generating desired oscillations in the conduit. The gains may be either positive or negative. The controller 1505 then supplies the gains to multiplying digital-to-analog converters 1515. In other implementations, two or more multiplying digital-to-analog converters arranged in series may be used to implement a single, more sensitive multiplying digital-to-analog converter.

[0245] The controller 1505 also generates drive signals using the digitized sensor signals. The controller 1505 provides these drive signals to digital-to-analog converters 1520 that convert the signals to analog signals that are supplied to the multiplying digital-to-analog converters 1515.

[0246] The multiplying digital-to-analog converters 1515 multiply the analog signals by the gains from the controller 1505 to produce signals for driving the conduit. Amplifiers 1525 then amplify these signals and supply them to the drivers 46. Similar results could be obtained by having the controller 1505 perform the multiplication performed by the multiplying digital-to-analog converter, at which point the multiplying digital-to-analog converter could be replaced by a standard digital-to-analog converter.

[0247] FIG. 15B illustrates the control approach in more detail. Within the controller 1505, the digitized sensor signals are provided to an amplitude detector 1550, which determines a measure, a(t), of the amplitude of motion of the conduit using, for example, the technique described above. A summer 1555 then uses the amplitude a(t) and a desired amplitude a₀to calculate an error e(t) as:

[0248] e(t)=a₀−a(t).

[0249] The error e(t) is used by a proportional-integral (“PI”) control block 1560 to generate a gain K₀(t). This gain is multiplied by the difference of the sensor signals to generate the drive signal. The PI control block permits high speed response to changing conditions. The amplitude detector 1550, summer 1555, and PI control block 1560 may be implemented as software processed by the controller 1505, or as separate circuitry.

[0250] 1. Control Procedure

[0251] The meter 1500 operates according to the procedure 1600 illustrated in FIG. 16. Initially, the controller receives digitized data from the sensors (step 1605). Thereafter, the procedure 1600 includes three parallel branches: a measurement branch 1610, a drive signal generation branch 1615, and a gain generation branch 1620.

[0252] In the measurement branch 1610, the digitized sensor data is used to generate measurements of amplitude, frequency, and phase, as described above (step 1625). These measurements then are used to calculate the mass flow rate (step 1630) and other process variables. In general, the controller 1505 implements the measurement branch 1610.

[0253] In the drive signal generation branch 1615, the digitized signals from the two sensors are differenced to generate the signal (step 1635) that is multiplied by the gain to produce the drive signal. As described above, this differencing operation is performed by the controller 1505. In general, the differencing operation produces a weighted difference that accounts for amplitude differences between the sensor signals.

[0254] In the gain generation branch 1620, the gain is calculated using the proportional-integral control block. As noted above, the amplitude, a(t), of motion of the conduit is determined (step 1640) and subtracted from the desired amplitude a₀(step 1645) to calculate the error e(t). Though illustrated as a separate step, generation of the amplitude, a(t), may correspond to generation of the amplitude in the measurement generation step 1625. Finally, the PI control block uses the error e(t) to calculate the gain (step 1650).

[0255] The calculated gain is multiplied by the difference signal to generate the drive signal supplied to the drivers (step 1655). As described above, this multiplication operation is performed by the multiplying D/A converter or may be performed by the controller.

[0256] 2. PI Control Block

[0257] The objective of the PI control block is to sustain in the conduit pure sinusoidal oscillations having an amplitude a₀. The behavior of the conduit may be modelled as a simple mass-spring system that may be expressed as:

[0258] {umlaut over (x)}+2ζω_n{dot over (x)}+ω_n²x=0,

[0259] where x is a function of time and the displacement of the mass from equilibrium, ω_nis the natural frequency, and ζ is a damping factor, which is assumed to be small (e.g., 0.001). The solution to this force equation as a function of an output y(t) and an input i(t) is analogous to an electrical network in which the transfer function between a supplied current, i(s), and a sensed output voltage, y(s), is: $\frac{y\left(s\right)}{i\left(s\right)}=\frac{ks}{s^2+2{\mathrm{\zeta \omega }}_{n}s+{\omega }_n^2}.$

[0260] To achieve the desired oscillation in the conduit, a positive-feedback loop having the gain K₀(t) is automatically adjusted by a ‘slow’ outer loop to give:

[0261] {umlaut over (x)}+(2ζω_n−kK₀(t)){dot over (x)}+ω_n²x=0.

[0262] The system is assumed to have a “two-time-scales” property, which means that variations in K₀(t) are slow enough that solutions to the equation for x provided above can be obtained by assuming constant damping.

[0263] A two-term PI control block that gives zero steady-state error may be expressed as: $K_0\left(t\right)=K_{p}e\left(t\right)+K_i{\int }_0^{t}e\left(t\right)dt,$

[0264] where the error, e(t) (i.e., a₀-a(t)), is the input to the PI control block, and K_pand K_iare constants. In one implementation, with a₀=10, controller constants of K_p=0.02 and K_i=0.0005 provide a response in which oscillations build up quickly. However, this PI control block is nonlinear, which may result in design and operational difficulties.

[0265] A linear model of the behavior of the oscillation amplitude may be derived by assuming that x(t) equals Aε^jωt, which results in:

[0266] {dot over (x)}={dot over (A)}e^jωt+jωe^jωt, and {umlaut over (x)}=[Ä−ω²A]e^jωt+2jω{dot over (A)}e^jωt.

[0267] Substituting these expressions into the expression for oscillation of the loop, and separating into real and imaginary terms, gives:

[0268] jω{2{dot over (A)}+(2 ζω_n−kK₀)A}=0, and Ä+(2ζω_n−kK₀){dot over (A)}+(ω_n²−ω²)A=0.

[0269] A(t) also may be expressed as: $\frac{\stackrel{.}{A}}{A}=-{\mathrm{\zeta \omega }}_n+\frac{k{K}_o}{2}t.$

[0270] A solution of this equation is: $\log A\left(t\right)=\left(-{\mathrm{\zeta \omega }}_n+\frac{k{K}_o}{2}\right)t.$

[0271] Transforming variables by defining a(t) as being equal to log A(t), the equation for A(t) can be written as: $\frac{da}{dt}=-{\mathrm{\zeta \omega }}_n+\frac{k{K}_o\left(t\right)}{2},$

[0272] where K₀is now explicitly dependent on time. Taking Laplace transforms results in: $a\left(s\right)=\frac{-{\mathrm{\zeta \omega }}_n-k{K}_o\left(s\right)/2}{s},$

[0273] which can be interpreted in tennis of transfer-functions as in FIG. 17. This figure is of particular significance for the design of controllers, as it is linear for all K₀and a, with the only assumption being the two-time-scales property. The performance of the closed-loop is robust with respect to this assumption so that fast responses which are attainable in practice can be readily designed.

[0274] From FIG. 17, the term ζω_n, is a “load disturbance” that needs to be eliminated by the controller (i.e., kK₀/2 must equal ζω_nfor a(t) to be constant). For zero steady-state error this implies that the outer-loop controller must have an integrator (or very large gain). As such, an appropriate PI controller, C(s), may be assumed to be K_p(1+1/sT_i), where T_iis a constant. The proportional term is required for stability. The term ζω_n, however, does not affect stability or controller design, which is based instead on the open-loop transfer function: $C\left(s\right)G\left(s\right)=\frac{a\left(s\right)}{e\left(s\right)}=\frac{k{K}_p\left(1+s{T}_i\right)}{2s^2{T}_i}=\frac{k{K}_p/2\left(s+1/T_i\right)}{s^2}.$

[0275] The root locus for varying K_pis shown in FIG. 18. For small K_p, there are slow underdamped roots. As K_pincreases, the roots become real at the point P for which the controller gain is K_p=8/(kT_i). Note in particular that the theory does not place any restriction upon the choice of T_i. Hence the response can, in principle, be made critically damped and as fast as desired by appropriate choices of K_pand T_i.

[0276] Although the poles are purely real at point P, this does not mean there is no overshoot in the closed-loop step response. This is most easily seen by inspecting the transfer function between the desired value, a₀, and the error e: $\frac{e\left(s\right)}{a_0\left(s\right)}=\frac{s^2}{s^2+0.5k{K}_p\left(s+1/T_i\right)}=\frac{s^2}{p_2\left(s\right)},$

[0277] where p₂is a second-order polynomial. With a step input, a₀(s)=α/s, the response can be written as αp′(t), where p(t) is the inverse transform of 1/p₂(s) and equals a₁exp(−λ₁t)+a₂exp(−λ₂t). The signal p(t) increases and then decays to zero so that e(t), which is proportional to p′, must change sign, implying overshoot in a(t). The set-point a₀may be prefiltered to give a pseudo set-point a₀*: $a_0^{*}\left(s\right)=\frac{1}{1+s{T}_i}{a}_0\left(s\right),$

[0278] where T_iis the known controller parameter. With this prefilter, real controller poles should provide overshoot-free step responses. This feature is useful as there may be physical constraints on overshoot (e.g., mechanical interference or overstressing of components).

[0279] The root locus of FIG. 18 assumes that the only dynamics are from the inner-loop's gain/log-amplitude transfer function (FIG. 16) and the outer-loop's PI controller C(s) (i.e., that the log-amplitude a=log A is measured instantaneously). However, A is the amplitude of an oscillation which might be growing or decaying and hence cannot in general be measured without taking into account the underlying sinusoid. There are several possible methods for measuring A, in addition to those discussed above. Some are more suitable for use in quasi-steady conditions. For example, a phase-locked loop in which a sinusoidal signal s(t)=sin(ω_nt+φ₀) locks onto the measured waveform y(t)=A(t) sin(ω_nt+φ₁) may be employed. Thus, a measure of the amplitude a=log A is given by dividing these signals (with appropriate safeguards and filters). This method is perhaps satisfactory near the steady-state but not for start-up conditions before there is a lock.

[0280] Another approach uses a peak-follower that includes a zero-crossing detector together with a peak-following algorithm implemented in the controller. Zero-crossing methods, however, can be susceptible to noise. In addition, results from a peak-follower are available only every half-cycle and thereby dictate the sample interval for controller updates.

[0281] Finally, an AM detector may be employed. Given a sine wave y(t)=A sin ω_nt, an estimate of A may be obtained from  ;₁0.5πF{abs(y)}, where F{} is a suitable low-pass filter with unity DC gain. The AM detector is the simplest approach. Moreover, it does not presume that there are oscillations of any particular frequency, and hence is usable during startup conditions. It suffers from a disadvantage that there is a leakage of harmonics into the inner loop which will affect the spectrum of the resultant oscillations. In addition, the filter adds extra dynamics into the outer loop such that compromises need to be made between speed of response and spectral purity. In particular, an effect of the filter is to constrain the choice of the best T_i.

[0282] The Fourier series for abs(y) is known to be: $A\mathrm{abs}\left(\sin {\omega }_{n}t\right)=\frac{2A}{\pi }[1+\frac{2}{3}\cos 2{\omega }_{n}t-\frac{2}{15}\cos 4{\omega }_{n}t+\frac{2}{35}\cos 6{\omega }_{n}t+\dots ].$

[0283] As such, the output has to be scaled by π/2 to give the correct DC output A, and the (even) harmonic terms a_kcos 2kω_nt have to be filtered out. As all the filter needs to do is to pass the DC component through and reduce all other frequencies, a “brick-wall” filter with a cut-off below 2ω_nis sufficient. However, the dynamics of the filter will affect the behavior of the closed-loop. A common choice of filter is in the Butterworth form. For example, the third-order low-pass filter with a design break-point frequency ω_bis: $F\left(s\right)=\frac{1}{1+2s/{\omega }_b+2s^2/{\omega }_b^2+s^3/{\omega }_b^3}.$

[0284] At the design frequency the response is 3 dB down; at 2ω_bit is −18 dB (0.12), and at 4ω_bit is −36 dB (0.015) down. Higher-order Butterworth filters have a steeper roll-off, but most of their poles are complex and may affect negatively the control-loop's root locus.

[0285] G. Zero Offset Compensation

[0286] As noted above, zero offset may be introduced into a sensor voltage signal by drift in the pre-amplification circuitry and by the analog-to-digital converter. Slight differences in the pre-amplification gains for positive and negative voltages due to the use of differential circuitry may worsen the zero offset effect. The errors vary between transmitters, and with transmitter temperature and component wear.

[0287] Audio quality (i.e., relatively low cost) analog-to-digital converters may be employed for economic reasons. These devices are not designed with DC offset and amplitude stability as high priorities. FIGS. 19A-19D show how offset and positive and negative gains vary with chip operating temperature for one such converter (the AD1879 converter). The repeatability of the illustrated trends is poor, and even allowing for temperature compensation based on the trends, residual zero offset and positive/negative gain mismatch remain.

[0288] If phase is calculated using the time difference between zero crossing points on the two sensor voltages, DC offset may lead to phase errors. This effect is illustrated by FIGS. 20A-20C. Each graph shows the calculated phase offset as measured by the digital transmitter when the true phase offset is zero (i.e., at zero flow).

[0289] FIG. 20A shows phase calculated based on whole cycles starting with positive zero-crossings. The mean value is 0.00627 degrees.

[0290] FIG. 20B shows phase calculated starting with negative zero-crossings. The mean value is 0.0109 degrees.

[0291] FIG. 20C shows phase calculated every half-cycle. FIG. 20C interleaves the data from FIGS. 20A and 20B. The average phase (−0.00234) is closer to zero than in FIGS. 20A and 20B, but the standard deviation of the signal is about six times higher.

[0292] More sophisticated phase measurement techniques, such as those based on Fourier methods, are immune to DC offset. However, it is desirable to eliminate zero offset even when those techniques are used, since data is processed in whole-cycle packets delineated by zero crossing points. This allows simpler analysis of the effects of, for example, amplitude modulation on apparent phase and frequency. In addition, gain mismatch between positive and negative voltages will introduce errors into any measurement technique.

[0293] The zero-crossing technique of phase detection may be used to demonstrate the impact of zero offset and gain mismatch error, and their consequent removal. FIGS. 21A and 21B illustrate the long term drift in phase with zero flow. Each point represents an average over one minute of live data. FIG. 21A shows the average phase, and FIG. 21B shows the standard deviation in phase. Over several hours, the drift is significant. Thus, even if the meter were zeroed every day, which in many applications would be considered an excessive maintenance requirement, there would still be considerable phase drift.

[0294] 1. Compensation Technique

[0295] A technique for dealing with voltage offset and gain mismatch uses the computational capabilities of the digital transmitter and does not require a zero flow condition. The technique uses a set of calculations each cycle which, when averaged over a reasonable period (e.g., 10,000 cycles), and excluding regions of major change (e.g., set point change, onset of aeration), converge on the desired zero offset and gain mismatch compensations.

[0296] Assuming the presence of up to three higher harmonics, the desired waveform for a sensor voltage SV(t) is of the form:

[0297] SV(t)=A₁sin(ωt)+A₂sin(2ωt)+A₃sin(3ωt)+A₄sin(4ωt)

[0298] where A₁designates the amplitude of the fundamental frequency component and A₂-A₄designate the amplitudes of the three harmonic components. However, in practice, the actual waveform is adulterated with zero offset Z₀(which has a value close to zero) and mismatch between the negative and positive gains G_nand G_p. Without any loss of generality, it can be assumed that G_pequals one and that G_nis given by:

[0299] G_n=1+ε_G,

[0300] where ε_Grepresents the gain mismatch.

[0301] The technique assumes that the amplitudes A_iand the frequency ω are constant. This is justified because estimates of Z₀and ε_Gare based on averages taken over many cycles (e.g., 10,000 interleaved cycles occurring in about 1 minute of operation). When implementing the technique, the controller tests for the presence of significant changes in frequency and amplitude to ensure the validity of the analysis. The presence of the higher harmonics leads to the use of Fourier techniques for extracting phase and amplitude information for specific harmonics. This entails integrating SV(t) and multiplying by a modulating sine or cosine function.

[0302] The zero offset impacts the integral limits, as well as the functional form. Because there is a zero offset, the starting point for calculation of amplitude and phase will not be at the zero phase point of the periodic waveform SV(t). For zero offset Z₀, the corresponding phase offset is, approximately, ${\varphi }_{Z_o}=-\sin \left(\frac{Z_o}{A_1}\right).$

[0303] For small phase, ${\varphi }_{Z_o}=-\frac{Z_o}{A_1},$

[0304] with corresponding time delay $t_{Z_o}=\frac{{\varphi }_{Z_o}}{\omega }.$

[0305] The integrals are scaled so that the limiting value (i.e., as Z₀and ε_Gapproach zero) equals the amplitude of the relevant harmonic. The first two integrals of interest are: $I_{1Ps}=\frac{2\omega }{\pi }{\int }_{t_{Z_o}}^{\frac{\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)·\sin \left[\omega \left(t-t_{Z_o}\right)\right]dt,\mathrm{and}$ $I_{1Ns}=\frac{2\omega }{\pi }\left(1+{ϵ}_G\right){\int }_{\frac{\pi }{\omega }+t_{Z_o}}^{2\frac{\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)·\sin \left[\omega \left(t-t_{Z_o}\right)\right]dt.$

[0306] These integrals represent what in practice is calculated during a normal Fourier analysis of the sensor voltage data. The subscript 1 indicates the first harmonic, N and P indicate, respectively, the negative or positive half cycle, and s and c indicate, respectively, whether a sine or a cosine modulating function has been used.

[0307] Strictly speaking, the mid-zero crossing point, and hence the corresponding integral limits, should be given by π/ω−t_Z0, rather than π/ω+t_Zo. However, the use of the exact mid-point rather than the exact zero crossing point leads to an easier analysis, and better numerical behavior (due principally to errors in the location of the zero crossing point). The only error introduced by using the exact mid-point is that a small section of each of the above integrals is multiplied by the wrong gain (1 instead of 1+ε_Gand vice versa). However, these errors are of order Z₀²ε_Gand are considered negligible.

[0308] Using computer algebra and assuming small Z₀and ε_G, first order estimates for the integrals may be derived as: $I_{1{\mathrm{Ps}}_{-}\mathrm{est}}=A_1+\frac{4}{\pi }{Z}_o\left[1+\frac{2}{3}\frac{A_2}{A_1}+\frac{4}{15}\frac{A_4}{A_1}\right],\mathrm{and}$ $I_{1{\mathrm{Ns}}_{-}\mathrm{est}}=\left(1+{ϵ}_G\right)[A_1-\frac{4}{\pi }{Z}_o\left[1+\frac{2}{3}\frac{A_2}{A_1}+\frac{4}{15}\frac{A_4}{A_1}\right]].$

[0309] Useful related functions including the sum, difference, and ratio of the integrals and their estimates may be determined. The sum of the integrals may be expressed as:

[0310] Sum_1s=(I_1Ps+I_1Ns),

[0311] while the sum of the estimates equals: ${\mathrm{Sum}}_{1s_{-}\mathrm{est}}=A_1\left(2+{ϵ}_G\right)-\frac{4}{\pi }{Z}_o{ϵ}_G\left[1+\frac{2}{3}\frac{A_2}{A_1}+\frac{4}{15}\frac{A_4}{A_1}\right].$

[0312] Similarly, the difference of the integrals may be expressed as:

[0313] Diff_1s=I_1Ps−I_1Ns,

[0314] while the difference of the estimates is: ${\mathrm{Diff}}_{1s_{-}\mathrm{est}}=-A_1{ϵ}_G+\frac{4}{\pi }{Z}_o\left(2+{ϵ}_G\right)\left[1+\frac{2}{3}\frac{A_2}{A_1}+\frac{4}{15}\frac{A_4}{A_1}\right].$

[0315] Finally, the ratio of the integrals is: ${\mathrm{Ratio}}_{1s}=\frac{I_{1\mathrm{Ps}}}{I_{1\mathrm{Ns}}},$

[0316] while the ratio of the estimates is: ${\mathrm{Ratio}}_{1s}=\frac{1}{1+{ϵ}_G}\left[1+Z_o\left[\frac{8}{15}15A_1+10A_2+\frac{4A_4}{{\mathrm{\pi A}}_1^2}\right]\right].$

[0317] Corresponding cosine integrals are defined as: $I_{1\mathrm{Pc}}=\frac{2\omega }{\pi }{\int }_{t_{Z_o}}^{\frac{\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)\cos \left[\omega \left(t-t_{Z_o}\right)\right]dt,\mathrm{and}$ $I_{1\mathrm{Nc}}=\frac{2\omega }{\pi }\left(1+{ϵ}_G\right){\int }_{\frac{\pi }{\omega }+t_{Z_o}}^{\frac{2\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)\cos \left[\omega \left(t-t_{Z_o}\right)\right]dt,$

[0318] with estimates: $I_{1{\mathrm{Pc}}_{-}\mathrm{est}}=-Z_o+\frac{40A_2+16A_4}{15\pi },\mathrm{and}$ $I_{1{\mathrm{Nc}}_{-}\mathrm{est}}=\left(1+{ϵ}_G\right)\left[Z_o+\frac{40A_2+16A_4}{15\pi }\right],$

[0319] and sums: ${\mathrm{Sum}}_{1c}=I_{1\mathrm{Pc}}+I_{1\mathrm{Nc}},\mathrm{and}$ ${\mathrm{Sum}}_{1c_{-}\mathrm{est}}={ϵ}_G\left[Z_o+\frac{40A_2+16A_4}{15\pi }\right].$

[0320] Second harmonic integrals are: $I_{2\mathrm{Ps}}=\frac{2\omega }{\pi }{\int }_{t_{Z_o}}^{\frac{\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)\sin \left[2\omega \left(t-t_{Z_o}\right)\right]dt,\mathrm{and}$ $I_{2\mathrm{Ns}}=\frac{2\omega }{\pi }\left(1+{ϵ}_G\right){\int }_{\frac{\pi }{\omega }+t_{Z_o}}^{\frac{2\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)\sin \left[2\omega \left(t-t_{Z_o}\right)\right]dt,$

[0321] with estimates: $I_{2{\mathrm{Ps}}_{-}\mathrm{est}}=A_2+\frac{8}{15\pi }{Z}_o\left[-5+9\frac{A_3}{A_1}\right],\mathrm{and}$ $I_{2{\mathrm{Ps}}_{-}\mathrm{est}}=\left(1+{ϵ}_G\right)\left[A_2-\frac{8}{15\pi }{Z}_o\left[-5+9\frac{A_3}{A_1}\right]\right],$

[0322] and sums: ${\mathrm{Sum}}_{2s}=I_{2\mathrm{Ps}}+I_{2\mathrm{Ns}},\mathrm{and}$ ${\mathrm{Sum}}_{2{\mathrm{Ps}}_{-}\mathrm{est}}=A_2\left(2+{ϵ}_G\right)-\frac{8}{15\pi }{ϵ}_G{Z}_o\left[-5+9\frac{A_3}{A_1}\right].$

[0323] The integrals can be calculated numerically every cycle. As discussed below, the equations estimating the values of the integrals in terms of various amplitudes and the zero offset and gain values arc rearranged to give estimates of the zero offset and gain terms based on the calculated integrals.

[0324] 2. Example

[0325] The accuracy of the estimation equations may be illustrated with an example. For each basic integral, tlree values are provided: the “true” value of the integral (calculated within Mathcad using Romberg integration), the value using the estimation equation, and the value calculated by the digital transmitter operating in simulation mode, using Simpson's method with end correction.

[0326] Thus, for example, the value for I_1Pscalculated according to: $I_{1\mathrm{Ps}}=\frac{2\omega }{\pi }{\int }_{t_{Z_o}}^{\frac{\pi }{\omega }+t_{Z_o}}\left(SV\left(t\right)+Z_o\right)\sin \left[\omega \left(t-t_{Z_o}\right)\right]dt$

[0327] is 0.101353, while the estimated value (I_1Psest) calculated as: $I_{1{\mathrm{Ps}}_{-}\mathrm{est}}=A_1+\frac{4}{\pi }{Z}_o\left[1+\frac{2}{3}\frac{A_2}{A_1}+\frac{4}{15}\frac{A_4}{A_1}\right]$

[0328] is 0.101358. The value calculated using the digital transmitter in simulation mode is 0.101340. These calculations use the parameter values illustrated in Table C.

[0329] The exact, estimate and simulation results from using these parameter values are illustrated in Table D.

[0330] Thus, at least for the particular values selected, the estimates given by the first order equations are extremely accurate. As Z₀and ε_Gapproach zero, the errors in both the estimate and the simulation approach zero.

[0331] 3. Implementation

[0332] The first order estimates for the integrals define a series of non-linear equations in terms of the amplitudes of the harmonics, the zero offset, and the gain mismatch. As the equations are non-linear, an exact solution is not readily available. However, an approximation followed by corrective iterations provides reasonable convergence with limited computational overhead.

[0333] Conduit-specific ratios may be assumed for A₁-A₄. As such, no attempt is made to calculate all of the amplitudes A₁-A₄. Instead, only A₁and A₂are estimated using the integral equations defined above. Based on experience of the relative amplitudes, A₃may be approximated as A₂/2, and A₄may be approximated as A₂/10.

[0334] The zero offset compensation technique may be implemented according to the procedure 2200 illustrated in FIG. 22. During each cycle, the controller calculates the integrals I_1Ps, I_1Ns, I_1Pc, I_1Nc, I_2Ps, I_2Nsand related functions sum_1s, ratio_1s, sum_1cand sum_2s(step 2205). This requires minimal additional calculation beyond the conventional Fourier calculations used to determine frequency, amplitude and phase.

[0335] Every 10,000 cycles, the controller checks on the slope of the sensor voltage amplitude A₁, using a conventional rate-of-change estimation technique (step 2210). If the amplitude is constant (step 2215), then the controller proceeds with calculations for zero offset and gain mismatch. This check may be extended to test for frequency stability.

[0336] To perform the calculations, the controller generates average values for the functions (e.g., sum_1s) over the last 10,000 cycles. The controller then makes a first estimation of zero offset and gain mismatch (step 2225):

[0337] Z₀=−Sum_1c/2, and ε_G=1/Ratio_1s−1

[0338] Using these values, the controller calculates an inverse gain factor (k) and amplitude factor (amp_factor) (step 2230):

[0339] k=1.0/(1.0+0.5*ε_G), and amp_factor=1+50/75*Sum_2s/Sum_1s

[0340] The controller uses the inverse gain factor and amplitude factor to make a first estimation of the amplitudes (step 2235):

[0341] A₁=k*[Sum_1s/2+2/π*Z₀*ε_G*amp_factor], and A₂=k*[Sum_2s/2−4/(3*π)*Z₀*ε_G

[0342] The controller then improves the estimate by the following calculations, iterating as required (step 2240):

[0343] x₁=Z₀,

[0344] x₂=ε_G,

[0345] ε_G=[1+8/π*x₁/A₁*amp_factor]/Ratio_1s−1.0,

[0346] Z₀=−Sum_1c/2+x₂*(x₁+2.773/π*A₂)/2,

[0347] A₁=k*[Sum_1s/2+2/π*x₁*x₂*amp_factor],

[0348] A₂=k*[Sum_2s/2−4/(15*π)*x₁*x₂*(5-4.5*A₂)].

[0349] The controller uses standard techniques to test for convergence of the values of Z₀and ε_G. In practice the corrections are small after the first iteration, and experience suggests that three iterations are adequate.

[0350] Finally, the controller adjusts the raw data to eliminate Z₀and ε_G(step 2245). The controller then repeats the procedure. Once zero offset and gain mismatch have been eliminated from the raw data, the functions (i.e., sum_1s) used in generating subsequent values for Z₀and ε_Gare based on corrected data. Accordingly, these subsequent values for Z₀and ε_Greflect residual zero offset and gain mismatch, and are summed with previously generated values to produce the actual zero offset and gain mismatch. In one approach to adjusting the raw data, the controller generates adjustment parameters (e.g., S1_off and S2_off) that are used in converting the analog signals from the sensors to digital data.

[0351] FIGS. 23A-23C, 24A and 24B show results obtained using the procedure 2200. The short-term behavior is illustrated in FIGS. 23A-23C. This shows consecutive phase estimates obtained five minutes after startup to allow time for the procedure to begin affecting the output. Phase is shown based on positive zero-crossings, negative zero-crossings, and both.

[0352] The difference between the positive and negative mean values has been reduced by a factor of 20, with a corresponding reduction in mean zero offset in the interleaved data set. The corresponding standard deviation has been reduced by a factor of approximately 6.

[0353] Longer term behavior is shown in FIGS. 24A and 24B. The initial large zero offset is rapidly corrected, and then the phase offset is kept close to zero over many hours. The average phase offset, excluding the first few values, is 6.14e⁻⁶, which strongly suggests that the procedure is successful in compensating for changes in voltage offset and gain imbalance.

[0354] Typical values for Z₀and ε_Gfor the digital Coriolis meter are Z₀=−7.923e⁻⁴and ε_G=−1.754e⁻⁵for signal SV₁, and Z₀=−8.038e⁻⁴and ε_G=+6.93e⁻⁴for signal SV₂.

[0355] H. Dynamic Analysis

[0356] In general, conventional measurement calculations for Coriolis meters assume that the frequency and amplitude of oscillation on each side of the conduit are constant, and that the frequency on each side of the conduit is identical and equal to the so-called resonant frequency. Phases generally are not measured separately for each side of the conduit, and the phase difference between the two sides is assumed to be constant for the duration of the measurement process. Precise measurements of frequency, phase and amplitude every half-cycle using the digital meter demonstrate that these assumptions are only valid when parameter values are averaged over a time period on the order of seconds. Viewed at 100 Hz or higher frequencies, these parameters exhibit considerable variation. For example, during normal operation, the frequency and amplitude values of SV₁may exhibit strong negative correlation with the corresponding SV₂values. Accordingly, conventional measurement algorithms are subject to noise attributable to these dynamic variations. The noise becomes more significant as the measurement calculation rate increases. Other noise terms may be introduced by physical factors, such as flowtube dynamics, dynamic non-linearities (e.g. flowtube stiffness varying with amplitude), or the dynamic consequences of the sensor voltages providing velocity data rather than absolute position data.

[0357] The described techniques exploit the high precision of the digital meter to monitor and compensate for dynamic conduit behavior to reduce noise so as to provide more precise measurements of process variables such as mass flow and density. This is achieved by monitoring and compensating for such effects as the rates of change of frequency, phase and amplitude, flowtube dynamics, and dynamic physical non-idealities. A phase difference calculation which does not assume the same frequency on each side has already been described above. Other compensation techniques are described below.

[0358] Monitoring and compensation for dynamic effects may take place at the individual sensor level to provide corrected estimates of phase, frequency, amplitude or other parameters. Further compensation may also take place at the conduit level, where data from both sensors are combined, for example in the calculation of phase difference and average frequency. These two levels may be used together to provide comprehensive compensation.

[0359] Thus, instantaneous mass flow and density measurements by the flowmeter may be improved by modelling and accounting for dynamic effects of flowmeter operation. In general, 80% or more of phase noise in a Coriolis flowmeter may be attributed to flowtube dynamics (sometimes referred to as “ringing”), rather than to process conditions being measured. The application of a dynamic model can reduce phase noise by a factor of 4 to 10, leading to significantly improved flow measurement performance. A single model is effective for all flow rates and amplitudes of oscillation. Generally, computational requirements are negligible.

[0360] The dynamic analysis may be performed on each of the sensor signals in isolation from the other. This avoids, or at least delays, modelling the dynamic interaction between the two sides of the conduit, which is likely to be far more complex than the dynamics at each sensor. Also, analyzing the individual sensor signals is more likely to be successful in circumstances such as batch startup and aeration where the two sides of the conduit are subject to different forces from the process fluid.

[0361] In general, the dynamic analysis considers the impact of time-varying amplitude, frequency and phase on the calculated values for these parameters. While the frequency and amplitude are easily defined for the individual sensor voltages, phase is conventionally defined in terms of the difference between the sensor voltages. However, when a Fourier analysis is used, phase for the individual sensor may be defined in terms of the difference between the midpoint of the cycle and the average 180° phase point.

[0362] Three types of dynamic effects are measurement error and the so-called “feedback” and “velocity” effects. Measurement error results because the algorithms for calculating amplitude and phase assume that frequency, amplitude, and phase are constant over the time interval of interest. Performance of the measurement algorithms may be improved by correcting for variations in these parameters.

[0363] The feedback effect results from supplying energy to the conduit to make up for energy loss from the conduit so as to maintain a constant amplitude of oscillation. The need to add energy to the conduit is only recognized after the amplitude of oscillation begins to deviate from a desired setpoint. As a result, the damping term in the equation of motion for the oscillating conduit is not zero, and, instead, constantly dithers around zero. Although the natural frequency of the conduit does not change, it is obscured by shifts in the zero-crossings (i.e., phase variations) associated with these small changes in amplitude.

[0364] The velocity effect results because the sensor voltages observe conduit velocity, but are analyzed as being representative of conduit position. A consequence of this is that the rate of change of amplitude has an impact on the apparent frequency and phase, even if the true values of these parameters are constant.

[0365] 1. Sensor-Level Compensation for Amplitude Modulation

[0366] One approach to correcting for dynamic effects monitors the amplitudes of the sensor signals and makes adjustments based on variations in the amplitudes. For purposes of analyzing dynamic effects, it is assumed that estimates of phase, frequency and amplitude may be determined for each sensor voltage during each cycle. As shown in FIG. 25, calculations are based on complete but overlapping cycles. Each cycle starts at a zero crossing point, halfway through the previous cycle. Positive cycles begin with positive voltages immediately after the initial zero-crossing, while negative cycles begin with negative voltages. Thus cycle n is positive, while cycles n−1 and n+1 are negative. It is assumed that zero offset correction has been performed so that zero offset is negligible. It also is assumed that higher harmonics may be present.

[0367] Linear variation in amplitude, frequency, and phase are assumed. Under this assumption, the average value of each parameter during a cycle equals the instantaneous value of the parameter at the mid-point of the cycle. Since the cycles overlap by 180°, the average value for a cycle equals the starting value for the next cycle.

[0368] For example, cycle n is from time 0 to 2π/ω. The average values of amplitude, frequency and phase equal the instantaneous values at the mid-point, π/ω, which is also the starting point for cycle n+1, which is from time π/ω to 3π/ω. Of course, these timings are approximate, since ω also varies with time.

[0369] a. Dynamic Effect Compensation Procedure

[0370] The controller accounts for dynamic effects according to the procedure 2600 illustrated in FIG. 26. First, the controller produces a frequency estimate (step 2605) by using the zero crossings to measure the time between the start and end of the cycle, as described above. Assuming that frequency varies linearly, this estimate equals the time-averaged frequency over the period.

[0371] The controller then uses the estimated frequency to generate a first estimate of amplitude and phase using the Fourier method described above (step 2610). As noted above, this method eliminates the effects of higher harmonics.

[0372] Phase is interpreted in the context of a single waveform as being the difference between the start of the cycle (i.e., the zero-crossing point) and the point of zero phase for the component of SV(t) of frequency ω, expressed as a phase offset. Since the phase offset is an average over the entire waveform, it may be used as the phase offset from the midpoint of the cycle. Ideally, with no zero offset and constant amplitude of oscillation, the phase offset should be zero every cycle. In practice, however, it shows a high level of variation and provides an excellent basis for correcting mass flow to account for dynamic changes in amplitude.

[0373] The controller then calculates a phase difference (step 2615). Though a number of definitions of phase difference are possible, the analysis assumes that the average phase and frequency of each sensor signal is representative of the entire waveform. Since these frequencies are different for SV₁and SV₂, the corresponding phases are scaled to the average frequency. In addition, the phases are shifted to the same starting point (i.e., the midpoint of the cycle on SV₁). After scaling, they are subtracted to provide the phase difference.

[0374] The controller next determines the rate of change of the amplitude for the cycle n (step 2620): $\begin{array}{c}{\mathrm{roc}}_{-}{\mathrm{amp}}_n\approx \frac{\mathrm{amp}\left(\mathrm{end}\mathrm{of}\mathrm{cycle}\right)-\mathrm{amp}\left(\mathrm{start}\mathrm{of}\mathrm{cycle}\right)}{\mathrm{period}\mathrm{of}\mathrm{cycle}}\\ =\left({\mathrm{amp}}_{n+1}-{\mathrm{amp}}_{n-1}\right){\mathrm{freq}}_n.\end{array}$

[0375] This calculation assumes that the amplitude from cycle n+1 is available when calculating the rate of change of cycle n. This is possible if the corrections are made one cycle after the raw amplitude calculations have been made. The advantage of having an accurate estimate of the rate of change, and hence good measurement correction, outweighs the delay in the provision of the corrected measurements, which, in one implementation, is on the order of 5 milliseconds. The most recently generated information is always used for control of the conduit (i.e., for generation of the drive signal).

[0376] If desired, a revised estimate of the rate of change can be calculated after amplitude correction has been applied (as described below). This results in iteration to convergence for the best values of amplitude and rate of change.

[0377] b. Frequency Compensation for Feedback and Velocity Effects

[0378] As noted above, the dynamic aspects of the feedback loop introduce time varying shifts in the phase due to the small deviations in amplitude about the set-point. This results in the measured frequency, which is based on zero-crossings, differing from the natural frequency of the conduit. If velocity sensors are used, an additional shift in phase occurs. This additional shift is also associated with changes in the positional amplitude of the conduit. A dynamic analysis can monitor and compensate for these effects. Accordingly, the controller uses the calculated rate of amplitude change to correct the frequency estimate (step 2625).

[0379] The position of an oscillating conduit in a feedback loop that is employed to maintain the amplitude of oscillation of the conduit constant may be expressed as:

[0380] X=A(t) sin(ω₀t−θ(t)),

[0381] where θ(t) is the phase delay caused by the feedback effect. The mechanical Q of the oscillating conduit is typically on the order of 1000, which implies small deviations in amplitude and phase. Under these conditions, θ(t) is given by:

[0382] θ(t)≈−{dot over (A)}(t)/2ω₀A(t).

[0383] Since each sensor measures velocity: $\begin{array}{c}\mathrm{SV}\left(t\right)=\stackrel{.}{X}\left(t\right)=\stackrel{.}{A}\left(t\right)\sin \left[{\omega }_{0}t-\theta \left(t\right)\right]+\left[{\omega }_0-\stackrel{.}{\theta }\left(t\right)\right]A\left(t\right)\cos \left[{\omega }_{0}t-\theta \left(t\right)\right]\\ ={\omega }_0{A\left(t\right)\left[{\left(1-\frac{\stackrel{.}{\theta }}{{\omega }_0}\right)}^2+{\left(\frac{\stackrel{.}{A}\left(t\right)}{{\omega }_{0}A\left(t\right)}\right)}^2\right]}^{1/2}\cos \left({\omega }_{0}t-\theta \left(t\right)-\gamma \left(t\right)\right),\end{array}$

[0384] where γ(t) is the phase delay caused by the velocity effect: $\gamma \left(t\right)={\tan }^{-1}\left(\frac{\stackrel{.}{A}\left(t\right)}{{\omega }_{0}A\left(t\right)\left(1-\frac{\stackrel{.}{\theta }}{{\omega }_0}\right)}\right).$

[0385] Since the mechanical Q of the conduit is typically on the order of 1000, and, hence, variations in amplitude and phase are small, it is reasonable to assume: $\frac{\stackrel{.}{\theta }}{{\omega }_0}⪡1\mathrm{and}\frac{\stackrel{.}{A}\left(t\right)}{{\omega }_{0}A\left(t\right)}⪡1.$

[0386] This means that the expression for SV(t) may be simplified to:

[0387] SV(t)≈ω₀A(t) cos(ω₀t−θ(t)−γ(t)),

[0388] and for the same reasons, the expression for the velocity offset phase delay may be simplified to: $\gamma \left(t\right)\approx \frac{\stackrel{.}{A}\left(t\right)}{{\omega }_{0}A\left(t\right)}.$

[0389] Summing the feedback and velocity effect phase delays gives the total phase delay: $\varphi \left(t\right)=\theta \left(t\right)+\gamma \left(t\right)\approx -\frac{\stackrel{.}{A}\left(t\right)}{2{\omega }_{0}A\left(t\right)}+\frac{\stackrel{.}{A}\left(t\right)}{{\omega }_{0}A\left(t\right)}=\frac{\stackrel{.}{A}\left(t\right)}{2{\omega }_{0}A\left(t\right)},$

[0390] and the following expression for SV(t):

[0391] SV(t)≈ω₀A(t) cos [ω₀t−φ(t)].

[0392] From this, the actual frequency of oscillation may be distinguished from the natural frequency of oscillation. Though the former is observed, the latter is useful for density calculations. Over any reasonable length of time, and assuming adequate amplitude control, the averages of these two frequencies are the same (because the average rate of change of amplitude must be zero). However, for improved instantaneous density measurement, it is desirable to compensate the actual frequency of oscillation for dynamic effects to obtain the natural frequency. This is particularly useful in dealing with aerated fluids for which the instantaneous density can vary rapidly with time.

[0393] The apparent frequency observed for cycle n is delineated by zero crossings occurring at the midpoints of cycles n−1 and n+1. The phase delay due to velocity change will have an impact on the apparent start and end of the cycle: $\begin{array}{c}{\mathrm{obs}}_{—}{\mathrm{freq}}_n={\mathrm{obs}}_{—}{\mathrm{freq}}_{n-1}+\frac{{\mathrm{true}}_{—}{\mathrm{freq}}_n}{2\pi }\left({\varphi }_{n+1}-{\varphi }_{n-1}\right)\\ ={\mathrm{obs}}_{—}{\mathrm{freq}}_{n-1}+\frac{{\mathrm{true}}_{—}{\mathrm{freq}}_{n-1}}{2\pi }\left(\frac{{\stackrel{.}{A}}_{n+1}}{4\pi {\mathrm{true}}_{—}{\mathrm{freq}}_n{A}_{n+1}}-\frac{{\stackrel{.}{A}}_{n-1}}{4\pi {\mathrm{true}}_{—}{\mathrm{freq}}_n{A}_{n-1}}\right)\\ ={\mathrm{obs}}_{—}{\mathrm{freq}}_{n-1}+\frac{1}{8{\pi }^2}\left(\frac{{\stackrel{.}{A}}_{n+1}}{A_{n+1}}-\frac{{\stackrel{.}{A}}_{n-1}}{A_{n-1}}\right).\end{array}$

[0394] Based on this analysis, a correction can be applied using an integrated error term: ${\mathrm{error}}_{—}{\mathrm{sum}}_n={\mathrm{error}}_{—}{\mathrm{sum}}_{n-1}-\frac{1}{8{\pi }^2}\left(\frac{{\stackrel{.}{A}}_{n+1}}{A_{n+1}}-\frac{{\stackrel{.}{A}}_{n-1}}{A_{n-1}}\right),\mathrm{and}$ ${\mathrm{est}}_{—}{\mathrm{freq}}_n={\mathrm{obs}}_{—}{\mathrm{freq}}_n-{\mathrm{error}}_{—}{\mathrm{sum}}_n,$

[0395] where the value of error_sum at startup (i.e., the value at cycle zero) is: ${\mathrm{error}}_{—}{\mathrm{sum}}_0=-\frac{1}{8{\pi }^2}\left(\frac{{\stackrel{.}{A}}_0}{A_0}+\frac{{\stackrel{.}{A}}_1}{A_1}\right).$

[0396] Though these equations include a constant term having a value of ⅛π², actual data has indicated that a constant term of ⅛π is more appropriate. This discrepancy may be due to unmodelled dynamics that may be resolved through further analysis.

[0397] The calculations discussed above assume that the true amplitude of oscillation, A, is available. However, in practice, only the sensor voltage SV is observed. This sensor voltage may be expressed as:

[0398] SV(t)≈ω₀A(t) cos(ω₀t−φ(t))

[0399] The amplitude, amp_SV(t), of this expression is:

[0400] amp_SV(t)≈ω₀A(t).

[0401] The rate of change of this amplitude is:

[0402] roc_amp_SV(t)≈ω₀{dot over (A)}(t)

[0403] so that the following estimation can be used: $\frac{\stackrel{.}{A}\left(t\right)}{A\left(t\right)}\approx \frac{{\mathrm{roc}}_{—}{\mathrm{amp}}_{—}\mathrm{SV}\left(t\right)}{{\mathrm{amp}}_{—}\mathrm{SV}\left(t\right)}.$

[0404] c. Application of Feedback and Velocity Effect Frequency Compensation

[0405] FIGS. 27A-32B illustrate how application of the procedure 2600 improves the estimate of the natural frequency, and hence the process density, for real data from a meter having a one inch diameter conduit. Each of the figures shows 10,000 samples, which are collected in just over 1 minute.

[0406] FIGS. 27A and 27B show amplitude and frequency data from SV₁, taken when random changes to the amplitude set-point have been applied. Since the conduit is full of water and there is no flow, the natural frequency is constant. However, the observed frequency varies considerably in response to changes in amplitude. The mean frequency value is 81.41 Hz, with a standard deviation of 0.057 Hz.

[0407] FIGS. 28A and 28B show, respectively, the variation in frequency from the mean value, and the correction term generated using the procedure 2600. The gross deviations are extremely well matched. However, there is additional variance in frequency which is not attributable to amplitude variation. Another important feature illustrated by FIG. 28B is that the average is close to zero as a result of the proper initialization of the error term, as described above.

[0408] FIGS. 29A and 92B compare the raw frequency data (FIG. 29A) with the results of applying the correction function (FIG. 29B). There has been a negligible shift in the mean frequency, while the standard deviation has been reduced by a factor of 4.4. From FIG. 29B, it is apparent that there is residual structure in the corrected frequency data. It is expected that further analysis, based on the change in phase across a cycle and its impact on the observed frequency, will yield further noise reductions.

[0409] FIGS. 30A and 30B show the corresponding effect on the average frequency, which is the mean of the instantaneous sensor voltage frequencies. Since the mean frequency is used to calculate the density of the process fluid, the noise reduction (here by a factor of 5.2) will be propagated into the calculation of density.

[0410] FIGS. 31A and 31B illustrate the raw and corrected average frequency for a 2″ diameter conduit subject to a random amplitude set-point. The 2″ flowtube exhibits less frequency variation that the 1″, for both raw and corrected data. The noise reduction factor is 4.0.

[0411] FIGS. 32A and 32B show more typical results with real flow data for the one inch flowtube. The random setpoint algorithm has been replaced by the normal constant setpoint. As a result, there is less amplitude variation than in the previous examples, which results in a smaller noise reduction factor of 1.5.

[0412] d. Compensation of Phase Measurement for Amplitude Modulation

[0413] Referring again to FIG. 26, the controller next compensates the phase measurement to account for amplitude modulation assuming the phase calculation provided above (step 2630). The Fourier calculations of phase described above assume that the amplitude of oscillation is constant throughout the cycle of data on which the calculations take place. This section describes a correction which assumes a linear variation in amplitude over the cycle of data.

[0414] Ignoring higher harmonics, and assuming that any zero offset has been eliminated, the expression for the sensor voltage is given by:

[0415] SV(t)≈A₁(1+λ_At) sin(ωt)

[0416] where λ_Ais a constant corresponding to the relative change in amplitude with time. As discussed above, the integrals I₁and I₂may be expressed as: $I_1=\frac{2\omega }{\pi }{\int }_0^{\frac{2\pi }{\omega }}\mathrm{SV}\left(t\right)\sin \left(\omega t\right)dt,\mathrm{and}$ $I_2=\frac{2\omega }{\pi }{\int }_0^{\frac{2\pi }{\omega }}\mathrm{SV}\left(t\right)\cos \left(\omega t\right)dt.$

[0417] Evaluating these integrals results in: $I_1=A_1\left(1+\frac{\pi }{\omega }{\lambda }_A\right),\mathrm{and}$ $I_2=-A_1\frac{1}{2\omega }{\lambda }_{A.}$

[0418] Substituting these expressions into the calculation for amplitude and expanding as a series in λ_Aresults in: $\mathrm{Amp}=A_1(1+\frac{\pi }{\omega }{\lambda }_A+\frac{1}{8{\omega }^2}{\lambda }_A^2+\dots ).$

[0419] Assuming λ_Ais small, and ignoring all tennis after the first order term, this may be simplified to: $\mathrm{Amp}=A_1\left(1+\frac{\pi }{\omega }{\lambda }_A\right).$

[0420] This equals the amplitude of SV(t) at the midpoint of the cycle (t=π/ω). Accordingly, the amplitude calculation provides the required result without correction.

[0421] For the phase calculation, it is assumed that the true phase difference and frequency are constant, and that there is no voltage offset, which means that the phase value should be zero. However, as a result of amplitude modulation, the correction to be applied to the raw phase data to compensate for amplitude modulation is: $\mathrm{Phase}={\tan }^{-1}\left(\frac{{\lambda }_A}{2\left({\mathrm{\pi \lambda }}_A+\omega \right)}\right).$

[0422] Assuming that the expression in brackets is small, the inverse tangent function can be ignored.

[0423] A more elaborate analysis considers the effects of higher harmonics. Assuming that the sensor voltage may be expressed as:

[0424] SV(t)=(1+λ_At)[A₁sin(ωt)+A₂sin(2ωt)+A₃sin(3ωt)+A₄sin(4ωt)]

[0425] such that all harmonic amplitudes increase at the same relative rate over the cycle, then the resulting integrals may be expressed as: $I_1=A_1\left(1+\frac{\pi }{\omega }{\lambda }_A\right),\mathrm{and}$ $I_2=\frac{-1}{60\omega }{\lambda }_A\left(30A_1+80A_2+45A_3+32A_4\right)$

[0426] for positive cycles, and $I_2=\frac{-1}{60\omega }{\lambda }_A\left(30A_1-80A_2+45A_3-32A_4\right)$