Method and device for controlling a piezo-actuator

[0001] The present invention is directed to a method, a control unit, and a fuel injection system, respectively, where a piezoelectric actuator is electrically recharged by the application of an electric current in order to change its length. Such a method, in which the derivative with respect to time of the voltage applied to the piezoelectric actuator is changed within a charging or discharging operation, is known from German Patent 199 21 456.

[0002] The method and the devices according to the present invention having the characterizing features of the independent claims have the advantage over the related art that they lower the noise emissions of the injection system exactly in those operating situations where they are significantly influenced by the triggering of the piezoelectric actuators utilized. In addition, a fundamental advantage lies in the fact that, in common rail injection systems in particular, the system behavior, i.e., the accuracy of triggering, as well as the metering of the injected quantities remain unaffected, especially at high rail pressures, i.e., that even at high rotational speeds or high loads on the internal combustion engine the required timing tolerances with respect to triggering, as well as the accuracy of the metered quantity, are easily complied with.

[0003] Advantageous refinements of and improvements to the methods and devices listed in the independent claims are rendered possible by the measures described in the dependent claims.

[0004] Exemplary embodiments of the present invention are illustrated in the drawing and explained in greater detail in the following description. The figures show:

[0005] FIG. 1 two voltage-time diagrams

[0006] FIG. 2 a flow diagram

[0007] FIG. 3 a block diagram

[0008] FIG. 4 an additional block diagram.

[0009] FIG. 1a shows a voltage-time diagram. It shows the variation of voltage over time across a piezoelectric actuator which controls the injection of fuel into the combustion chamber of an internal combustion engine via a valve. Two standard triggering characteristics are illustrated; in the first triggering, voltage U is linearly increased within charging time 1 from zero to a value U1 which is maintained for a certain time (e.g., U1200 V). During subsequent discharging time 2, the voltage applied to the piezoelectric actuator is again linearly reduced to zero. The second triggering has an intermediate level U2 (e.g., U2100 V), to which the voltage is initially increased within charging time 3. After reaching this voltage level, the voltage is increased by the difference value U3 (e.g., U3100 V) within additional charging time 4, to be only subsequently reduced in two steps to the value zero within discharging times 5 and 6. FIG. 1b shows similar voltage characteristics having identical voltage levels U1 and U2, respectively. However, the charging times and discharging times 7, 8, 9, 11, 12, and 13 are longer than the charging times and discharging times 1 through 6 in FIG. 1a. The absolute value of the derivatives with respect to time of the voltage characteristics in the charging times and discharging times is therefore less than in FIG. 1a. Any triggering characteristics that may be represented by broken lines are basically conceivable and the above description is appropriately applicable.

[0010] As a general rule, in injection systems having piezoelectric actuators, a control valve which controls the movement of the nozzle needle is not triggered directly, but via a hydraulic coupler, as described in German Patent Application 197 32 802, for example. This coupler has essentially two functions: First, it reinforces the lift of the piezoelectric actuator and second, it decouples the control valve from the static thermal expansion of the actuator. The triggering voltage necessary for accurate positioning of the control valve and thus for implementing a desired injection is, as a rule, heavily dependent on the fuel pressure and, in a common rail system, on the rail pressure of the fuel. This is explained by the fact that the control valve works against or with the rail pressure, depending on the switching direction of the valve. As a rule, the derivative with respect to time of the triggering voltage is to be selected in such a way that the charging time and discharging time correspond exactly to the time constant of the mechanical system. The vibration induced in the system is minimized in this case. For different reasons, it is indeed desirable to keep the charging time and discharging time as short as possible, in particular to implement triggering periods as short as possible, in order to supply the smallest injected quantities, which is especially important at high rail pressures. On the other hand, the noise emission increases notably with the gradient, i.e., the derivative with respect to time of the voltage since, due to the high speed of the actuator movement, the control valve is also moved with similar speed. This effect is interfering in certain operating situations of the engine. In this connection, the expression operating situation is not to be understood as a certain period of time within a triggering of the piezoelectric actuator, but rather the operating condition, generally present through several injection cycles, such as idling, for example, which is characterized by small load and low rotational speed. Triggering according to FIG. 1a is to be used in normal driving operation under load, while in the operating situation idling, a triggering according to FIG. 1b having a flatter triggering gradient is preferred to achieve a reduction in noise emission, particularly here where the noise caused by triggering of the injection system is noticeable compared to other vehicle noises.

[0011] FIG. 2 illustrates the procedure of triggering of a piezoelectric actuator which, in a common rail injector for example, controls the injection of diesel fuel into the combustion chamber of the diesel engine. After switching on 10 of the engine, i.e., the injection system, it is first verified in query 20 whether a charging/discharging operation is requested. If this is the case, the operating condition of the engine is determined (process step 30). The operating condition of the engine is characterized by the rotational speed and/or the load on the engine and/or by the fuel pressure in the injection system. Further characterizing variables may be the temperature of the piezoelectric actuator, the temperature of the fuel, or other characteristic data. In subsequent process step 40, the setpoint of the derivative with respect to time of the voltage which is to be applied to the piezoelectric actuator is determined as a function of the operating condition of the engine. The gradient setpoint is set here in such a way that the noise development due to the movement of mechanical components is minimized while the functionality of the injection system is preserved. When certain threshold values of the rotational speed, the load torque, and/or the rail pressure are reached here (e.g., rotational speed <2000 rpm, the load is less than 10% of the maximum load and the rail pressure is below 500 bar), then a smooth transition of the gradient setpoint, in comparison to normal operation, is implemented, so that below the threshold values mentioned, the derivative with respect to time of the voltage to be applied changes over continuously to smaller values. The charging time or the discharging time varies typically (e.g., at 50% of the maximum load) between 80 s and 100 s, while it assumes values between 100 s and 150 s below the threshold values. In subsequent query 50, it is checked whether it is the first request of the injection system after switching on. If yes, a driver signal is calculated for a driver which triggers charging/discharging means to be applied to the piezoelectric actuator. The driver signal is calculated here in such a way that a sufficient electric current is fed to the piezoelectric actuator in order to achieve the determined setpoint of the derivative with respect to time or the charging/discharging time of the voltage to be applied. In additional step 80, the driver that triggers the charging/discharging means is triggered until the final value of the electric voltage across the piezoelectric actuator is reached. In an additional step 90, the actual value of time is determined, which was necessary to charge or discharge the piezoelectric actuator to the voltage to be achieved. The program subsequently returns to query 20.

[0012] If in query 50 the result is No, then the system deviation, i.e., the deviation of the last actual value of the time needed for the recharging, from the calculated setpoint, is determined and is taken into account in subsequent process step 70 for calculating the driver signal for the next recharging of the piezoelectric actuator.

[0013] The change in triggering only in certain operating points, such as idling (characterized above by the threshold values mentioned), is entirely sufficient, since, due to triggering, only in these points does the noise, imitated by the injector, significantly influence the overall noise of the drive unit. In partial load or full load operation, however, the overall noise is by far dominated by the combustion noise. The present invention is based on the idea that in order to implement a more constant charging/discharging time in the range of the system time, the triggering gradients, i.e., the charging/discharging times are not changed, as previously, as a function of the voltage, but are switched over to a flatter gradient in certain operating situations, in particular during idling. In doing so, the noise emission may be significantly reduced. The rail pressure is also relatively low during idling, so that even during longer charging/discharging times, the smallest injected quantities may be implemented and the narrow tolerances to be adhered to with regard to the injected quantities may be ensured.

[0014] Alternatively to a smooth transition of the gradient or the time setpoint between normal operation and idling, a hard switch-over to smaller gradients may also be provided when one or several of the threshold values fall below a certain value.

[0015] FIG. 3 shows a control unit 200 which is connected to a driver 120 and charging/discharging means 110. The control unit has a monitoring unit 150 which is supplied with operating condition variables 210. These operating condition variables are the rotational speed, the load torque, the rail pressure, and/or the temperature of the piezoelectric actuator, and/or the fuel temperature, and/or other parameters. Monitoring unit 150 determines the setpoints for the charging/discharging times and the charging/discharging gradients and transmits these to logic circuit 130. Logic circuit 130 is connected to an actual value detecting unit 140, which, as illustrated in FIG. 3, may be integrated into the control unit, but may also be situated separately in the immediate proximity of charging/discharging means 110. Actual value detecting unit 140 is connected to charging/discharging means 110. Logic circuit 130 may receive a request signal from higher-level engine control units (not shown) via line 220. Logic circuit 130 is connected to a driver 120 which, in turn, is interconnected with charging/discharging means 110 which apply a voltage to piezoelectric actuator 100 as a function of time.

[0016] The setpoint for the charging/discharging time is determined in monitoring unit 150, taking into consideration the variables rotational speed, load, and rail pressure, and the monitoring unit transmits the determined value to logic circuit 130. Upon request, logic circuit 130 calculates a driver signal via signal line 220 taking into consideration the actual value of the charging/discharging time or the charging/discharging gradient measured by actual value detection unit 140. Logic circuit 130 conveys the driver signal to driver 120 which then triggers charging/discharging means 110 in order to implement the voltage gradients to be achieved across piezoelectric actuator 100.

[0017] To regulate the control gradients during the recharging phases, variables other than rotational speed load and/or rail pressure may be alternatively used for determining the operating condition of the engine and/or the injection system.

[0018] FIG. 4 shows a component 131 of logic circuit 130 in the form of a block diagram. The actual value detected by actual value detection unit 140 and the setpoint calculated by monitoring unit 150 are fed to a summing node 255 via lines 250 and 260, respectively. The summing node calculates the system deviation, i.e., the difference between the setpoint and the actual value and feeds this difference to PI regulator 270, i.e., a proportional amplifier, which is connected in parallel to an integrator. The output of PI regulator 270 is connected to a second summing node 275 which adds the output value of the PI regulator to the setpoint from monitoring unit 150. Prior to or following the recharging procedure to be calculated, the voltage levels are fed via lines 280 or 290 to a third summing node 285 which calculates their difference and feeds it to multiplier 295 which, in turn, calculates the charge necessary for the recharging procedure from the difference and the value of the capacitance of the piezoelectric actuator fed via line 300. Divider 305 divides the value of the electric charge, obtained from multiplier 295, by the value of the charging/discharging time obtained from summing node 275, so that the information about the current value necessary for the recharging procedure at the piezoelectric actuator may be picked off at output 310 of divider 305. Output 310 of divider 305 is connected to driver 120 and is available to it for triggering charging/discharging means 110 (see FIG. 3). Lines 280, 290, and 300 are connected either to storage elements in which the voltage and capacitance values to be retrieved are stored, or they are connected to separate circuit elements (not shown) which recalculate or define the voltage and capacitance values, as a function of the triggering demand and switching condition.

[0019] Component 131 implements the process steps illustrated in FIG. 2. The charging and discharging time are regulated by a PI regulator, the difference between the voltage levels to be bridged, and the actuator capacitance of the associated charging and discharging current being determined.