Apparatus and Method for Determining Velocity of a Platform

The present invention relates to an apparatus and method for determining the velocity of a platform and, in particular, to an apparatus and method for determining the velocity of a platform without using a ground station.

The Global Positing System (GPS) is a satellite-based navigation system that continuously transmits timing, frequency, and satellite position information to potential users. The GPS consists of a full constellation of twenty-four (24) satellites in half geosynchronous orbits. The position of the GPS satellites is controlled and monitored by the Department of Defense (DoD). GPS satellites continuously emit coded GPS signals. The GPS signal contains timing information that allows a user to determine the time elapsed for the GPS signal to transverse the distance between the GPS satellite and the user (the platform). By knowing the time the GPS signal left the GPS satellite, the time the GPS signal arrived at the user, and the speed of the GPS signal, the user can determine the distance from itself to the GPS satellite. By knowing the position of the GPS satellite (ephemeris data), and the distance from itself to the GPS satellite, the user can successfully triangulate its own position.

A user can also determine its own velocity by processing carrier frequency information. All GPS satellites broadcast on two carrier frequencies, referred to as the L1 and L2 carrier frequencies. The carrier frequencies are closely monitored by the DoD to ensure carrier frequency integrity. However, the GPS satellite frequency as measured by the user will change proportional to the relative velocity between the GPS satellite and user. This is known as the Doppler shift. In a similar manner as the above position solution, the user can determine its own velocity by knowing the GPS satellite to user relative velocity (Doppler shift), and the velocity of the GPS satellite. The user knows the velocity of the GPS satellite by the ephemeris information that the GPS satellite transmits to the user. As such, measurements of Doppler shift in the carrier frequency may be used to determine velocity.

To obtain a velocity measurement more accurate than that acquired by measuring and using carrier frequency Doppler shift alone, a ground reference station is utilized. Most high accuracy systems require the use of a ground station to process differential corrections commonly referred to as Differential GPS (DGPS). The satellite-user GPS system has systematic errors that degrade the position and velocity solution. These errors can be corrected for by establishing a fixed reference station that knows its position. The reference station also knows its own velocity, that of zero velocity. The reference station monitors the GPS satellite signals and compares computed range and range-rate solutions against measured solutions. Because the reference station knows its own position and velocity accurately, it can determine the common errors in the GPS satellite system. The reference station then transmit these corrections to the user.

In an effort to increase the accuracy of velocity measurements, inertial measurement units have been added to the platform to add more information to the velocity solution. In these GPS/Inertial Navigation Systems (GPS/INS), the inertial system measures platform acceleration and rotation using accelerometers and gyroscopes. The GPS/INS then integrates the acceleration measurements to calculate velocity, and integrates again to calculate position, while using the lower rate GPS measurements to bound the high rate inertial systems. However, small bias errors in the raw inertial measurements accumulate over time to yield a drifting velocity and position solution. In order to achieve high accurate velocity measurements, the inertial systems must be high grade and high cost. These GPS/INS may further use Doppler shift alone or Doppler shift in conjunction with a ground station to obtain a more accurate velocity measurement.

Accordingly, there exists a need for an apparatus and method for computing an accurate velocity (and position) solution of a platform without using a ground station by estimating the common errors that DGPS would measure instead of using a ground station to measure these common errors. Further, there is a need for an apparatus and method capable of providing an accurate velocity measurement for a platform at a relatively high update rate. Additionally, there exists a need for an apparatus and method for making inertial measurements of a platform and combining the inertial measurements with GPS measurements made at the platform to provide an accurate velocity measurement of the platform using low cost GPS receivers and inertial systems. Also needed is an apparatus and method using GPS carrier phase (CP) range rate measurements, without using a ground station, that estimates constant CP range rate errors and removes bias errors in inertial measurements to provide an accurate velocity measurement of a platform.

According to the present invention, there is provided an apparatus and method for determining the velocity of a platform using a satellite-based navigation system. The apparatus includes a GPS receiver located at the platform that receives a plurality of GPS signals from a corresponding plurality of GPS satellites and outputs GPS navigation data. An inertial measurement unit located at the platform outputs inertial navigation data relating to the inertial movement of the platform. From the GPS navigation data and the inertial navigation data, the acceleration of the platform is processed. A processor further processes the GPS navigation data and the acceleration of the platform to generate the velocity of the platform.

In a preferred embodiment, the present invention includes a GPS receiver located at the platform for receiving a plurality of GPS signals from a corresponding plurality of GPS satellites and outputting GPS navigation data. An inertial measurement unit is located at the platform for outputting inertial navigation data relating to the inertial movement of the platform. A processor generates a first acceleration measurement of the platform from the GPS navigation data and generates a second acceleration measurement of the platform from the inertial navigation data. The apparatus further includes an acceleration bias remover that receives and combines the first acceleration measurement and the second acceleration measurement to substantially remove the acceleration bias generated by the inertial measurement unit. A processor processes the GPS navigation data and the substantially bias-free acceleration to generate a velocity signal indicative of the velocity of the platform. Preferable, the processor includes a Kalman filter for filtering GPS systematic errors from the velocity signal.

For a more complete understanding of the present invention and the advantages thereof, reference is made to the following detailed description taken in conjunction with the accompanying drawings wherein:

FIG. 1 is a block diagram illustrating the apparatus in accordance with the present invention;

FIG. 2 is a block diagram illustrating a processor shown in FIG. 1 in accordance with the present invention;

FIG. 3 illustrates an acceleration bias remover and a processor for generating the velocity of the platform;

FIG. 4 illustrates the acceleration bias remover including a Kalman filter;

FIG. 5 illustrates a method for determining the velocity of a form without using a ground station; and

FIG. 6 illustrates the processing step shown in FIG. 5.

With reference to the drawings, like reference characters designate like or similar parts throughout the drawings.

With reference to FIG. 1, there is shown an apparatus in accordance with the present invention. A platform 10 is shown having an inertial measurement unit (IMU) 12 and a GPS receiver 14 located at the platform 10. As will be appreciated, the platform 10 can be a land, sea and/or air based vehicle or object. The GPS receiver 14 includes an antenna 16 receiving a plurality of GPS signals 18 from a plurality of GPS satellites 20. The GPS receiver 14 generates GPS navigation data from the GPS signals 18 and transmits the GPS navigation data over bus 22 to a processor 24. The IMU 12 measures inertial acceleration and rotation of the platform 10 and transmits this inertial navigation data to the processor 24 over the bus 22. The processor 24 calculates or computes the velocity of the platform from the inertial navigation data and the GPS navigation data received from the IMU 12 and the GPS receiver 14, respectively.

Now referring to FIG. 2, there is shown a more detailed description of the processor 24 shown in FIG. 1. The processor 24 receives body acceleration and platform attitude data 32 generated by the IMU 12. The body acceleration and platform attitude is measured by the IMU 12 in a reference field with respect to the platform 10 itself. That is, the body acceleration and platform attitude measurements are referenced to a particular reference point peculiar to the platform 10. For example, if the platform 10 was an aircraft, the body acceleration would generally have three acceleration components--(1) the pointing direction of the aircraft, (2) direction of the right wing and (3) direction straight up from the aircraft. Further, the platform attitude includes roll, pitch, and yaw data measurements.

Using this information, the processor 24 transforms body acceleration of the platform 10 to locally level acceleration data 42 of the platform. The locally level acceleration data 42 is a three-component acceleration measurement including east direction, north direction and up direction acceleration components. As will be appreciated, any desired coordinate system or systems may be used, however, those disclosed are preferred.

The processor 24 also receives ephemeris data 26, pseudorange data 28 and carrier phase range rates 30 generated by the GPS receiver 14. The ephemeris data includes the set of orbital parameters for each of the GPS satellites 20. The pseudorange data 28 includes pseudorange measurements calculated by the GPS receiver 14. Depending on the particular GPS receiver 14 used, the carrier phase range rates 30 may either be true carrier phase range rates or carrier phase range measurements generated by the GPS receiver 14. If the receiver 14 is of a type that generates carrier phase range measurements, the present invention will differentiate the carrier phase range measurements to generate the carrier phase range rates 30. In the preferred embodiment, the GPS receiver 14 generates carrier phase range rates 30. From the received ephemeris data 26, the processor 24 calculates or computes satellite (SV) velocity 34 and SV position 36 of each of the GPS satellites. Using the SV velocity 34, SV position 36, and pseudorange data 28, the processor 24 calculates cosine direction vectors 38. The direction vectors 38 are the directional vectors of the platform 10 to each of the GPS satellites.

The processor 24 receives the carrier phase range rates 30 and generates an atmospheric delay estimate 39. The atmospheric delay estimate 39 includes two components--(1) ionospheric group delay and tropospheric delay. The residual ionospheric group delay is calculated using the difference between the carrier phase range rates 30 and both the L1 and the L2 carrier frequencies, sometimes referred to as dual frequency difference or correction. The tropospheric delay is a function of SV altitude, temperature, SV elevation angle and other atmospheric conditions, and is calculated according to a model well known in the art. Once the atmospheric delay estimate 39 is calculated, the processor 24 compensates for the delay estimate 39 to generate atmospheric compensated carrier phase range rates 40.

As shown in FIG. 2, the locally level acceleration data 42, the SV velocity 34, the direction vectors 38, the compensated carrier phase range rates 40 and the uncompensated carrier phase range rates 30 are all input to a subprocessor 44 for calculating the velocity of the platform 10. As will be appreciated, the subprocessor 44 and the processor 24 can be the same processor or two separate processors. Further, the numbers in parentheses (e.g. 5 Hz) indicates preferred update rates for these signals or calculations. It will be understood that other update rates may be used depending on the equipment used and/or speed of the processors.

Now referring to FIG. 3, there is illustrated the subprocessor 44 shown in FIG. 2. The subprocessor 44 performs a least squares calculation using the carrier phase range rates 30, the SV velocity 34 and the direction vectors 38 to calculate GPS-based acceleration data 46 of the platform 10. The subprocessor 44 includes an acceleration bias remover 48 (or acceleration smoother) receiving the locally level acceleration 42 (IMU-based) and the GPS-based acceleration 46 and generating an acceleration 49 of the platform 10. The acceleration bias remover 48 filters acceleration bias errors generated by the IMU 12 and inherent in the locally level acceleration 42. The acceleration bias remover 48 "smooths" the acceleration measurements and generates a smoothed acceleration measurement 49 for the platform 10.

Now referring to FIG. 4, there is shown a more detailed illustration of the acceleration bias remover 48. In the preferred embodiment, the acceleration bias remover 48 includes a one-state Kalman filter 54 that removes or filters the bias from the IMU acceleration 42 by bounding it with the GPS-based acceleration 46 to generate the acceleration 49 of the platform. The equations for the acceleration bias remover are as follows:

where:

A.sup.+ =Acceleration estimate after update

A.sup.- =Acceleration estimate before update

a.sup.imu =Inertial acceleration

Q=Process noise

P=Acceleration estimate covariance

a.sup.gps =Pseudorange derived acceleration

n=current epoch

n-1=previous epoch

R=Measurement noise

Now referring back to FIG. 3, the bias-removed acceleration 49 of the platform 10, the SV velocity 34, the direction vectors 38 and the compensated carrier phase range rates 40 are input to a processor 50. The processor 50 calculates the velocity of the platform 10. The processor 50 further includes a Kalman filter 52 for estimating constant carrier phase range rate errors and filtering these errors. These errors include the systematic errors found in the satellite-user GPS system. The following equation shows the typical carrier phase measurement and the associated errors:

where:

R.sub.T =True Range

N.sub..lambda. =Integer ambiguity

E.sub.I =Residual error due to ionospheric delay after dual frequency correction

E.sub.T =Residual error due to ionospheric delay after tropospheric correction model

E.sub.CL =Error due to user clock offset

E.sub.EPH =Precision Positioning Service ephemeris error including unmodeled SV clock error

E.sub.RC =Correlated receiver error (i.e. multipath, tracking loop errors, inter-channel bias)

E.sub.RUC =Uncorrelated receiver errors (measurement noise)

Because all terms in Equation 7 are close to a constant bias, each term can be estimated using the Kalman filter 52 and removed from the carrier phase range rates. The following sets forth the preferred state vector for the Kalman filter 52:

Now referring to FIG. 5, there is shown a method 200 for determining the velocity of a platform without using a ground reference station. In a first step 202, a plurality of GPS signals 18 are received at the platform 10 by the antenna 16 of the GPS receiver 14. At a step 204, a step 206 and a step 208, the GPS receiver 14 generates ephemeris data 26, pseudorange data 28 and carrier phase range rate data 30, respectively. From this GPS data, the GPS-based acceleration 46 of the platform is computed at step 210.

In a step 212, inertial movement of the platform 10 is measured by the IMU 12 located at the platform 10. Inertial navigation data is generated at step 214 from the inertial movement measured by the IMU 12. From the inertial navigation data, the IMU-based acceleration of the platform 10 is computed at step 216 to generate the locally level acceleration 42.

In a step 218, the GPS-based acceleration 46 and the locally level acceleration (IMU-based acceleration) 42 are combined and filtered to substantially remove acceleration bias generated by the IMU 12. As will be appreciated, the filtering step 218 is performed by the acceleration bias remover 48 including the one-state Kalman filter 54. The equations for the Kalman filter 54 are described in detail in the preceding description. In step 220 the acceleration 49 of the platform 10 is computed. The acceleration 49 of the platform 10 and the GPS navigation data (ephemeris data 26, pseudorange data 28 and carrier phase range rate data 30) are processed in step 222 by the processor 24 and the velocity of the platform 10 is computed.

Now referring to FIG. 6, there is illustrated a detailed description of the processing step 222 shown in FIG. 5. The processor 24 computes the SV position 36 for each of the GPS satellites in step 230 and computes the SV velocity 34 of each of the GPS satellites in step 232. Further, in step 234, carrier phase range rates 40 are computed for each of the GPS satellites. In this same step, an ionospheric group delay and tropospheric delay 39 are calculated, respectively, using the L1 and L2 carrier frequencies and a tropospheric delay model that is well known in the art. The delays 39 are used to compensate for atmospheric delay associated with the GPS system and atmospheric compensated carrier phase range rates 40 (for each GPS satellite) are generated. In step 236, the processor 24 computes direction vectors relating to the direction of the platform 10 with respect to each of the GPS satellites.

Filtering, in step 238, substantially removes constant errors generated by the GPS system that occur in the carrier phase range rate measurement. These systematic errors have been described in detail in the above description of the carrier phase measurement equation. The filtering step 238 is performed, preferably, using an 8+n state Kalman filter 52, where n equals the number of GPS satellites used. The Kalman filter 52 includes six states accounting for acceleration and velocity of the platform 10, two states for frequency offset and frequency offset drift of GPS user clock, and n range rate bias states. In step 240, the velocity of the platform 10 is computed using the GPS navigation data and the acceleration of the platform in conjunction with the Kalman filter 52. Since most GPS receivers cannot provide accurate carrier phase range rate measurements at a very high rate, the propagation cycle of the Kalman filter 52 is used to provide a preferred update rate of about between 40-60 Hertz. This is accomplished by simply propagating forward the solution with the estimated acceleration and velocity of the platform 10.

Although a preferred embodiment of the present invention has been described in the foregoing detailed description and illustrated in the accompanying drawings, it will be understood by those skilled in the art that the invention is not limited to the embodiment disclosed but is capable of numerous rearrangements, substitutions and modifications without departing from the spirit of the invention.