Adaptive GPS and INS integration system

• US 6,900,760 B2
• Filed: 05/23/2001
• Issued: 05/31/2005
• Est. Priority Date: 06/07/2000
• Status: Expired due to Term

• Current Assignee: Qinetiq Limited (QinetiQ Group PLC)
• Original Assignee: Qinetiq Limited (QinetiQ Group PLC)
• Inventors: Groves, Paul D.
• Primary Examiner(s): Issing, Gregory C.
• Application Number:
  US10/297,663
• Publication Number:
  US 20040012522A1
• Time in Patent Office: 1,469 Days
• Field of Search: 342/357.14, 701/216
• US Class Current: 342/357.65
• CPC Class Codes: G01C 21/165 combined with non-inertial navigation instruments
• CPC Class Codes: G01S 19/26 involving a sensor measurement for aiding acquisition or tracking
• CPC Class Codes: G01S 19/47 the supplementary measurement being an inertial measurement, e.g. tightly coupled inertial

Abstract

This invention relates to the field of Inertial Navigation Systems (INS) and satellite navigation systems such as the Global Positioning System (GPS) and in particular relates to methods of integrating GPS and INS data in order to provide more accurate navigation solutions. The invention provides a method of improving the accuracy of a tightly coupled integrated INS and satellite radio navigation system in cases where the satellite navigation receiver has adaptive tracking loop bandwidths.

Claims

1. A method of integrating inertial navigation system and satellite navigation system data in a tightly coupled architecture by means of a Kalman filter, the satellite navigation data being received on a receiver comprising tracking loops having adaptive bandwidths, comprising

monitoring the tracking loop bandwidths or modelling them as a function of the receiver measured signal to noise density ratio (c/n₀) outputs, and

varying the rate of response of the Kalman filter to measurements from the satellite navigation system in response to changes in the tracking loop bandwidths such that correlated measurement noise in the Kalman filter is avoided.

2. A method as claimed in claim 1 wherein the Kalman filter is varied for each tracking loop.

3. A method as claimed in claim 1, wherein the iteration rate of the Kalman filter is varied in inverse proportion to the tracking loop bandwidth.

4. A method as claimed in claim 1, wherein measurement data averaged is averaged over the iteration period.

5. A method as claimed in claim 1, the Kalman filter having a Kalman gain matrix and wherein said Kalman gain matrix is weighted by multiplying it by the time between successive measurements divided by the time between successive uncorrelated measurements.

6. A method as claimed in claim 1, the Kalman filter having a Kalman gain matrix and wherein said Kalman gain matrix is represented as one n_t column matrix for pseudo-range measurements and a second n_t column matrix for pseudo-range rate measurements, where n_t is the number of satellites tracked, and is then weighted by an adaption matrix A, the adaption matrix being an n_t x n_t diagonal matrix where

the elements of the pseudo-range adaption matrix are

A_ρ

7. A method as claimed in claim 1, the Kalman filter comprising a Kalman gain matrix having a measurement noise covariance, R, and wherein said measurement noise covariance, R, is divided by an adaption matrix A as claimed in claim 5.

8. A method as claimed in claim 1, wherein the measurement noise covariance, R, is weighted by multiplying by the time between successive uncorrelated measurements divided by the time between successive measurements.

9. A method as claimed in claim 1, wherein the satellite navigation system is a Global Positioning System.

10. A method of integrating inertial navigation system and satellite navigation system data in a tightly coupled architecture by means of a Kalman filter, the satellite navigation data being received on a receiver comprising tracking loops having adaptive bandwidths, comprising

monitoring the tracking loop bandwidths or modelling them as a function of the receiver measured signal to noise density ratio (c/n₀) outputs , and

modelling time-correlated measurement noise within the Kalman filter explicitly such that the assumed correlation time within the Kalman filter is varied in inverse proportion to the tracking loop bandwidths.

11. A method as claimed in claim 10 wherein the correlation time is varied independently for measurements from each tracking loop.

12. A method as claimed in claim 10, wherein the correlated measurement noise is estimated as Kalman filter states.

13. A method as claimed in claim 10, wherein the correlated measurement noise is modelled using a Schmidt-Kalman filter algorithm.