METHOD OF DETERMINING NAVIGATION PARAMETERS FOR A CARRIER AND HYBRIDIZATION DEVICE ASSOCIATED WITH KALMAN FILTER BANK

• US 20120105278A1
• Filed: 07/09/2010
• Published: 05/03/2012
• Est. Priority Date: 07/10/2009
• Status: Active Grant

• Current Assignee: Safran Electronics & Defense
• Original Assignee: Sagem Defense Securite (Safran SA)
• Inventors: Riedinger, Didier, Vourch, Sébastien, Hoblos, Ghaleb
• Granted Patent: US 9,000,978 B2
• Time in Patent Office: Days
• US Class Current: 342/357.3
• CPC Class Codes: G01C 21/165 combined with non-inertial navigation instruments
• CPC Class Codes: G01S 19/20 Integrity monitoring, fault detection or fault isolation of space segment
• CPC Class Codes: G01S 19/47 the supplementary measurement being an inertial measurement, e.g. tightly coupled inertial

Abstract

A process for determination of navigation parameters of a carrier by a hybridisation device comprising a bank (3) of Kalman filters, each working out a hybrid navigation solution from inertial measurements calculated by a virtual platform (2) and raw measurements of signals emitted by a constellation of satellites supplied by a satellite-positioning system (GNSS), characterised in that it comprises the steps of:

• • determination for each satellite of at least one probability ratio between a hypothetical breakdown of given type of the satellite and a hypothetical absence of breakdown of the satellite,
  • declaration of a breakdown of given type on a satellite as a function of the probability ratio associated with this breakdown and of a threshold value,
  • estimation of the impact of the breakdown declared on each hybrid navigation solution, and
  • correction of hybrid navigation solutions as a function of the estimation of the impact of the breakdown declared.

Claims

1. A process for determination of navigation parameters of a carrier by a hybridisation device comprising a bank (3) of Kalman filters each working out a hybrid navigation solution from inertial measurements calculated by a virtual platform (2) and raw measurements of signals emitted by a constellation of satellites supplied by a satellite-positioning system (GNSS), characterised in that it comprises the steps of:

determination, for each satellite, of at least one (Ir, Ir′

) probability ratio between a hypothetical breakdown of given type of the satellite and a hypothetical absence of breakdown of the satellite,

declaration of a breakdown of given type on a satellite as a function of the (Ir, Ir′

) probability ratio associated with this breakdown and a threshold value,

estimation of the impact of the breakdown declared on each hybrid navigation solution, and

correction of the hybrid navigation solutions as a function of the estimation of the impact of the breakdown declared.

2. The process as claimed in the preceding claim, in which, for each satellite, at least one (Ir, Ir′

) probability ratio is determined at each incrementation of the Kalman filter on a sliding storage window, a breakdown of given type being declared if the sum of the (Ir, Ir′

) probability ratios associated with this breakdown on the sliding storage window is greater than the associated threshold value.

3. The process as claimed in the preceding claim, also comprising, if a breakdown is declared, a step of exclusion of raw measurements of signals emitted by the satellite declared in breakdown.

4. The process as claimed in the preceding claim, also comprising a step breakdown confirmation test comprising conducting a statistical test on innovations over a determined period following a declaration of breakdown, each innovation reflecting the spread between an observation corresponding to a raw measurement originating from a satellite not declared in breakdown and a posterriori estimation of said observation worked out by a Kalman filter using raw measurements originating from all the satellites of the constellation, the declaration of the satellite in breakdown being cancelled if the statistical test is negative.

5. The process as claimed in any one of the preceding claims, also comprising a calculation step, for each satellite, of cross innovation and covariance of cross innovation reflecting the spread between an observation corresponding to a raw measurement originating from the satellite and an a posterriori estimation of said observation worked out by a Kalman filter not using the raw measurement originating from the satellite, the (Ir, Ir′

) probability ratio being determined as a function of cross innovation and of cross covariance innovation provided by said Kalman filter not using the raw measurement originating from the satellite and from estimation of the impact of the breakdown of given type on said cross innovation.

6. The process as claimed in the preceding claim, in which the impact of the breakdown of given type on said cross innovation is estimated by determination, from data provided by said Kalman filter not using the raw measurement originating from the satellite, of a dynamic innovation matrix, and estimation of an amplitude of the breakdown of given type on a sliding estimation window corresponding to a given number (N) of incrementations of said Kalman filter not using the raw measurement originating from the satellite, the dynamic innovation matrix representing a link between the amplitude of the breakdown and the spread caused by the breakdown on cross innovation.

7. The process as claimed in the preceding claim, in which the given number (N) of incrementations of said Kalman filter corresponds to a duration lasting less than a predetermined detection period (T).

8. The process as claimed in claims 5 and 6, in which estimation of the amplitude of the breakdown of given type is done on the sliding estimation window by a minimisation method by the least-squares method or else by a Kalman filter.

9. The process as claimed in any one of the preceding claims, in which the estimation of the impact of the breakdown on each hybrid navigation solution comprises determination, for each hybrid navigation solution, from data provided by the Kalman filter working out the hybrid navigation solution, of a dynamic navigation matrix, and estimation of an amplitude and covariance of the breakdown of given type on a sliding estimation window corresponding to a given number of incrementations of said Kalman filter working out the hybrid navigation solution, the dynamic navigation matrix representing a link between the amplitude of the breakdown and the spread caused by the breakdown on said hybrid navigation solution.

10. The process as claimed in any one of the preceding claims, in which two (Ir, Ir′

) probability ratios are determined for each satellite, one (Ir) ratio being associated with a hypothetical breakdown of bias type and the other (Ir′

) ratio being associated with a hypothetical breakdown of ramp type.

11. The process as claimed in claims 8 and 9, in which, if a breakdown of bias type is declared during the declaration step, and the estimation step of the impact of the breakdown on each hybrid navigation solution is done on a sliding estimation window from the moment of declaration of the breakdown of bias type.

12. The process as claimed in claims 8 and 9, in which, if a breakdown of ramp type is declared, the step of estimation of the impact of the breakdown 23 on each hybrid navigation solution is done on the sliding estimation window preceding the moment of declaration of the breakdown of ramp type.

13. The process as claimed in any one of the preceding claims, in which, if several sums of (Ir, Ir′

) probability ratios are greater than their associated threshold value, a single breakdown is declared during the declaration step, said breakdown corresponding to the greater of the sums of probability ratios.

14. A hybridisation device comprising a virtual platform (2) capable of calculating inertial measurements, a Kalman filter (3) designed to devise a hybrid navigation solution from inertial measurements and raw measurements of signals emitted by constellations of satellites delivered by a satellite-positioning system (GNSS), characterised in that it comprises:

a detection module (41) configured to determine for each satellite at least one (Ir, Ir′

) probability ratio between a hypothetical breakdown of a given type by ratio and hypothetical absence of breakdown of the satellite and to declare a breakdown of given type as a function of the (Ir, Ir′

) probability ratio associated with the breakdown of given type is of a threshold value,

an accommodation module (5) configured to estimate the impact of the declared breakdown on the hybrid navigation solution devised by the Kalman filter, and to correct the hybrid navigation solution as a function of the estimation of the impact of the declared breakdown.