Systems and methods for attitude fault detection based on integrated GNSS/inertial hybrid filter residuals
Systems and methods for attitude fault detection based on integrated GNSS/inertial hybrid filter residuals are provided. In one embodiment, a fault detection system for aircraft attitude measurement system comprises: a sensor monitor coupled to a first inertial measurement unit, the sensor monitor comprising: a navigation error model for the first inertial measurement unit, the model configured to model a plurality of error states including at least an attitude error state vector, an velocity error state vector, and a position error state vector determined from data generated by the first inertial measurement unit; and a propagator-estimator configured to propagate and update error states based on GNSS data; and a residual evaluator configured to input measurement error residual values generated by the propagator-estimator, wherein the residual evaluator outputs an alert signal when the measurement error residual values exceed a threshold.
1. A fault detection system for aircraft attitude measurement system, the fault detection system comprising:
a sensor monitor coupled to a first inertial measurement unit of the aircraft attitude measurement system, the sensor monitor comprising:
a navigation error model for the first inertial measurement unit, the navigation error model configured to model a plurality of error states including at least an attitude error state vector, an velocity error state vector, and a position error state vector determined from data generated by the first inertial measurement unit; and
a propagator-estimator configured to propagate and update the plurality of error states from the navigation error model based on GNSS data; and
a residual evaluator configured to input measurement error residual values generated by the propagator-estimator, wherein the residual evaluator outputs an alert signal when the measurement error residual values exceed a predetermined statistical threshold.
2. The fault detection system of claim 1 , wherein the GNSS data is either pseudo ranges, GNSS position data or GNSS velocity data.
3. The fault detection system of claim 1 , wherein the propagator-estimator is a Kalman filter.
4. The fault detection system of claim 1 , wherein a pure inertial attitude output from the first inertial measurement unit is added as a measurement to the propagator-estimator.
5. The fault detection system of claim 1 , further comprising a display;
wherein the alert signal produces an alert on the display that indicates that the first inertial measurement unit is faulted.
6. The fault detection system of claim 1 , further comprising:
a second sensor monitor coupled to a second inertial measurement unit of the aircraft attitude measurement system, the second sensor monitor comprising:
a second navigation error model for the second inertial measurement unit, the navigation error model configured to model a plurality of error states including at least an attitude error state vector, an velocity error state vector, a position error state vector determined from data generated by the second inertial measurement unit; and
a second propagator-estimator configured to propagate and update the plurality of error states from the second navigation error model based on the GNSS data.
7. The fault detection system of claim 6 , wherein the residual evaluator outputs the alert signal when the residual error values from the propagator-estimator associated with the first inertial measurement unit exceed the predetermined statistical threshold but residual error values from the second propagator-estimator associated with the second inertial measurement unit do not exceed the predetermined statistical threshold.
8. The fault detection system of claim 1 , wherein the measurement error residual values include measurement errors for one or both of pitch measurements and roll measurements generated by the first inertial measurement unit.
9. The fault detection system of claim 1 , wherein the sensor monitor is integral to the first inertial measurement unit.
10. An inertial measurement system, the system comprising:
an inertial measurement unit on-board an aircraft, the inertial measurement unit configured to output attitude measurements of the aircraft;
a sensor monitor coupled to the inertial measurement unit, the sensor monitor comprising:
a navigation error model for the inertial measurement unit, the navigation error model configured to model a plurality of error states including at least an attitude error state vector, an velocity error state vector, and a position error state vector determined from data generated by the first inertial measurement unit; and
a propagator-estimator configured to propagate and update the plurality of error states from the navigation error model based on GNSS data; and
a residual evaluator configured to input measurement error residual values generated by the propagator-estimator, wherein the residual evaluator outputs an alert signal when the measurement error residual values exceed a predetermined statistical threshold.
11. The system of claim 10 , further comprising a display;
wherein the alert signal produces an alert on the display that indicates that the first inertial measurement unit is faulted.
12. The system of claim 10 , wherein the residual evaluator outputs the alert signal when the residual error values from the propagator-estimator associated exceed the predetermined statistical threshold but residual error values from a second propagator-estimator associated with a second inertial measurement unit do not exceed the predetermined statistical threshold.
13. The system of claim 10 , wherein the sensor monitor is integral to the inertial measurement unit.
14. A fault detection method for an aircraft attitude measurement system, method comprising:
monitoring attitude solution data generated by a first inertial measurement unit of an aircraft attitude measurement system;
executing a navigation error model for the first inertial measurement unit, the navigation error model configured to model a plurality of error states including at least an attitude error state vector, an velocity error state vector, and a position error state vector determined from data generated by the first inertial measurement unit;
generating measurement error residual values using a propagator-estimator, wherein the propagator-estimator is configured to iteratively update the plurality of error states from the navigation error model based on GNSS data; and
comparing the measurement error residual values against a predetermined statistical threshold and generating an alert signal when the measurement error residual values exceed the predetermined statistical threshold.
15. The method of claim 14 , wherein the GNSS data is either pseudo ranges, GNSS position data or GNSS velocity data.
16. The method of claim 14 , wherein the propagator-estimator is a Kalman filter.
17. The method of claim 14 , wherein a pure inertial attitude output from the first inertial measurement unit is added as a measurement to the propagator-estimator.
18. The method of claim 14 , wherein the alert signal produces an alert on a display that indicates that the first inertial measurement unit is faulted.
19. The method of claim 14 , wherein the measurement error residual values include measurement errors for one or both of pitch measurements and roll measurements generated by the first inertial measurement unit.
20. The method of claim 14 , further comprising:
producing the alert signal when the measurement error residual values from the propagator-estimator exceed the predetermined statistical threshold but measurement error residual values from a second propagator-estimator associated with a second inertial measurement unit do not exceed the predetermined statistical threshold.