Evaluation of a Fault Detection Method for Redundant IMU Systems

Johanna Matthaei, Ulf Bestmann, Peter Hecker

Abstract:	This work presents a method of fault detection with only two MEMS based inertial measurement units (IMUs) in combination with Global Navigation Satellite System (GNSS) data. The two identically constructed IMUs are mounted in a skid steered rover being part of a swarm for Mars exploration. The challenge is the detection of malfunctioning of a sensor and the identification of the source of error with two redundant sensors. The IMU’s sensors are segmented in two groups, the accelerometers and the gyroscopes. Therefore four different results of a navigation solution are possible. The base of the navigation system is a common coupling method of inertial sensor data and GNSS data with an Extended Kalman Filter (EKF). With all possible IMU data combinations, four EKFs are operating parallel. The observation of the residuals in the EKF allows a statement about the error source.
Published in:	Proceedings of the ION 2015 Pacific PNT Meeting April 20 - 23, 2015 Marriott Waikiki Beach Resort & Spa Honolulu, Hawaii
Pages:	1024 - 1030
Cite this article:	Matthaei, Johanna, Bestmann, Ulf, Hecker, Peter, "Evaluation of a Fault Detection Method for Redundant IMU Systems," Proceedings of the ION 2015 Pacific PNT Meeting, Honolulu, Hawaii, April 2015, pp. 1024-1030.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In