Previous Abstract Return to Session A6 Next Abstract

Session A6: Adaptive KF Techniques, Data Integrity, and Error Modeling

Multi-Faults Detection and Isolation for Redundant Strapdown Inertial Navigation System
Jianhua Cheng , Xiangyu Sun, Daidai Chen, Chun Cheng, Hongjie Mou, Ping Liu
Location: Big Sur
Alternate Number 2

Traditional generalized likelihood test method can detect faults while can not isolate multiple faults in redundant strapdown inertial navigation system. In this paper, a multi-faults detection and isolation method based on generalized likelihood test and linear prediction approach is proposed. A generalized likelihood test method is used to detect the fault of the system, and the linear prediction method is used to estimate the values of inertial sensors. The faulty inertial sensors can be isolated through comparing the estimated values with measurements of inertial sensors, and the faulty information is recorded for system reconstruction. Through simulation and experiment, it is proved that new method can accurately detect and isolate multiple faults and thus guarantee the reliability of strapdown inertial navigation system.



Previous Abstract Return to Session A6 Next Abstract