패리티 공간 기법과 분산형 필터를 이용한 관성센서의 신호 및 모델 기반의 고장검출 기법 연구

Abstract

Integrity monitoring method of the inertial sensor is necessary to identify the system reliability, because it is impossible to identify the system integrity without detecting and isolating the sensor fault. Thus, this paper deals with the design of signal and model based fault detection algorithm for redundant IMU(Inertial Measurement Unit) using PSA(Parity Space Approach) and decentralized filter. Generally, the benefits of FDI(Fault Detection and Isolation) algorithm using PSA are well known. However it can not detect the multiplicative faults. On the other hands, decentralized filter is not affected by the type of fault. Therefore the integrated FDI method is more powerful FDI solution for the inertial sensor than individual. In the simulation the redundant IMU has a cone-shape configuration and decentralized filter is based on Kalman filter. And the performance of FDI algorithm was verified using MATLAB simulink S/W simulator.