Wavelet and Parity Space based Actuator Fault Detection for Unmanned Aerial Vehicle

• 1、Instrument Science and Electrical Engineering School,Beihang University,Beijing, 100191

Abstract：For improving the security and reliability of the Unmanned Aerial Vehicle (UAV) flight Control system (FCS), the study of the UAV actuator fault diagnosis problem is very important. This paper combined the wavelet transform with parity space based method for UAV actuator fault detection. Though computing the stationary wavelet transform of the residual with low order parity relation, the frequency band of the low order parity vector was divided into multiple narrower frequency bands, correspondingly a group of residuals containing different frequency information of fault were obtained, then residual evaluation function and thresholds were selected for residual evaluation, then actuator fault detection was accomplished. This approach can not only overcome the defect of traditional parity space based method that the faults within broader frequency range could hardly be detected, but also achieve good performance index with low computational effort. Finally, taking a certain type of UAV as an example, a simulation was produced.

Keywords： fault detection unmanned aerial vehicle actuator wavelet transform parity space