Fault Detection, Isolation, and Data Fusion for Unmanned Aerial Vehicle Air Data Systems

Abstract

In this chapter, Global Positioning System (GPS)/inertial navigation system (INS) measurements, which have high accuracy, and air data system (ADS) measurements, which have low accuracy but high frequency, are integrated using the Kalman filtering technique in order to obtain high-accuracy measurement data at high frequency. It will be shown that the designed system, which is based on the indirect Kalman filtering technique, is successful in finding the wind speed data using the known error dynamics of the system and the determined statistical values. The fault detection and isolation (FDI) algorithms are developed by using the proposed system and diagnostic tests are performed in order to evaluate the performance of the system under different sensor measurement fault scenarios. For fusing the data coming from different measurement sets, a federated Kalman filter is used. The best data are produced by using the FDI algorithms and leaving the faulty data out before fusing with the federated filter.