SENSOR/ACTUATOR FAULT DETECTION, ISOLATION AND ACCOMMODATION APPLIED TO UAV MODEL

In this study sensor/actuator fault detection, isolation and accommodation is performed using innovation sequence analysis in a small Unmanned Aerial Vehicle (UAV) model. Sensor/actuator fault detection and isolation is an important part of a fault tolerant control system. In the study ways of determining the faults that occur in sensors and actuators, distinguishing between them and isolating the faults in the sensors is described. Kalman filtering technique is used to build a fault detection and isolation algorithm and a robust Kalman filter is designed to distinguish between actuator and sensor faults. A simplified robust adaptive Kalman filter is also built to estimate the states in case of measurement system malfunctions. The small UAV model is used for simulations and faults occurring in speed sensors and elevators are simulated.