A fault detection and isolation scheme for lateral vehicle dynamics of EVs using a quantitative parity space approach

Advanced control concepts for electric vehicles assume the correct functioning of sensors. In this contribution the problem of sensor failure detection and isolation in the case of lateral vehicle dynamics of EVs is tackled. The method depends on only basic vehicle parameters and assumes only a single failure in the vehicle system. A single sensor failure can be detected and the false alarm rate is minimized. The approach is based on the parity space method implemented with residual observers. Noise of sensors is taken into account and appropriate bounds in the feature space are derived in order to get a robust fault detection indicator. Simulation results and an experimental setup show the validness of the approach.