Fault detection and isolation system for four-wheels drive electric vehicles

In this paper, we investigate the problem of fault detection and isolation (FDI) of actuator and sensor faults which influence the lateral behavior of automated vehicles. Fast identification of faulty actuator or sensor is crucial for the stability of the vehicle and is considered as an important step in the fault-tolerant vehicle control design strategy. The steering system actuator and motor torque faults and yaw rate and lateral accelerations sensor faults are considered. To reduce the impact of model uncertainties, perturbation and sensor noise, a robust unknown input observer and H_/H-infinity fault detection filter are selected and applied for generating the residual signals. In addition, a decision algorithm (residual evaluation) based on the generalized likelihood ratio test (GLRT) is used for detecting and isolating the faulty signal. Numerical simulations demonstrate the effectiveness of the proposed approach.