Sensitivity applied model-based fault diagnosis algorithm for vehicle control system sensors

Abstract

Important vehicle control systems, such as ESC, MDPS, and ECS, enhance convenience, vehicle stability, and driver's safety. These functions work with vehicle sensors that inform motorists of the vehicle's condition. Therefore, the failure of the input sensor in each system can affect the functions of the entire system, increasing the risk of accidents as well as causing inconvenience to the driver. In this paper, a new method in determining and isolating faulty sensors in a vehicle's control system was investigated. The proposed method calculated the constraints and the residuals by applying the model-based fault diagnosis method to the sensors of the chassis system. To solve the inaccuracies of fault isolation in residuals, the sensitivity analysis was applied to the residual incident matrix to derive a fault isolation table, and to establish a fault isolation strategy. To verify the fault diagnosis algorithm by using the fault diagnotor developed in this study, the fault was identified and verified in the HILS environment with the RCP device.