PREDICTION OF ANTILOCK BRAKING SYSTEM CONDITION WITH THE VEHICLE STATIONARY USING A MODEL-BASED APPROACH

摘要

An Antilock Braking System (ABS) is one of the most important safety facilities equipped in modern vehicles. A self-test is therefore embedded into its ECU to identify any electronic malfunction. However there is no effective method to predict or check its mechanical conditions routinely to ensure its functionality. Because the ABS system is merely actuated above a particular speed in emergency stops, the current brake test facilities are not adequate for ABS test. Because of the dangers involved it would not be acceptable to use a public road to implement such a practice for fault detection so an alternative means must be sought. To provide a safe and convenient solution this paper proposes a novel method to predict ABS faults whilst the vehicle is stationary. In this situation a model-based approach is applied to predict various faults from the ABS, especially from its hydraulic subsystem. As such, a mathematic model is developed to describe the operating processes of ABS including possible faulty conditions. An autonomous control strategy is also designed to actuate the control module independently without the knowledge of the control algorithms embedded in an ABS control module. This approach is evaluated through a Simulink simulation.