Multiple-Model-Based Diagnosis of Multiple Faults With High-Speed Train Applications Using Second-Level Adaptation

Abstract

Due to the time-varying characteristics and the interacted nature of multiple faults in the high-speed train (HST), the fault modeling, isolation, and severity estimation cannot be described accurately using a single model, which may result in poor performance of the conventional fault diagnosis methods. This article introduces the idea of multiple models and second-level adaptation techniques to diagnose multiple faults of the HST traction motor. First, a reduced model description for the multiple faults is given. Then, a multiple fault isolation framework is developed to simplify the time-varying fault parameters space segmentation. Based on the decoupled fault set, a fault estimation scheme with second-level adaptation is used to provide a reliable alarm priority for different fault scenarios. A case study is performed to verify the effectiveness of the proposed approach.