Fault Diagnosis Strategy for Train-ground Communication System of Urban Mass Transit Using Reliability Analysis

This paper proposes a novel diagnosis strategy for Train-ground Communication System (TCS) of urban mass transit, which makes full use of the advantages of both fault tree for modeling and Bayesian Networks (BN) for the inference ability. Specifically, it adopts the dynamic fault tree model to capture the dynamic failure mechanisms, mines some qualitative structure and quantitative parameters provided by reliability analysis, and incorporates sensors data as well as test cost into fault diagnosis to design an efficient diagnosis algorithm. From experimental results, it can easily be inferred that the proposed methodology can locate the fault in the minimum diagnostic cost as compared to other approaches. This diagnosis strategy can provide an effective intelligent assistant decision for the fault location of TCS, avoid blind attempts and increase greatly its service efficiency.