Fault Diagnosis Strategy of Hybrid Electric Vehicle Based on FTA−BN

To realize fault fusion diagnosis for multiple components of hybrid electric vehicle (HEV) chassis system, a fault diagnosis strategy based on fault tree analysis and Bayesian network (FTA−BN) was proposed. The fault diagnosis system of HEV chassis was designed based on VB language and SQL Server open source database platform. Based on the constructed fault tree model of HEV chassis system, the fault events were converted into network nodes through the mapping relationship between fault tree and Bayesian network, and the fault events were introduced into the expert system. Based on the mixed fuzzy reasoning and bounded depth first search method, through the qualitative and quantitative analysis of fault symptoms, the reliability of fault events was calculated based on FTA−BN, and the causes with high reliability were selected as the diagnostic conclusions, and the experimental verification was carried out. The results show that the proposed FTA−BN fault diagnosis strategy can be used for the fusion diagnosis of multiple components of HEV chassis system, and the accuracy of fault diagnosis is 0.850, which is higher than that of manual experience judgment (0.675).