车载燃油系统故障诊断智能算法开发及验证

This paper proposes an intelligent multi-class fault diagnosis algorithm for the diesel engine fuel system based on PCA and ELM. Firstly, a fault diagnosis model is established in Simulink, utilizing real vehicle data to support offline verification. Subsequently, an architecture for the intelligent fault diagnosis algorithm is designed, following the OSA-CBM standard. Based on this architecture, the online fault diagnosis algorithm is developed in the Simulink and tested using a hardware-in-the-loop (HIL) platform. The verification results show that the proposed intelligent multi-class fault diagnosis algorithm achieves high accuracy and reliability in both offline simulations and HIL testing, indicating its potential for in-vehicle applications.