分析了基于模型的故障诊断方法的原理和特点以及其他故障诊断方法（基于规则、基于人工神经网、基于案例）的局限性．针对基于模型的故障诊断，提出了运用人工神经网和遗传算法的模糊建模方法，简化了建模复杂性．给出了基于模糊模型的故障诊断推理方法，从而能有效地处理复杂系统的故障诊断 The principle and characteristics of the model-based fault diagnosis method and the limitations of other fault diagnosis methods (rule-based, artificial neural network based, case-based) are analyzed. Aiming at the model-based fault diagnosis, a fuzzy modeling method using artificial neural network and genetic algorithm is proposed, which simplifies the modeling complexity. The fault diagnosis reasoning method based on fuzzy model is given, which can effectively deal with the fault diagnosis of complex systems