基于三种理论的汽油机故障诊断系统应用研究

Application Research on Fault Diagnosis System of Gasoline Engine Based on Three Theories

阐述了模糊逻辑、BP神经网络和专家系统的相关理论,探讨了结合三种理论构建系统的优点,介绍了电控汽油机故障诊断的相关理论。在上述理论的基础上,阐述了电控汽油机模糊神经网络诊断专家系统的构建方法,并运用MATLAB软件针对汽油机故障征兆-故障模式样本集编写了BP神经网络训练程序对样本进行了学习训练和仿真。

Theories of the fuzzy logic,BP neural network and expert system are introduced,and the advantages of the combining system of three theories are discussed,and the relevance theory of electronic controlled gasoline engine fault diagnosis is introduced. On the basis of the above theories,the construction method of fuzzy neural network diagnostic expert system of electronic controlled gasoline engine is elaborated,and aiming at the gasoline engine fault symptoms-failure mode sample set,BP neural network training program is proposed on samples of learning and training by using MATLAB software.