汽车电控发动机故障诊断研究

李嘉锋　韦忠劲

摘 要：為了改善电控发动机的故障诊断水平，为汽车作业创造更为优质的条件，本文选取RBF神经网络作为研究工具，根据电控发动机作业原理，构建故障诊断函数，在MATLAB环境下编译诊断程序。为了检验本文提出的故障诊断方案可靠性，本研究设置了3种故障状态，对这些故障诊断进行仿真。仿真结果显示，本文提出的故障诊断方案符合诊断精度、诊断效率改善要求，可以推广应用。

关键词：RBF神经网络 故障诊断 电控发动机

Research on Fault Diagnosis of Automobile Electronic Control Engine

Li Jiafeng Wei Zhongjin

Abstract：To improve the fault diagnosis level of the electronic control engine and create better conditions for automobile operation， this paper selects RBF neural network as the research tool， constructs the fault diagnosis function according to the operation principle of the electronic control engine， and compiles the diagnosis program in the MATLAB environment. In order to test the reliability of the fault diagnosis scheme proposed in this paper， this research has set up three kinds of fault states to simulate this fault diagnosis. The simulation results show that the fault diagnosis scheme proposed in this paper meets the requirements of improving diagnosis accuracy and diagnosis efficiency and can be popularized and applied.

Key words：RBF neural network， fault diagnosis， electronic control engine