基于神经网络的发动机故障诊断分析

Diagnosis and analysis on the engine failures based on neutral network

首先简要介绍了汽车发动机故障诊断系统及基于神经网络的发动机故障诊断分析方法,其次利用神经网络知识建立起带有单隐层的三层神经网络;在理论分析的基础上,利用MATLAB软件的simulink和神经网络工具箱,分别对发动机故障进行检测和分析,仿真结果表明:利用BP神经网络对发动机故障进行检测具有检测精度高、速度快的特点。

This paper briefly introduces diagnostis system and the analytical method for the engine failures based on the neutral network. Three - layer neutral network model is created by using the neuronic knowledge. Based on the theoretical analysis foundation, the author uses simulink and neutral network toolboxes to detect and analyze the engine failure. Results found that the application of BP network in engine failure has some specific natures such as high -accuracy and similar intellectual ability.