神经网络专家系统在发动机故障诊断中的应用

将神经网络与专家系统相结合形成了神经网络专家系统，其核心部分采用BP神经网络．以发动机为诊断对象，获取故障数据后，采用3层BP神经网络模型对网络进行训练，经过几种不同网络参数的对比后，设定了合理的网络训练参数，训练后获得了期望的误差值0．01，并采用VB6．0软件设计了专家系统的人机交互界面，为专家系统在发动机故障诊断中的应用提供了有益的参考．

Application of Neural Network Expert System to Fault Diagnosis of Engine

A Neural Network Expert System is established when Neural Network is combined with Expert System,the core of system adopts BP Neural Network.The diagnosis of engine is regarded as an example,after obtaining diagnosis data,it adopts three layers BP Neural Network model to carry on the training to the network,after contrasting a few different network parameters,sets reasonable network training parameters,acquires the expected error margin worth 0.01 after training,and adopts person machine that the software of VB 6.0 designed the Expert System to hand over each other's interface,and the result provides the beneficial reference to Expert System in the application of the vehicle engine fault diagnosis.