用模糊神经网络诊断发电机励磁主

介绍了一种以可控硅整流桥各桥臂两端电压信号作为模糊神经网络的输入信号的实时功率单元故障诊断系统。信号经过模数转换后送给模糊化处理单元，再传递到神经网络；神经网络根据以前学习、训练的知识，得出故障诊断结果。该系统解决了目前我国农村技术力量难以判断农用水利发电机组现代励磁控制系统中电力电子模块故障的难题。文中还介绍了用DSP（TMS320F2407）作为核心器件的控制电路的实现方法。

Fuzzy Neural Network Fault Diagnosis System for Excitation Electric Circuit Power Units of Generator

In the paper,a real-time monitoring and diagnosis system for excitation power units based on fuzzy-neural network according to the voltage signal of each power rectifier is introduced.The voltage signal will be sent to fuzzy-treatment unit after A/D transfer.Then,it is sent to the neural network.The neural network can diagnose faults according to the knowledge learned and trained previously.This system solve the problem that it is difficult to judge the fault of power electronic module in the modern excitation control system of generator unit in the countryside.The realization method of the system using DSP(TMS320F2407)as core unit is introduced in the paper.