基于遗传神经网络的汽轮发电机组故障诊断研究  

Fault Diagnosis Scheme for Turbo-Generator Based on Genetic Neural Network

作　　者：陈长征[1] 勾轶[1] 王毅[1] 唐任远[1] 

机构地区：[1]沈阳工业大学诊断与控制中心,沈阳110023

出　　处：《高电压技术》2003年第8期1-2,25,共3页High Voltage Engineering

基　　金：辽宁省自然科学基金资助 (编号 :2 0 0 110 2 0 2 9)

摘　　要：讨论了汽轮机故障诊断中的特征提取、数据处理和遗传神经网络结构。诊断系统的核心即神经网络分类系统可根据征兆区分不同故障。由于事先不知道故障征兆关系 ,所以诊断系统应有学习功能 ,基于遗传神经网络可满足这种要求。基于此提出的新故障诊断方法并应用汽轮发电机组故障诊断中 ,诊断结果与实际相符 。The main topics discussed in this paper include data extraction, data processing, structure of neural networks and genetic algorithm of the fault diagnosis for turbo-generator. A fault diagnosis system contains a classification system that can distinguish different fault based on observed symptoms of the process under investigation. Since the fault symptom is not always known beforehand, a system is required, which can be learned from experimental or simulated data. A genetic neural networks-based diagnosis is advantageous. In this paper, a new diagnosis scheme is presented and applied to a turbo-generator. Promising results are obtained which show that the presented method is possible approaches to perform effective fault diagnosis in turbo-generator.

关 键 词：汽轮发电机组 故障诊断 遗传算法 神经网络 数据处理 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]