基于神经网络的变压器故障诊断研究.rar

  • 需要金币1000 个金币
  • 资料包括:完整论文,开题报告,任务书
  • 转换比率:金钱 X 10=金币数量, 即1元=10金币
  • 论文格式:Word格式(*.doc)
  • 更新时间:2013-08-29
  • 论文字数:17767
  • 当前位置论文阅览室 > 毕业设计 > 电气工程 >
  • 课题来源:(小时代)提供原创文章

支付并下载

摘要:现代设备技术水平不断提高,生产率、自动化要求越来越高,相应地,故障也随之增加。变压器作为电力系统中非常复杂而且非常重要的设备,其工作状态对电力系统、企事业单位生产及居民生活具有十分重要的影响。如何提前对变压器故障进行预测和在故障发生后迅速判断故障原因是提高工作效率、减少经济损失的一个重要途径。因此研究变压器故障诊断对保证系统安全、可靠、经济运行,提高经济效益具有重要意义。

   概率神经网络(probabilistic neural networks)结构简单、训练简洁,利用概率神经网络模型的强大的非线性分类能力,将故障样本空间映射到故障模式空间中,可形成一个具有较强容错能力和结构自适应能力的诊断网络系统,从而提高故障诊断的准确率。本文在对油中溶解气体分析法进行深入分析后,以改良三比值法为基础,建立基于概率神经网络的故障诊断模型。然后,选取23组变压器故障原始样本数据对概率神经网络模型进行“学习”训练,获得了具有预测诊断功能的网络模型;选取10组变压器在线监测数据作为测试数据,并查看了训练数据网络的分类效果图和预测数据网络的分类效果图,结果只有两个样本判断错误,即只有两种变压器的故障类型判断错误,验证了基于概率神经网络在变压器故障预测诊断处理中的有效性。

关键词 故障诊断 概率神经网络 变压器 油中溶解气体分析

 

Abstract:With the technical level of modern facility improves continually, the fault probability increases greatly. Power transformer has a very significant influence to power system, enterprise s production and people s life. How to forecast transformer s fault ahead and find the fault reason quickly after the fault is a good way to increase work efficiency and lighten the economy losing.

   Probabilistic neural network has the advantages of simple structure, simple training, the use of a probabilistic neural network model for strong nonlinear classification, fault sample space is mapped to a fault in the pattern space, can form a strong fault tolerant ability and structure of adaptive diagnosis system, so as to improve the accuracy of fault diagnosis. Based on the dissolved gas in oil analysis in-depth analysis, in order to improve the ratio of three as the basis, establish the fault diagnosis based on probabilistic neural network model. Then, select 23 group of transformer fault original sample data on the probabilistic neural network model of" learning" training, obtain the predictive diagnosis of functional network model; select 10 group of transformer on-line monitoring data as test data, and show the training data network classification effect diagram and the predicted data network classification effect chart, only the results of a sample of two errors of judgment, that only two transformer fault type judgement error, verification based on probabilistic neural network in transformer fault forecast and diagnosis treatment effectiveness.

Keywords fault diagnosis, probability neural networks(PNN),power transformer,Dissolved Oas Analysis(DGA)