基于神经网络的输电线路故障检测研究  

Research of Fault Detection in Transmission Line Based on Neural Network

作　　者：董天祯[1] 郭江鸿[2] 吕娟[2] 付强 宝青兰 

机构地区：[1]哈尔滨工程大学模糊信息分析与智能识别研究室,哈尔滨150001 [2]哈尔滨工程大学计算机科学与技术学院,哈尔滨150001 [3]内蒙古通辽市奈曼旗农电局,内蒙古通辽028300 [4]内蒙古通辽市奈曼旗蒙古族中学,内蒙古通辽028300

出　　处：《系统仿真学报》2009年第15期4903-4906,4911,共5页Journal of System Simulation

摘　　要：在研究现有架空输电线路故障检测系统不足的基础上,提出了一种基于神经网络的输电线路故障检测系统。该系统以复杂结构的10KV架空输电线路为研究对象,利用分布在线路上的数据采集设备得到线路电气参数的相应数据并进行编码;通过载波通信将数据传送给控制中心;在控制中心对各参数数据解码和去噪;然后将每个参数进行分段数据采样输入经反向传播算法(即LMBP算法)训练的三层结构的神经网络进行分析,最后进行统计处理输出故障信息。采用EMTP仿真实验表明,本文故障检测系统准确率较其它故障检测系统有较大提高。On the basis of analysis of the existing cable fault-detecting systems, a cable fault detecting system based on neural net work was proposed; the 10kv aerial cable was investigated under this system. The data caught by the device located on the aerial cable, was sent to the control center by carrier wave communication, where data interpolation was used to restore the parameter＇s real distribution. After segment sampling to each parameter, the sample data were given to the 3-layer neural network, the fault type was given by the statistical analysis. The sample set was used to train the neural net work by LMBP algorithm. By EMTP simulating software, the simulating experiment was carried out. The results show that the accuracy rating increases greatly.

关 键 词：输电线路 故障检测 神经网络 小波分析 EMTP仿真 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]