多传感器信息融合的低压电网窃电设备在线监测方法  

On Line Monitoring Method of Electric Stealing Equipment in Low Voltage Power Network Based on Multi Sensor Information Fusion

作　　者：贾永良 陶鹏 张洋瑞 张冰玉 JIA Yong-liang;TAO Peng;ZHANG Yang-rui;ZHANG Bing-yu(Marketing Service Center of State Grid Hebei Electric Power Co.,Ltd.,Hebei Shijiazhuang 050000,China)

机构地区：[1]国网河北省电力有限公司营销服务中心,河北石家庄050000

出　　处：《机械设计与制造》2024年第11期105-109,共5页Machinery Design & Manufacture

基　　金：国家电网有限公司总部科技项目(5100-202155492A-0-5-ZN)。

摘　　要：低压电网窃电设备的在线监测是电网安全工作过程中不可缺少的步骤,但监测过程易受噪声信号、电压强度、磁场干扰等问题的影响,为此提出基于多传感器信息融合的低压电网窃电设备在线监测方法。该方法利用多传感器采集窃电设备的运行信号并将信号融合,再通过局域波分解算法剔除信号中的噪声,避免噪声对监测过程产生影响,其次采用原子分解算法提取信号的特征,最后将信号特征输入到正弦基神经网络模型中,通过对信号的自动分类与监测完成低压电网窃电设备的在线监测。实验结果表明,所提方法的信号监测效果好、响应时间短。The online monitoring of stealing equipment in low-voltage power grid is an indispensable step in the process of power grid security,but the monitoring process is vulnerable to noise signals,voltage intensity,magnetic field interference and other issues.Therefore,an online monitoring method for power stealing equipment in low-voltage power grid based on multi-sensor information fusion is proposed.This method uses multi-sensor to collect the operating signals of power stealing equipment and fuse them,then uses local wave decomposition algorithm to eliminate the noise in the signal to avoid the impact of noise on the monitoring process,secondly uses atomic decomposition algorithm to extract the characteristics of the signal,and finally inputs the signal characteristics into the sinusoidal basis neural network model,and completes the online monitoring of power stealing equipment in low-voltage power grid through the automatic classification and monitoring of signals.The experimental results show that the proposed method has good signal monitoring effect and short response time.

关 键 词：多传感器信号采集 原子分解法 正弦基神经网络 误差函数 自动分类与监测 

分 类 号：TH16[机械工程—机械制造及自动化] TM93[电气工程—电力电子与电力传动]