摘要
为实现对电力计量装置电压异常的精准识别,开展基于随机森林的电力计量装置电压异常故障识别方法的设计研究。采样模块主要由电压取样、电流取样、低压取样、放大、A/D转换、通信等部分构成,根据实际需求,设计电力计量装置电压异常分量采样;引进随机森林,设计采样分量训练与分类;在设备的电压传送线和前端建立正常的通信连接,通过对电力计量装置电压状态的在线监测,实现对装置电压故障的识别。对比实验结果证明:设计的故障识别方法的实际应用效果良好,可以精准识别到电力计量装置的电压异常故障,且识别结果较为准确。
In order to realize accurate identification of voltage anomalies of power metering devices,a fault identification method for voltage anomalies of power metering devices based on random forest was designed and studied.The sampling module is mainly composed of voltage sampling,current sampling,low-voltage sampling,amplification,A/D conversion,communication,etc.According to the actual demand,power metering device voltage anomaly component sampling was designed.Random forest was introduced to design training and classification of sampling components.A normal communication connection was established between the voltage transmission line and the front end of the device,and the voltage fault of the device can be identified through the on-line monitoring of the voltage state of the power metering device.Compared with the experimental results,it is proved that the designed fault identification method has good practical application effect.The method can accurately identify the voltage abnormal fault of the power metering device,and the identification results are more accurate.
作者
殷毓灿
孔赟
YIN Yucan;KONG Yun(Yangzhou Power Supply Branch of State Grid Jiangsu Electric Power Co.,Ltd.,Yangzhou 225000,China)
出处
《通信电源技术》
2023年第5期88-90,共3页
Telecom Power Technology
关键词
随机森林
在线监测
故障识别
电力计量装置
random forest
on-line monitoring
fault identification
power metering device
