电力计量自动化终端设备运行异常智能诊断方法  

Intelligent Diagnosis Method for Abnormal Operation of Power Measurement Automation Terminal Equipment

作　　者：姜雪娇 徐佳隆 钟磊 江蕾 吴民 孙延松 Jiang Xuejiao

机构地区：[1]海南电网有限责任公司,海南海口570010

出　　处：《工业控制计算机》2025年第4期111-113,共3页Industrial Control Computer

摘　　要：在长期运行过程中,由于内部元件老化、外部环境干扰等原因,电力计量自动化终端设备容易出现异常。而设备运行信号中含有的噪声分量对于异常诊断造成了严重干扰,为此研究过程中在对信号进行EMD分解后,结合样本熵特征筛选与小波阈值滤波能够有效去除信号中的噪声分量,提升信号质量,为后续的设备运行异常诊断奠定坚实的数据基础。为此提出了一种新的电力计量自动化终端设备运行异常智能诊断方法,采集终端设备运行信号,对于采集到的信号进行EMD分解,结合样本熵特征筛选及小波阈值滤波实现信号去噪处理。利用主成分分析法对于去噪后的信号进行降维处理后识别异常信号,结合KNN与随机森林模型确定异常信号类型,从而实现设备异常智能诊断。A new intelligent diagnosis method for abnormal operation of power metering automation terminal equipment is proposed.Collect operating signals from power metering automation terminal equipment,perform EMD decomposition on the collected signals,and combine sample entropy feature screening and wavelet threshold filtering to achieve signal denoising processing.Using principal component analysis to reduce the dimensionality of the denoised signal and identify abnormal signals,combined with KNN and random forest models to determine the type of abnormal signals,thus achieving intelligent diagnosis of equipment abnormalities.

关 键 词：电力计量 自动化终端设备 设备运行异常 智能诊断 

分 类 号：TM933[电气工程—电力电子与电力传动]