基于贝叶斯的非侵入式负荷监测冲击波形分类方法  

A Bayesian-based classification method of impulse waveforms for non-intrusive load monitoring application

作　　者：张博 梁凯 ZHANG Bo;LIANG Kai(Xuzhou Power Supply Company of State Grid Jiangsu Electric Power Co.,Ltd,Xuzhou,Jiangsu 221006;Wuxi Power Supply Company of State Grid Jiangsu Electric Power Co.,Ltd,Wuxi,Jiangsu 214002)

机构地区：[1]国网江苏省电力有限公司徐州供电分公司,江苏徐州221006 [2]国网江苏省电力有限公司无锡供电分公司,江苏无锡214002

出　　处：《电气技术》2022年第7期64-68,共5页Electrical Engineering

摘　　要：针对非侵入式负荷监测中冲击类负荷类型辨识问题,提出基于贝叶斯分类器的冲击波形分类方法。首先,定义电流冲击波形的冲击幅值、上升时间、跌落幅值和跌落时间四个特征量,用于建立多特征贝叶斯分类模型。然后,将采集的不同电器的电流冲击样本分为多组,用多组样本特征值的均值作为贝叶斯分类模型的参数。最后,在单相电能表硬件平台上实现分类算法,在实验室场景下对定频空调和变频空调进行分类测试。结果表明,本文所提方法可有效识别两种类型冲击,验证了所提方法的可行性。Aiming at the problem of impact load type identification in non-intrusive load monitoring,an impact waveform classification method based on Bayesian classifier is proposed.Firstly,the characteristic parameters of an impulse current waveform such as impulse amplitude,rising time,dropping amplitude,and falling time are defined to establish a multi-feature Bayesian classification model.Secondly,the current impulse waveform samples of different appliances are divided into several groups,and the mean values of the characteristic values of the samples are used as the parameters of the classification model.Finally,the classification algorithm is implemented on the hardware platform of single-phase watt hour meter.The classification tests of fixed frequency air conditioner and variable frequency air conditioner are carried out in the laboratory scene.The results show that the proposed method can effectively identify two types of impulse waveforms,which verifies the feasibility of the proposed method.

关 键 词：非侵入式负荷监测 贝叶斯分类器 负荷冲击电流 智能电能表 

分 类 号：TM714[电气工程—电力系统及自动化]