基于支持向量机的多回路关口电能表在线监测方法研究

根据多回路关口电能表监测的多通道采样连续性，采样信号的多特征及其耦合性，提出一种基于支持向量机的多回路关口电能表在线监测方法，该方法分别设计了基于监测系统的信号采集单元、数据通信单元、数据处理单元及上位机显示单元，搭建了电能表的多个故障缺陷模型，通过信号采集单元将缺陷模型的三相电流及电压实时数据传送给数据处理单元，数据处理单元对收集的数据进行基础运算后形成对应的有功功率样本数据，并对其进行小波包能量谱离线分析后形成故障特征池，构建了基于支持向量机算法的多分类支持向量机模型并将得到的模型内嵌于上位机显示单元，监测系统运行时上位机显示单元直接基于模型的关联规则，与转换后的采样信息进行特征匹配，实现对电能表运行工况的实时监测。经实例验证结果表明该方法能够实时、准确的对多回路关口电能表进行在线监测，具有较强的应用价值及前景。

Research on on-line monitoring method of multi-loop gateway watt-hour meter based on SVM

According to the multi-channel sampling continuity of multi-loop gateway electric energy meter monitoring, the multi-features and coupling of sampled signals, an online monitoring method of multi-loop gateway electric energy meter based on support vector machine is proposed. The online monitoring method of the meter is designed based on The signal acquisition unit, data communication unit, data processing unit and host computer display unit of the monitoring system; the article first builds multiple fault defect models of the electric energy meter, and uses the signal acquisition unit composed of three-phase current and voltage sensors to divide the three fault models. The real-time data of phase current and voltage are transmitted to the data processing unit. The data processing unit performs basic calculations on the collected data to form the corresponding model active power sample data and uses the wavelet packet energy spectrum for offline analysis to form a fault feature pool. Based on this, construct A multi-class support vector machine model based on the support vector machine algorithm is built and the obtained model is embedded in the upper computer display unit; the upper computer display unit directly matches the converted sampling information based on the association rules of the model when the monitoring system is running. Realize the real-time monitoring of the operating conditions of the electric energy meter. The results show that the method can quickly and accurately monitor the multi-circuit gateway electric energy meter online, and it has strong application value and prospects.