用于支持向量机算法的电流指纹采集系统设计  

Design of Current Fingerprint Acquisition System for SVM Algorithm

作　　者：李闯 LI Chuang(Shanghai Lingzhou Medical Technology Co.,Ltd.,Shanghai 201206,China)

机构地区：[1]上海领周医疗科技有限公司,上海201206

出　　处：《仪表技术》2023年第4期27-31,共5页Instrumentation Technology

摘　　要：为了智能监控电网入户处违禁电器的使用以保证用电安全,需要对入网负载的电压和电流特征进行采样,再运用支持向量机算法对采集到的数据进行负载识别。为了保证算法学习的有效性和负载识别的准确性,设计了一种有效的采样系统,兼顾安全性、可靠性和准确性,且具备产业化基础。对常用电器进行电压电流采样实验,并送入支持向量机算法系统中深度学习,结果表明,该采样系统在负载识别上具有较高的准确性和可靠性,有望拥有广阔的应用场景。In order to intelligently monitor the use of prohibited electrical appliances at the entrance of the power grid to ensure electricity safety,it is necessary to sample the voltage and current characteristics of the incoming load,and then use support vector machine algorithm to identify the collected data.In order to ensure the effectiveness of algorithm learning and the accuracy of load recognition,an effective sampling system has been designed,which takes into account safety,reliability,and accuracy,and has an industrial foundation.Voltage and current sampling experiments are made on commonly used electrical appliances,which are sent to the support vector machine algorithm system for deep learning.The results show that the sampling system has high accuracy and reliability in load recognition,and is expected to have broad application scenarios.

关 键 词：支持向量机算法 电流指纹采集 窗口滤波 电压频率转换 负载识别 差分电路 线性光耦 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TP181[自动化与计算机技术—计算机科学与技术]