吴珊珊 张聪 赵欢欢 杨采薇

摘  要：电力在对人们生活的影响日益加深的同时，也带来了更大的火灾隐患。相关数据统计显示，电气火灾的第一大成因就是故障电弧。为提高串联型故障电弧检测的准确度，针对串联故障电弧的隐蔽性、随机性与不稳定性，文章提出基于谐波与SVM的串联故障电弧检测方法。通过自主搭建串联故障电弧实验平台以采集正常运行电路与产生串联故障电弧时的谐波数据，并以此为参数作为SVM的输入变量建立模型。研究结果表明，基于SVM的串联故障电弧检测方法，识别串联故障电弧的准确率较高，具有一定的可行性与有效性。

关键词：串联故障电弧;谐波;SVM

中图分类号：TP181;TM501               文献标识码：A文章编号：2096-4706（2021）15-0051-04

Abstract： While the influence of electric power on people’s life is deepening， it also brings greater fire hazards. Relevant statistics show that the first major cause of electrical fire is fault arc. In order to improve the accuracy of series fault arc detection， aiming at the concealment， randomness and instability of series fault electric arc， a series fault electric arc detection method based on harmonic and SVM is proposed in this paper. The series fault electric arc experimental platform is built independently to collect the harmonic data of normal operation circuit and series fault electric arc， and take this as the parameter and take this parameter as the input variable of SVM to establish the model. The results show that the series fault electric arc detection method based on SVM has high accuracy of identifying series fault electric arc， which has a certain feasibility and effectiveness.

Keywords： series fault electric arc; harmonics; SVM

0  引  言

我国应急管理部消防救援局公布的2010年1月至10月全国火灾形势报告数据显示，电气火灾占火灾总数的32.1%，是火灾的第一大成因，而电弧故障所引起的火灾事故在电气火灾中的占比高达67%。并联与接地型故障电弧因其电流明显变大，传统电气保护设备能够灵敏地将其切断。而串联型故障电弧可认为是在回路中串入了一个非线性电阻而使回路电流减小，小于传统电气保护设备设定的短路和过载阈值，所以普通的断路器不能有效地预防串联型故障电弧，致使串联型故障电弧成为引发电气火灾的主要元凶之一。故而，对串联型故障电弧进行研究具有十分重要的意义。

目前国内外建筑配电系统中故障电弧的……