基于BWO-ELM算法与VR-GIS技术的电力光缆故障诊断及定位研究  

Research on Fault Diagnosis and Location of Power Optical Cable Based on BWO-ELM Algorithm and VR-GIS Technology

作　　者：蔡海良 胡凯 李军 邢小雷 CAI Hailiang;HU Kai;LI Jun;XING Xiaolei(Deqing County Power Supply Company of State Grid Zhejiang Electric Power Co.,Ltd.,Deqing 313200,China;Zhejiang Huayun Information Technology Co.,Ltd.,Hangzhou 310008,China;Deqing Xindian Electric Power Construction Co.,Ltd.,Deqing 313200,China)

机构地区：[1]国网浙江省电力有限公司德清县供电公司,浙江德清313200 [2]浙江华云信息科技有限公司,杭州310008 [3]德清欣电电力建设有限公司,浙江德清313200

出　　处：《计算机测量与控制》2022年第12期98-104,111,共8页Computer Measurement &Control

基　　金：国家电网重点科技项目(CY841000JS20210264)。

摘　　要：针对目前电力光缆故障模式识别精度低和故障点定位误差大的问题,提出一种基于BWO-ELM算法与VR-GIS系统的电力光缆故障诊断及定位方法,首先利用白鲸优化算法(BWO)优化极限学习机(ELM)初始参数,构建BWO-ELM多分类OTDR曲线分析的故障模式识别方法,获取光纤故障点的直线距离与类型,为故障定位奠定基础;其次,提出基于VR-GIS的光缆故障精确定位方法将光纤故障点的直线距离转换为光缆距离,并与实际地理位置匹配,得到实际故障点的坐标,同时直观展示于VR-GIS系统;最后通过仿真实验来验证所提方法的应用效果,结果表明所提方法故障模式识别精度约为98.66%,故障定位误差在±3 m上下浮动,平均误差为1.481%,较其他识别模型和故障定位方法具有较高的性能与准确率。Aiming at the problems of low accuracy of power cable fault pattern recognition and large error of fault location,a method of power cable fault diagnosis and location based on BWO-ELM algorithm and VR-GIS system is proposed.First,the BWO(beluga whale optimization)algorithm is used to optimize the initial parameters of the extreme learning machine(ELM),and the BWO-ELM multi classification OTDR curve analysis fault pattern recognition method is constructed,Obtain the straight-line distance and type of the optical fiber fault point to lay the foundation for fault location;Secondly,the accurate positioning method of optical cable fault based on VR-GIS is proposed,which converts the straight-line distance of the optical fiber fault point into the optical cable distance,and matches it with the actual geographical location to obtain the coordinates of the actual fault point,which is intuitively displayed in the VR-GIS system;Finally,the application effect of the proposed method is verified by simulation experiments.The results show that the fault pattern recognition accuracy of the proposed method is about 98.66%,the fault location error is floating up and down 3 M,and the average error is 1.481%.It has higher performance and accuracy than other recognition models and fault location methods.

关 键 词：电力光缆 故障模式识别 故障定位 BWO-ELM VR-GIS 

分 类 号：TN911[电子电信—通信与信息系统]