基于自组织特征映射的配电线路巡视周期分类研究  

Study on the Classification of Distribution Line Patrol Cycle Based on Self-organizing Feature Maps

作　　者：张宇泽 何亚樵 吴杰[1] 付学玲 ZHANG Yuze;HE Yaqiao;WU Jie;FU Xueling(State Grid Tianjin Electric Power Company,Chengxi Power Supply Company,Tianjin 300190,China)

机构地区：[1]国网天津市电力公司城西供电分公司,天津300190

出　　处：《电工技术》2020年第15期162-164,共3页Electric Engineering

摘　　要：针对差异化制定配电线路巡视周期问题,首先分析了配电线路巡视周期的影响因素,包括线路重要程度、平均负荷大小、故障发生频次、处缺频次和运行环境五个方面,然后提出了基于自组织特征映射的配电线路巡视周期分类方法,利用机器学习实现配电线路巡视周期的合理分类,为供电企业提高线路巡视效率提供技术支撑。Aiming at the problem of differentiated distribution line inspection cycle,the influencing factors of distribution line inspection cycle were analyzed,including five aspects of line importance,average load size,frequency of fault occurrence,frequency of shortage and operating environment.A self-organizing feature mapping-based classification method for distribution line inspection cycles was proposed.The machine learning method was used to achieve a reasonable classification of distribution line inspection cycles,providing technical support for power supply companies to improve line inspection efficiency.

关 键 词：配电线路 巡视周期 自组织特征映射网络 聚类 

分 类 号：TM726[电气工程—电力系统及自动化]