基于稳态特征的输电设备全维度状态监测系统  

A Full-Dimensional Status Monitoring System for Transmission Equipment Based on Steady-State Characteristics

作　　者：刘军 杨治田[3] 李旭 罗彦宏 付纪华 赵新志 马晓光 LIU Jun;YANG Zhitian;LI Xu;LUO Yanhong;FU Jihua;ZHAO Xinzhi;MA Xiaoguang(State Grid Shaanxi Maintenance Company,Xi’an 710065,Shaanxi,China;Xi’an University of Technology,Xi’an 710048,Shaanxi,China;State Grid Shaanxi Electric Power Company,Xi’an 710048,Shaanxi,China;State Grid Shaanxi Electric Power Economic Technology Research Institute,Xi’an 710065,Shaanxi,China)

机构地区：[1]国网陕西省电力公司检修公司,陕西西安710065 [2]西安理工大学,陕西西安710048 [3]国网陕西省电力公司,陕西西安710048 [4]国网陕西经济技术研究院,陕西西安710065

出　　处：《电网与清洁能源》2020年第7期24-29,共6页Power System and Clean Energy

基　　金：国家自然科学基金(51779206)。

摘　　要：研究了一种基于稳态特征的输电设备全维度状态监测系统,利用监测装置管理模块、监测状态量模块、运行监控模块、状态预警模块,实现了输电设备的全维度状态监测。对于该系统采集到的设备状态数据,基于稳态特征利用SVM和SMOTE算法进行了数据挖掘和预处理,包括数据清理、集成、变换、规约以及样本平衡处理,滤除其中的不一致、不确定、缺失数据,有效降低了数据冗余率,提高了监测过程负载均衡性。实验结果表明:该系统的负载均衡性能较好,数据冗余率与监测数据传输损耗量更低,微小故障发现率与设备维稳率更高。In this paper,a full-dimensional status monitoring system for transmission equipment based on steadystate characteristics is studied,and the full-dimensional status monitoring of transmission equipment is realized by using monitoring device management module,monitoring state quantity module,operation monitoring module and state warning module.The device state data collected by the system is mined and preprocessed based on steady-state features using SVM and SMOTE algorithms,including data cleaning,integration,transformation,protocol,and sample balancing processing to filter out the inconsistent,uncertain,missing,redundant data,effectively reduce the data redundancy rate and improve the load balancing of the monitoring process.The experimental results show that the load balancing performance of the system is better,the data redundancy rate and the monitoring data transmission loss are lower,the detection rate of minor faults and the maintenance rate of the equipment are higher.

关 键 词：稳态特征 状态监测 预处理 数据冗余 负载均衡 

分 类 号：TM85[电气工程—高电压与绝缘技术]