摘要
为了助力调控员处理电网故障,推动智能电网高效运作,针对主电网故障数据难以聚类,能耗控制效果较差的问题,搭建基于数据挖掘的智能电网故障处置辅助决策系统。利用温度、位移、电压传感器等采集主电网运行数据,传输至信息处理模块识别过滤信息,滤除冗余信息保留告警信息,并汇总、分类处理告警信息,处理后传输至告警信息展示模块中,集中展示;故障处置辅助决策模块接收告警信息,同时通过故障检测子模块的K-means聚类算法及卷积神经网络识别故障类型,通过辅助决策子模块输出相应处置预案信号,在线离线分析模块接收信号后,实时保存故障视频,储存至可视化故障回放子模块,并针对同类型故障自动生成故障预案,供后续自动判别处置方案,传输至用户模块,方便调控员实时调阅进行报表智能查询。实验结果证明,该系统能够实时检测主电网故障数据并做出相应辅助决策;能耗控制方面表现突出,聚类及数据挖掘效果优秀。
In order to help regulators deal with power grid faults and promote the efficient operation of smart grid,aiming at the problems that the fault data of the main grid is difficult to cluster and the effect of energy consumption control is poor,a smart grid fault disposal auxiliary decision-making system based on data mining is built.The temperature,displacement and voltage sensors are used to collect the operation data of the main power grid and transmit it to the information processing module to identify and filter the information,filter the redundant information,retain the alarm information,summarize and classify the alarm information,and then transmit it to the alarm information display module for centralized display.The auxiliary decision-making module for fault disposal receives the alarm information,identifies the fault type through the K-means clustering algorithm and convolution neural network of the fault detection sub module,outputs the corresponding disposal plan signal through the auxiliary decision-making sub module,saves the fault video in real time after the online and offline analysis module receives the signal,stores it in the visual fault playback sub module,and automatically generates the fault plan for the same type of fault,It is used for subsequent automatic discrimination and disposal scheme,and transmitted to the user module to facilitate the real-time access of the regulator for intelligent query of the report.The experimental results show that the system can detect the fault data of the main power grid in real-time and make corresponding auxiliary decisions.Moreover,it has outstanding performance in energy consumption control and excellent clustering and data mining results.
作者
陶文伟
吴金宇
江泽铭
曹扬
仇伟杰
TAO Wenwei;WU Jinyu;JIANG Zeming;CAO Yang;QIU Weijie(China Southern Power Grid Company Limited,Guangzhou 510530 China)
出处
《测试技术学报》
2023年第2期127-134,共8页
Journal of Test and Measurement Technology
关键词
数据挖掘
智能电网
故障处置
辅助决策
聚类算法
卷积神经网络
data mining
smart grid
fault handling
decision support
clustering algorithm
convolutional neural network
