摘要
针对电网结构日益复杂,故障处置预案解析、识别及智能推送手段匮乏导致故障处置预案利用率低的问题,文中提出一种基于自然语言理解的故障处置预案语义建模方法。首先,基于文本卷积神经网络训练包含通知、开停机、调整机组出力、控制潮流、控制电压、控制频率6种意图的预案文本,构建预案语义理解模型。然后,基于正则表达式对意图识别后的预案文本进行解析,得到预案中的设备对象和操作目标。最后,基于文本相似度技术映射预案中设备对象与调度控制系统中标准名称。在实际调控系统的应用表明,所提预案语义建模方法能够准确地解析预案、识别语义,并通过感知电网故障智能推送相应的故障处置预案。
Aiming at the problem of low utilization of fault handling plan due to the increasingly complex power grid structure and the lack of analysis,identification and intelligent push means of fault handling plan,a semantic modeling method of fault handling plan based on natural language understanding is proposed in this paper.Firstly,based on the text convolution neural network,the plan text with six intentions of notification,start and stop,adjusting unit output,controlling power flow,controlling voltage and controlling frequency is trained,and the semantic understanding model of the plan is constructed.Then,the plan text after intention recognition is analyzed based on regular expression to obtain the equipment object and operation target in the plan.Finally,based on the text similarity technology,the equipment object in the plan is mapped with the standard name in the dispatching control system.The application in the actual dispatching control system shows that the proposed semantic modeling method can accurately analyze and identify the semantics of the plan,and intelligently push the corresponding fault handling plan by sensing the power grid fault.
作者
胡怀伟
富英
张越
刘立轩
张沛
HU Huaiwei;FU Ying;ZHANG Yue;LIU Lixuan;ZHANG Pei(Inner Mongolia Power(Group)Co.,Ltd.,Ulanqab Electric Power Bureau,Ulanqab 012000,China;State Grid Electric Power Research Institute Co.,Ltd.,NARI Group Corporation Co.,Ltd.,Nanjing 211106,China;Beijing Kedong Electric Power Control System Co.,Ltd.,Beijing 100192,China;School of Electrical Engineering,Beijing Jiaotong University,Beijing 100044,China)
出处
《电力信息与通信技术》
2022年第5期68-73,共6页
Electric Power Information and Communication Technology
基金
内蒙古电力(集团)有限责任公司科技项目“基于大数据的监控信号统计分析技术研究及应用子项目基于地调的智能“调控员”的研究与实现”(2020-43)。
关键词
故障处置预案
文本卷积神经网络
语义理解
文本相似度计算
fault handling plan
text convolutional neural network
semantic understanding
text similarity calculation
