基于LSTM和ArcGIS的区域输电网污秽状态智能评估方法  

Intelligent Assessment Methods of Regional Transmission Networks Pollution State Based on LSTM and ArcGIS

作　　者：文思伦 张楚岩 刁明光[1] 周振安 徐惠勇[1] 刘慧芳[1] WEN Silun;ZHANG Chuyan;DIAO Mingguang;ZHOU Zhen′an;XU Huiyong;LIU Huifang(School of Information Engineering,China University of Geosciences(Beijing),Beijing 100083,China)

机构地区：[1]中国地质大学(北京)信息工程学院,北京100083

出　　处：《湖南电力》2024年第3期55-63,共9页Hunan Electric Power

基　　金：国家自然科学基金项目(51907178)。

摘　　要：为了提高电力系统的污闪预防水平,以北京市区域内架空输电线路绝缘子表面污秽状态的评估方法为研究对象,提出一种基于气象条件和盐密数据分布的区域污秽状态智能评估方法。首先,分析影响绝缘子表面积污的主要外部因素,并利用灰色关联分析法选取出6种关联度最大的影响因素。其次,建立一种基于长短期记忆神经网络的绝缘子污秽度预测模型,该模型的预测结果与实测数据吻合度较高,可以很好地反映电网绝缘子的污秽状态。最后,将气象条件与地理信息相结合,使用ArcGIS绘制北京地区的污区分布预测图,实现污秽度评估的动态表达,并在校园内搭建微型自然积污平台以验证文中方法。结果表明,所提方法可以实现区域输电网污秽度数据的周期动态监测,提前预判重度污区出现位置,有助于提升电网防污闪工作的预见性、精准性及智能化水平。In order to improve the pollution flashover prevention level of power system,this paper takes the assessment method of pollution state on insulator surface of overhead transmission lines in Beijing as the research object,and proposes a dynamic intelligent assessment method of regional pollution degree based on meteorological conditions and salt density data distribution.First,this paper analyzes the main external factors affecting insulator surface pollution,and uses grey correlation analysis to select six factors with the highest correlation degree.Secondly,a prediction model of insulator pollution degree based on long short-term memory neural network is established in this paper.The prediction results of this model are in good agreement with the measured data,which can well reflect the pollution state of insulators in power grid.Finally,by combining meteorological conditions with geographic information,this paper draws the forecast map of pollution area distribution in Beijing with ArcGIS,and realizes the dynamic expression of pollution degree evaluation.Moreover,a miniature natural fouling platform is built on campus to verify the proposed method.The results show that the research work in this paper can realize the periodic dynamic change monitoring of pollution degree data of regional transmission grid,predict the occurrence location of heavy pollution areas in advance,and help to improve the predictability,accuracy and intelligence level of anti-pollution flasticity work of power grid.

关 键 词：输电线路外绝缘 气象条件 灰色关联分析法 长短期记忆神经网络(LSTM) 污秽度预测模型 ARCGIS 

分 类 号：TM726[电气工程—电力系统及自动化]