基于Q-Learning的电网工程数据处理与分析方法研究  

Research on data processing and analysis method of power grid engineering based on Q⁃Learning

作　　者：涂潜 占亚波 陈东亮 宗震 TU Qian;ZHAN Yabo;CHEN Dongliang;ZONG Zhen(Construction Branch,State Grid Anhui Electric Power Co.,Ltd.,Hefei 230071,China)

机构地区：[1]国网安徽省电力有限公司建设分公司,安徽合肥230071

出　　处：《电子设计工程》2023年第12期124-127,132,共5页Electronic Design Engineering

基　　金：2021年国网安徽省电力有限公司建设分公司咨询课题(B6120A210003)。

摘　　要：针对因电网工程数据管控智能化程度不足而导致的工程延期等现象频繁发生的问题,文中展开了一种基于Q-Learning的电网工程数据管控与分析方法研究。该方法在构建的电网工程数据管控与分析系统中,利用主成分分析法(Principal Component Analysis,PCA)提取数据的主要影响因素。将其输入Q-Learning算法进行智能分析,从而得到数据分析结果并用于制定管控措施。同时以某电力公司的输变电工程费用数据作为样本进行实验验证,结果表明,采用所提方法分析得到的工程总费用约为2千万元,相较预算减少了3%,且管控效果优于其他对比方法,能够为工程数据提供支撑与辅助决策。Aiming at the problem that the lack of intelligence of power grid engineering data management and control leads to frequent engineering delays,this paper studies a power grid engineering data management and control and analysis method based on Q⁃Learning.In the power grid engineering data control and analysis system,the Principal Component Analysis(Principal Component Analysis,PCA)method is used to extract the main influencing factors of the data,and input them into the Q⁃Learning algorithm for intelligent analysis,so as to obtain the data analysis results for formulating control measures.Taking the cost data of power transmission and transformation project of a power company as a sample for experimental verification,the results show that the total cost of the project analyzed by the method proposed in this paper is about 20 million yuan,which is 3%lower than the budget.The control effect is better than other comparison methods,and can provide engineering data support and auxiliary decision⁃making.

关 键 词：电网工程 数据管控 智能分析 主成分分析法 Q-Learning算法 输变电工程 

分 类 号：TP381[自动化与计算机技术—计算机系统结构] TN919[自动化与计算机技术—计算机科学与技术]