基于改进关联规则挖掘的电网气象风险评估  

Meteorological Risk Assessment of Power Grid Based on Improved Association Rule Mining

作　　者：吴子龙 王正宇 WU Zilong;WANG Zhengyu(School of Electric Power,South China University of Technology,Guangzhou 510641)

机构地区：[1]华南理工大学电力学院,广州510641

出　　处：《舰船电子工程》2024年第2期143-147,共5页Ship Electronic Engineering

基　　金：广东省新能源电力系统智能运行与控制企业重点实验室开放基金项目(编号:GPKLIOCNEPS-2021-KF-01);南方电网公司重点科技项目“基于数据深度分析的配电网精准规划投资决策技术研究”(编号:YNKJXM20191215)资助。

摘　　要：为了给电力气象灾害预报预警提供依据,论文提出了一种基于改进关联规则挖掘的电网气象风险量化评估方法。首先,分析电网历史停电事件中气象要素与停电情况的数据信息构成,构建多维电网故障数据立方体。在此基础上,为了量化评估电网气象灾害风险,提出风险系数计算方法以改进多维关联规则挖掘算法,同时,优化关联规则生成和筛选环节,以提高数据挖掘效率和关联规则的可信度。然后,基于数据挖掘所得关联规则库,对电网气象灾害风险进行量化评估。最后,以某地区电网数据进行分析,验证了论文方法的有效性。In order to provide a basis for forecasting and early warning of power meteorological disasters,a quantitative evalua⁃tion method of power grid meteorological disaster risk based on improved multi-dimensional association rule mining is proposed.Firstly,the data information composition of meteorological elements and power outages in historical power outage events is ana⁃lyzed,and a multi-dimensional power grid fault data cube is constructed.On this basis,in order to quantitatively evaluate the mete⁃orological disaster risk of power grid,a risk coefficient calculation method is proposed to improve the multi-dimensional association rule mining algorithm.At the same time,the generation and screening of association rules are optimized to improve the efficiency of data mining and the credibility of association rules.Then,based on the association rule base obtained from data mining,the meteo⁃rological disaster risk of power grid is quantitatively evaluated.Finally,the effectiveness of the proposed method is verified by ana⁃lyzing the data of a regional power grid.

关 键 词：数据立方体 数据挖掘 多维关联规则 气象灾害 风险评估 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] TM76[自动化与计算机技术—计算机科学与技术]