基于数据挖掘的电网故障关联性分析与研究

Power System Fault Correlativity Analysis and Research Based on Data Mining

将数据挖掘技术中的关联规则应用到电网故障分析中,从故障数据中发现分类属性与决策属性间的频繁模式、相关性或因果关系,以便从宏观上把握电网故障元素间的关联特性.讨论了关联规则应用于电网故障分析的体系结构及实现的具体步骤,重点对电网故障关联性分析中的频繁项挖掘算法进行了研究,对传统的Apriori算法进行了改进,提出了一种高效的基于数组的类频繁项集挖掘算法.

Association rules of data mining technology were applied to fault analysis. The frequent pattern, the correlation or causal relationship between classification properties and decision-making properties were founded from fault data, the correlation property of power system fault elements was grasped on macro view. The fault analysis architecture by using association rules and realization steps were discussed in this paper, focusing on the frequent itemset mining algorithm research of power system fault correlativity analysis. The conventional algorithm Apriori was improved, and a highly efficient algorithm called class frequent itemset mining based on array was proposed.