基于H-mine算法的变电站二次系统故障关联分析

Fault Correlation Analysis of Substation Secondary System Based on H-mine Algorithm

在智能变电站二次系统发生故障时,故障数据存在冗余、丢失、甚至是错误数据等情况会导致二次系统故障诊断不便的问题。基于数据挖掘中的关联分析理论,提出了一种利用H-mine算法的关联分析方法。首先借助二次系统故障历史数据,利用H-mine算法实现了对变电站二次系统故障数据的关联分析,求出故障信息与具体的故障原因的关联规则。其次按照置信度指标选出符合要求的关联规则,然后引入了Kulc和IR两种指标对结果进行分析,确保结果的可靠性,实现故障特征信息与故障原因的快速匹配,以便检修人员制定检修方案。通过算例分析表明,与传统关联分析算法相比,H-mine算法处理速度较快,可以迅速地实现信息匹配,并且Kulc和IR两种指标有效地保证了关联分析结果的准确性。

In the secondary system of intelligent substation,redundancy,loss and even wrong data of fault data will cause inconvenience of the secondary system fault diagnosis.Based on the theory of association analysis in data mining,we proposed an association analysis method using H-mine algorithm.Firstly,with the help of historical data of secondary system fault,the association analysis of secondary system fault data in substation is realized by using H-mine algorithm,and the association rules between fault information and specific fault causes are obtained.Secondly,we selected the association rules meeting the requirements according to the confidence index.Then,we introduced Kulc and IR to analyze the results to ensure the reliability of the results and to realize the rapid matching between fault feature information and fault causes,so that the maintenance personnel can make maintenance plans.The example analysis shows that H-mine algorithm processes much faster than the traditional association analysis algorithm.So it can quickly achieve information matching,and Kulc and IR indicators effectively ensure the accuracy of the association analysis results.