灰色关联分析下数字化电力网络过负荷数据挖掘

Data mining of digital power network overload data under gray correlation analysis

由于现有的挖掘方法挖掘误码率大于0.2 BER,挖掘精准程度低。文章为此研究了灰色关联分析下数字化电力网络过负荷数据挖掘,运用灰色关联分析法进行数据处理,挖掘出各因素之间的关联程度;对数据进行聚类分析,获得不同样本类型;运用训练集建立决策树模型,对产生数据进行差异度计算;运用数字化电力网络过负荷数据的挖掘判别函数进行数据挖掘。试验结果表明,在不同时间点中的数据序列幅值相同;10个小组过负荷数据的挖掘误码率为0~0.2 BER,能够对过负荷数据较为精准地挖掘,达到良好的挖掘效果。

Because the existing bit error rate of mining methods is greater than 0.2 BER and the mining accuracy is low,we study the overload data mining of digital power network under gray correlation analysis.The gray correlation analysis method is used to process the data and dig out the degree of correlation between various factors.The data were subjected to the clustering analysis to obtain the different sample types.The training set is used to build a decision tree model and calculate the difference degree of the generated data.Using the mining discrimination function of digital power network overload data for data mining.The experimental results show that the data sequence in different time points;the mining error rate of overload data in 10 groups is 0~0.2 BER,which can mine overload data accurately and achieve good mining effect.