基于机器学习的配电网监控信息智能分析规则库构建方法

Rule Base Construction Method for Intelligent Analysis of Distribution Network Monitoring Information Based on Machine Learning

为构建一种具有实时性的配电网监控信息智能分析规则库,提出了基于机器学习的配电网监控信息智能分析规则库构建方法。将规则库中全部配电网监控规则头排序并设成主链,将规则导进链表里生成规则集,保证各个监控信息数据包都存在一个分析规则。使用基于机器学习的配电网故障数据分类方法,识别配电网监控信息中的故障数据,并提取故障数据频繁项集。使用基于MapReduce的并行关联规则增量更新算法,更新分析规则库中的信息智能分析规则,保证分析规则库中的信息智能分析规则具有实时性。实验结果表明,所提方法的识别结果准确度、检出率均值都大于0.97,假阳性率都是0.01,可以及时识别出配电网监控系统实时检测故障信息,保证信息智能分析规则更新具有实时性。

In order to build a real-time intelligent analysis rule base of distribution network monitoring information, a construction method of intelligent analysis rule base of distribution network monitoring information based on machine learning is proposed. All the monitoring rule heads in the rule base are sorted and set as the main chain, and the rules are imported into the linked list to generate the rule set, so as to ensure that each monitoring information packet has an analysis rule;the machine learning based fault data classification method of distribution network is used to identify the fault data in the monitoring information of distribution network, and extract the frequent itemsets of fault data. The parallel incremental updating algorithm of association rules based on MapReduce is used to update the information intelligent analysis rules in the analysis rule base, so as to ensure the real-time performance of the information intelligent analysis rules in the analysis rule base. The experimental results show that the accuracy and the average detection rate of the proposed method are greater than 0.97, and the false positive rate is 0.01. The proposed method can identify the real-time fault information of the distribution network monitoring system in time, and ensure the real-time updating of information intelligent analysis rules.