基于模糊决策树的配电线路故障诊断研究

Research on fault diagnosis of distribution lines based on fuzzy decision tree

配电系统中准确、快速的故障识别在运行中起着不可或缺的作用,当故障事件发生后,停电区域将扩展到相邻区域,因此,为了使电力系统恢复健康状态,准确及时的故障识别越来越重要。文章提出了一种基于模糊决策树(FDT)的电力系统故障诊断方法。FDT将模糊规则库引入传统的决策树,通过在终端节点结合每个参与类的概率的多类决策。FDT在离线模式下使用电力系统分析,针对不同的操作条件进行训练,并且在节点总线系统、14节点总线系统上验证了所开发方案的性能。

Accurate and rapid fault identification plays an indispensable role in the operation of power distribution systems.When a fault event occurs,the power outage area will expand to adjacent areas.In order to restore the health of the power system,accurate and timely fault identification is becoming increasingly important.This article proposes a power system fault diagnosis method based on fuzzy decision tree(FDT).FDT introduces a fuzzy rule library into traditional decision trees,and trains the developed method in offline mode using power system analysis for different operating conditions by combining the probability of each participating class with multi class decision-making at terminal nodes.The performance of the proposed method is verified on 9 nodes bus systems and 14 nodes bus systems.