基于弱关联挖掘技术的电网故障自动诊断研究

Research on power grid fault automatic diagnosis based on weak association mining technology

针对电网故障进行诊断的过程中,故障信息存在不完整或不确定性,甚至存在关键信息丢失的情况,造成故障诊断难以得出正确结论的问题,提出一种基于弱关联挖掘技术的电网故障自动诊断方法。首先进行支持度计算,得到电网故障的表述参数,并将电网故障类别看作是贝叶斯网络的父节点,将挖掘的弱关联规则作为子节点,构建基于弱关联挖掘的贝叶斯网络模型,对各父节点的先验概率及各子节点的条件概率进行计算,完成对电网故障的自动诊断。仿真实验结果表明,采用所提方法对电网故障进行自动诊断,正确性高,容错性好,实用性强,具有很高的诊断精度。

In the process of fault diagnosis of power grids,it is difficult to obtain the correct conclusion of the fault diagnosis due to the imperfection or uncertainty of fault information,even the key information loss. A power grid fault automatic diagnosis method based on weak association mining technology is put forward. The support degree calculation is executed first to obtain the expression parameters of power grid fault. The power grid fault category is regarded as the father node of Bayesian network,and weak association mining rule as the child node to construct the Bayesian network model based on weak association mining.After that the prior probability of each father node and conditional probability of each child node are calculated to complete the automatic diagnosis of power grid failure. The simulation results show that the proposed method can automatically diagnose the power grid fault,and has high accuracy,good fault-tolerance,strong practicability,and high diagnostic accuracy.