改进的贝叶斯算法在智能变电站网络故障诊断系统中的研究

提高智能变电站网络故障诊断能力对于确保电力系统的稳定运行和供电可靠性具有重要意义,故障的分类是目前智能变电站网络故障诊断系统所面临的一个主要问题.常用分类算法存在着训练数据多样化,特征的选择标准具有不确定性,学习潜力匮乏等问题,文章在原有贝叶斯算法的基础上加入了特征的选择标准及学习过程,实验结果表明改进的贝叶斯算法在很大程度上能有效解决故障分类问题,从而提高智能变电站的网络故障诊断能力.

An Improved Bayesian Algorithm in the Intelligent Substation Network Fault Diagnosis System

To improve the fault diagnosis ability of the intelligent substation is very important for stable operation of power system and power supply reliability.One of the major problems that the power network fault diagnosis system faced is the fault classification problem.The existing classification algorithm has some shortcomings,such as the training samples imbalance,lack of consistent characteristics,weakness learning ability.The feature selection and learning strategies are added to Bayesian algorithm to propose an improved Bayesian fault classification algorithm.The experimental results show that the improved Bayesian algorithm can solve the fault classification problem effectively,and enhance the fault diagnosis capability of intelligent substation network.