基于改进PSO-FNN神经网络变压器故障检测研究  

Research of transformer fault diagnosis based on improved PSO-FNN neural network

作　　者：任东红 谢萍萍 邢兵锁[1] 李朝东 林鹏 REN Donghong;XIE Pingping;XING Bingsuo;LI Chaodong;LIN Peng(Tongling Polytechnic,Tongling 244000,China;Tongling Power Supply Company of State Grid,Tongling 244000,China)

机构地区：[1]铜陵职业技术学院,安徽铜陵244000 [2]国网铜陵供电公司,安徽铜陵244000

出　　处：《安徽水利水电职业技术学院学报》2023年第1期35-39,共5页Journal of Anhui Technical College of Water Resources and Hydroelectric Power

基　　金：安徽省教育教学研究项目(2020jyxm2049);安徽省重点科研项目(2022AH052754)。

摘　　要：针对电力变压器故障难以准确诊断的问题,提出了一种基于改进粒子群算法(PSO)的模糊神经网络(FNN)诊断模型。该模型运用模糊神经网络,同时结合变压器故障与变压器油中各气体成分之间的密切关系,确定了神经网络输入变量,同时在标准粒子群算法中引入遗传变异因子对模型进行训练,提高了训练精度。MATAB软件测试结果表明,模型预测精度较高,可进一步研究应用。Because the fault of power transformer is difficult to diagnose accurately,a fuzzy neural net-work diagnosis model based on improved particle swarm optimization(PSO)is proposed in the paper.The proposed model uses fuzzy neural network and the neural network input variables are determined based on the close relationship between transformer faults and gas components in transformer oil.At the same time an improved particle swarm optimization algorithm is proposed to train the neural net-work,genetic variation factor is introduced to avoid falling into local optimum and training accuracy is improved.The results show that the improved PSO-FNN transformer fault diagnosis model has high precision and certain application value.

关 键 词：变压器故障 改进粒子群算法 模糊神经网络 遗传变异 诊断 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]