CAJ | 학술논문

针对变压器故障诊断神经网络模型存在网络结构复杂、训练时间长等问题，提出基于粗糙集及RBF神经网络的变压器故障诊断方法。运用粗糙集理论中无决策分析，建立基于可分辨矩阵和信息熵的知识约简算法，进行数据挖掘，寻找最小约简；以处理后的数据集合作为训练样本，采用高斯函数作为径向基函数，分别求解方差及各层权值，建立变压器故障诊断模型。通过测试对比，此算法虽然略微降低诊断正确率，但网络结构简单、训练速度快、泛化能力强，对提高神经网络在变压器故障诊断中的应用性能有较好的指导意义。
침대변압기고장진단신경망락모형존재망락결구복잡、훈련시간장등문제，제출기우조조집급RBF신경망락적변압기고장진단방법。운용조조집이론중무결책분석，건립기우가분변구진화신식적적지식약간산법，진행수거알굴，심조최소약간；이처리후적수거집합작위훈련양본，채용고사함수작위경향기함수，분별구해방차급각층권치，건립변압기고장진단모형。통과측시대비，차산법수연략미강저진단정학솔，단망락결구간단、훈련속도쾌、범화능력강，대제고신경망락재변압기고장진단중적응용성능유교호적지도의의。
Aiming at the problems of complex network structure, long training time and other issues in classical neural networks, a method of RBF neural network combined with rough sets has been proposed in this paper for transformers fault diagnosis.Firstly, the minimum reduction is calculated after data mining based on distinguishes matrix and infor-mation entropy in rough set theory.Then, gaussian function has been taken as the radial basis function, and the vari-ance and weights of layers are calculated by the processed data collection as training samples .Finally,the transformer fault diagnosis model has been built.By comparing the test results, although the algorithm is slightly lower in the view of the diagnostic accuracy, the simple network structure, training speed and strong generalization ability of neural net -works to improve application performance in transformer fault diagnosis have better guidance.