摘要
本文分析了根据油中溶解气体、采用人工神经网络进行变压器的故障诊断时，不同的输入矢量构成方式对诊断结果的影响。介绍了主分量分析法的原理和步骤，由此构成的新矢量，降低了输入矢量的维数，提高了变量间的不相关性。实例检验结果表明，该方法是有效的。
The influence of different input manners on diagnosis results is discussed in this paper as the artificial neural network is adopted to diagnose the dissolved gases of transformer.The principle and procedure of principal component analysis method are introduced also.Comparing new input vector constructed by this method with that constructed directly,the dimension of input vector is decreased and capacity of orthogonal among the variables is enhanced.The practical examples show that this method is effective.
出处
《电工电能新技术》
CSCD
1999年第2期44-48,共5页
Advanced Technology of Electrical Engineering and Energy
基金
国家自然科学基金
关键词
故障诊断
主分量分析法
电力变压器
神经网络
fault diagnosis,dissolved gas analysis,principal component analysis,artificial neural network
