基于相关系数矩阵和概率神经网络的局部放电模式识别

针对变压器局部放电模式分类过程中特征参数维数过高的问题，提出了一种基于相关系数矩阵的参数降维方法。利用提取出的变压器局部放电信号的特征参数构造相关系数矩阵，通过分析放电信号18个特征参数间的相关性，删除具有相似分类能力的特征参数，之后引入分离度指标来衡量特征向量的分类能力大小，提取出6个具有较高分类能力的特征向量，最后通过概率神经网络进行模式识别。结果表明该降维方法有效降低了特征参数的维数，简化了分类器结构，在小样本情况下对于概率神经网络模式分类器具有较高的识别率，识别效果优于传统BP神经网络。

Pattern recognition for partial discharge based on correlation coefficient matrix and probabilistic neural networks

A new dimension reduction method based on correlation coefficient matrix is proposed aimed at the high-dimension of characteristic parameters in the process of pattern recognition for partial discharge in power transformer. The correlation coefficient matrix (CCM) is constructed using parameters extracted from partial discharge signal in power transformer. The parameters which have similar classification ability to each other are deleted with the help of correlation analysis among 18 characteristic parameters in CCM. Six parameters which have higher classification capabilities are extracted using the critical index and are used as the inputs for pattern classifiers of probabilistic neural networks. The results show that the parameter dimension is reduced and the classifier construction is simplified, and the recognition effect is better than that of the traditional back propagation neural network in the condition of small samples.