改进型概率神经网络在变压器故障诊断中的应用

Application of Improved Probabilistic Neural Network in Transformer Fault Diagnosis

为了提高变压器故障诊断的准确率,提升变压器故障预测的水平,提出一种改进型概率神经网络的变压器故障诊断方法。该方法以改良三比值法作为特征值,采用概率神经网络进行分类诊断。针对概率神经网络预测结果容易受到平滑因子的影响,利用灰狼搜索算法对概率神经网络中平滑因子参数进行优化,从而能够提高预测的准确度。实验结果表明,概率神经网络预测的准确率只有70%,改进型概率神经网络的准确率达到90%,准确率提高了20%。改进型概率神经网络的变压器故障诊断方法具有良好的性能,是一种行之有效的变压器故障诊断方法。

In order to improve the accuracy of transformer fault diagnosis and enhance the level of transformer fault prediction, an improved probabilistic neural network method for transformer fault diagnosis is proposed. In this method, the improved three-ratio method is used as the characteristic value, and probabilistic neural network is used for classification diagnosis. Because the prediction results of probabilistic neural network are easily affected by smoothing factor, grey wolf search algorithm is used to optimize the parameters of smoothing factor in probabilistic neural network, which can improve the accuracy of prediction. The experimental results show that the accuracy of the probabilistic neural network prediction is only 70%, the accuracy of the improved probabilistic neural network is up to 90%, and the accuracy is increased by 20%. The transformer fault diagnosis method based on improved probabilistic neural network has good performance and is an effective transformer fault diagnosis method.