电力变压器的动态隧道BP网络故障诊断算法

Algorithm of training BP Neural Network with dynamic tunneling technique for fault diagnosis of power transformers

变压器油中溶解气体分析是电力变压器绝缘故障诊断的重要方法。将人工神经网络中的BP算法应用于电力变压器故障诊断。由于BP算法训练神经网络具有训练易陷入局部极小,收敛速度缓慢的缺点,动态隧道技术运用到训练BP网络上,可以有效地改进BP网络易陷入局部极小的缺陷。经大量实例分析,并将其结果与传统的BP算法的结果进行比较,表明该算法能有效地对电力变压器单故障样本进行分类,具有较高的诊断准确率。

Dissolved Gas-in-oil Analysis （DGA） plays an important role in fault diagnosis of power transformers.BP （Back Propagation） algorithm is used to classify for insulation fault diagnosis in this paper.But typical BP algorithm has some defects,such as converging slowly and immersing in local vibration frequently.The algorithm using dynamic tunneling technique to train BP Neural Networks has been proved to have good performances in avoiding the local trap.So this paper adopts BP artificial neural network with dynamic tunneling technique in fault diagnosis of power transformers. A mass of fault samples are analyzed in the algorithm and the results are compared with those obtained by the traditional BPNN.The comparison result indicates that the algorithm using the dynamical tunneling technique has better classifying capability for single-fault samples as well as high diagnosis precision.