基于粒子群优化神经网络的变压器故障诊断方法

Transformer Fault Diagnosis Method Based on Particle Swarm Optimization Neural Network

针对传统变压器故障诊断方法存在处理效率低、诊断结果主观性强的问题,提出一种改进的粒子群(PSO)-BP神经网络变压器故障诊断方法,对智能变压器故障诊断方法展开研究。首先,采用无编码比值法,选用变压器油中各类溶解气体的比值来更准确地区分故障类型;其次,根据所选的测试集气体比值信息,利用灰色关联度法选出与测试集信息特征相似度高的信息作为训练集,剔除冗余信息,优化样本数据;最后,构建PSO-BP变压器故障诊断模型,利用改进的PSO算法优化BP神经网络中的权值和阈值,使诊断模型拥有更快的诊断速度和更高的准确率。使用采集的变压器油中溶解气体信息训练诊断网络,与传统BP神经网络法进行对比。结果表明,所提方法对变压器故障诊断的准确率明显提高。

Aiming at the problems of low processing efficiency and strong subjectivity in diagnosis results of traditional transformer fault diagnosis methods,an improved particle swarm optimization BP neural network transformer fault diagnosis method is proposed,and research is conducted on intelligent transformer fault diagnosis methods.Firstly,using the non coding ratio method,the ratio of various dissolved gases in transformer oil is selected to more accurately distinguish the type of fault.Secondly,based on the gas ratio information of the selected test set,the grey correlation method is used to select information with high similarity to the information features of the test set as the training set,eliminate redundant information,and optimize the sample data.Finally,a PSO-BP transformer fault diagnosis model is constructed,and an improved particle swarm optimization algorithm is used to optimize the weights and thresholds in the BP neural network,enabling the diagnostic model to have faster diagnostic speed and higher accuracy.Train a diagnostic network using dissolved gas information collected from transformer oil,and compare it with the traditional BP neural network method.The results show that the proposed method has significantly improved the accuracy of transformer fault diagnosis.