基于PSO-SVM的变压器故障诊断研究

Transformer Fault Diagnosis Based on PSO-SVM

支持向量机用于变压器故障诊断时,其参数的选择会影响到诊断的准确度。为了提高支持向量机的精确度和效率,将粒子群算法和支持向量机相结合,提出了基于粒子群优化支持向量机的故障诊断方法。用粒子群算法实现对支持向量机惩罚因子及径向基核函数的寻优,从而提高支持向量机的分类性能。仿真结果表明,此方法能够有效提高变压器故障诊断的准确率。

When Support Vector Machine（SVM） is used in fault diagnosis of transformer ,the parameter selection is very important and decides the fault diagnosis accuracy and efficiency .In order to enhance fault diagnosis precision ,a new fault diagnosis method is proposed by combining Particle Swarm Optimization （PSO） and SVM in this paper ,optimizing SVM based on PSO algorithm .PSO is used to optimize the penalty factor and the radial basis function ,to improve the classifying performance of SVM .Simulation results indicate that this method can effectively enhance the fault diagnosis accuracy .