基于IAFSA优化LS-SVM的变压器故障识别研究

Research on Transformer Fault Recognition Based on LS-SVM Optimized by Improved Artificial Fish Swarm Algorithms

为快速、准确识别变压器在运行过程中出现的不同故障,文章运用改进人工鱼群(IAFSA)对最小二乘法支持向量机(LS-SVM)的参数进行优化,并提出了一种新的变压器故障识别方法。首先,利用加权平均方法与调整函数对人工鱼群算法进行改进,并运用改进人工鱼群算法对最小二乘法支持向量机算法的惩罚因子与核函数进行优化,提高诊断模型的识别精度;其次,对变压器不同故障样本数据进行分析;最后与实际故障参数对比,验证该方法的准确性、可靠性。研究表明,基于改进人工鱼群算法优化LS-SVM的变压器故障识别方法可准确地识别变压器在运行过程中出现的故障,为变压器工作状态监测提供新的思路。

In order to quickly and accurately identify the different faults of transformers,this paper adopts the improved artificial fish swarm algorithm(IAFSA)to optimize the least squares support vector machine(LS-SVM),and proposes a new method of transformer fault identification.Firstly,the method of weighted average and adjustment function are used to improve the artificial fish swarm algorithm,and the IAFSA is used to optimize the penalty factor and the kernel function of the least squares support vector machine algorithm to improve the accuracy of recognition of the diagnosis model.Secondly,this method is used to analyze the different fault datas of transformer.Finally,the accuracy and reliability of the method are verified by comparing with the parameters of actual fault.The research shows that the method what the least squares support vector machine optimized by the IAFSA could accurately identify the faults occurring in the operation of transformer,and provides a new idea for the monitoring of transformer working status.