基于NGA优化支持向量机的电力变压器故障诊断

为提高电力变压器故障诊断的准确性,提出一种支持向量机（Support Vector Machines,SVM）的故障诊断方法.该方法用添加最优保存策略的小生境策遗传算法对SVM进行参数优化,确保种群中适应度高的个体能被保留到下一代,使优化对象比较容易稳定,以得到更优良的个体,提高诊断精度.通过与遗传算法优化SVM及标准小生境遗传算法优化SVM的诊断结果相比较,根据对比结果表明：所提方法对变压器故障数据的分类辨识效果更好.

Power transformer fault diagnosis based on support vector machine with NGA

In order to improve the accuracy of fault diagnosis of power transformer, the Support Vector Machine （SVM） was applied to its fault diagnosis. The niche genetic algorithm combined with Elite preservation is used to optimize the parameters of SVM. To ensure that the population of individuals with the highest fitness can be retained to the next generation. In order to obtain the excellent individuals,and make the optimization object more stable. Applied this method to the transformer fault diagnosis, and compared with the diagnosis results of other algorithms. The results show that, this method has better identification effect on transformer fault diagnosis.