基于灰狼算法的变压器特征选取方法研究

Research on Transformer Feature Selection Method Based on Grey Wolf Optimization Algorithm

本研究介绍一种灰狼算法在特征参量空间选取特征参量子集的方法,简化模型训练复杂度,并在Matlab中进行算法验证。结果表明:该算法能够将35个特征参量减小到5个,且模型准确率大于85%。

A method of selecting a subset of characteristic parameters in the characteristic parameter space by using gray Wolf algorithm was introduced and then verified in Matlab.The simulation results show that the algorithm can reduce the 35 feature parameters to 5,and the accuracy is greater than 85%.