摘要
Abstract
In order to solve the problem of low identification accuracy of transformer inrush current under strong interference,based on the principle analysis of transformer inrush current,a mathematical model for transformer inrush current identification that integrates multiple characteristic quantities of time domain,frequency domain and time frequency domain is proposed.The discontinuity angle is selected as the time domain characteristic quantity,correlation coefficient,amplitude of the second har-monic,phase difference between the second harmonic and the fundamental wave are selected as frequency domain characteris-tics,and the energy of characteristic quantity wavelet packet in time-frequency domain is extracted by wavelet packet with ex-cellent time-frequency performance,then the improved random forest method is used to realize the nonlinear mapping between the multivariate characteristic quantity and the transformer current category.The transformer inrush current and internal fault current simulation model is established.Through comparative analysis of identification under different interference levels,it is proved that the proposed method has higher recognition accuracy and stronger anti-interference ability.关键词
变压器/励磁涌流/特征量/小波包变换/随机森林Key words
transformer/excitation inrush current/characteristics/wavelet packet transform/random forest分类
信息技术与安全科学
