基于EEMD阈值去噪的电能质量多扰动分类

Classification of power quality multi-disturbance based on EEMD threshold denoising

针对含噪声的电能质量多扰动分类识别问题,提出一种基于EEMD阈值去噪的分类识别方法。首先依据源噪声信号在电能基波上不同频率和不同幅值叠加的特性,采用EEMD去噪法对信号源分解得到固有模态函数(IMF),消除高斯白噪声后,将得到的IMF分量转化为IMF能量值;最后,运用人工蜂群算法(ABC)优化在线极限学习机(OSELM)实现多扰动分类识别。MATLAB实例证明了提出方法的能够准确的对于扰动信号进行分类识别。

Aiming at the power quality multi-disturbance classification problem of noisy,a method based on EEMD threshold denoising is proposed.Firstly,based on the noise signal in the power characteristics of different frequency and amplitude of fundamental wave superposition,taking the EEMD of the signal source decomposition denoising method of intrinsic mode function(IMF)to the elimination of Gauss white noise,the IMF component into IMF energy value;finally,using the artificial bee colony algorithm(ABC)optimization the ultimate online learning machine(OSELM)is proposed to classify multi-disturbance.Simulation results in matlab show that the proposed method can classify and recognize the disturbance signals accurately.