基于掩膜分量的改进HHT方法在电能质量扰动信号定位中的应用

Application of Mask Component-based Refined HHT in the Locating of Power Quality Disturbance Signal

为克服EMD方法的缺点,提出了一种基于掩膜分量的改进HHT方法对电能质量进行检测。首先对信号进行神经网络预测延拓,并对延拓部分进行加窗处理,有效的抑制了端点效应在经验模态分解过程中带来的影响。然后用掩膜信号法对电能质量扰动信号进行分解,得到包含单一频率的精确的经验模态函数分量,再对各分量进行Hilbert变换,并对瞬时幅值求导就能确定出电能质量扰动的起止点。仿真结果表明,基于掩膜的改进HHT方法能有效克服端点效应和模态混叠对信号分解的影响,适用于各种暂态扰动的分析。

To overcome the shortcoming of EMD, this paper presents a mask component-based refined lqilbert-Huang transforms （HHT） to check power quality. First, the signals undergo extended forecasting in the neural network, and the extended part undergoes windowing process so that the end effect is suppressed in the EMD process. After that, the power quality disturbance signal is decomposed in the mask signaling method to obtain accurate intrinsic mode function （IMF） containing single frequency, before Hilbert transformation is performed and the start and end points of power quality disturbance are determined through derivation of the instantaneous amplitt, de. The simulation results indicate that the mask-based refined HHT can effectively overcome tbe affection of the end effect and mode mixing on signal decomoosition, and is suitable for analysis of instantaneous disturbances.