关键词: 电能质量, 暂态, 分形, 熵, power quality, transient, entropy

Abstract: Transient power quality may cause severe loss to sensitive customers; therefore, it is very important to recognize overlapped transient power quality under general terms. Starting from graphical recognition and based on fractal theory, waveform was segmented and the entropy of waveform containing disturbance was calculated, then the most prominent disturbance was identified. By subtracting identified disturbance, other disturbances can be identified in remaining waveforms through iterated cycle of analysis. Simulation results reveal that this algorithm is very robust and adaptable, which can identify overlapped transient power quality disturbance with minor magnitude under noisy environment, and the recognition rate is very inspiring.

Key words: fractal