【摘要】 针对串联型故障电弧影响供配电系统供电安全的问题,提出小波包信息熵理论与短时傅立叶变换相结合的串联型故障电弧频谱特征分析方法。首先研制了低压串联型故障电弧实验平台,开展了典型负载故障电弧模拟实验。其次,利用小波包技术对发生稳定串联型故障电弧前后的电流信号进行分解、重构和归一化处理。然后,计算各频带重构信号的信息熵,通过对比各频段燃弧前后重构信号的信息熵得出稳定故障电弧电流信号的特征频段为8-10.8KHz。最后,采用短时傅立叶变换对特征频段的重构信号进行时频分析,得出发生稳定故障电弧前后电流信号的频谱变化规律。经验证,在发生不稳定故障电弧时,8-10.8KHz仍为故障电弧的特征频段,频谱特征依然明显。结果表明,利用信息熵和短时傅立叶变换对低压串联型故障电弧进行频谱特征研究是可行的。更多还原

【Abstract】 In the view of power supply safety problems influenced by series arc fault, a kind of analysis method of series arc fault spectral characteristic was put forward by the combination of wavelet packet entropy and short-time Fourier transform(STFT). Firstly, a low voltage series arc fault experiment platform was developed. A series of simulation experiments of typical loads were carried out. Secondly, the series arc fault current signal before and after arc burning stably was decomposed, restructured and normalized by using the frequency band energy decomposition technique of wavelet packet. Thirdly, the information entropy of reconstructed signal of each frequency band was calculated. By comparing the information entropy before and after the arc fault, the characteristic frequency band of current signal was 8-10.8KHz. Finally, the spectrum variation of the characteristic frequency band before and after the arc fault was obtained by STFT. When the unstable arc occurred, 8-10.8 KHz was still the arc fault characteristic frequency and frequency spectrum characteristics are still obvious. The results showed that it was feasible to study low-voltage series fault arc spectrum characteristics by wavelet packet and STFT.更多还原