基于分数阶傅里叶变换的直流电弧特征分析

Analysis of DC Arc Feature Based on Fractional Fourier Transform

针对单纯从时域或频域方面检测电弧故障的缺陷,提出了采用分数阶傅里叶变换分析电弧的时频特征。通过试验模拟不同负载时直流串行与并行电弧故障,发现改变变换阶次使得电弧分数阶傅里叶域的特征比频域特征更加明显,有利于提取电弧特征。同时比较了负载性质及电弧类型对电弧分数阶傅里叶域特征的影响,及在线检测电弧故障时分数阶傅里叶变换和快速傅里叶变换算法复杂度与实时性。

For the shortcomings of detecting the arc fault simply with time domain or frequency domain methods,an analysis method of arc time-frequency feature was proposed based on fractional Fourier transform （FRFT）.DC serial and parallel arc faults under different load were carried out and the arc fault data was obtained from the experiment.Based on the analysis of the data,the feature of the arc on the fractional Fourier domain is more obvious than the frequency domain by changing the transform order p and makes the feature extracting more easier. Then,the influence on the fractional Fourier domain of different types of load and arc was analyzed.Finally,the algorithm complexity and real-time property were compared between（FRFT）and fast Fourier transform.