基于5次谐波分量概率分布的故障电弧识别方法的研究

发布时间：2018-05-17 06:40

本文选题：故障电弧 + 傅里叶分析　；参考：《沈阳工业大学》2017年硕士论文

【摘要】：自19世纪人类发明了电以来,电已经得到了很广泛的应用。在现如今人们的生活中家用电器得到了大量的普及,并且还有很多新型的家用电器不断涌现,电气火灾的发生的频率也越来越高、造成的危害也越来越大。这其中故障电弧是引发火灾的最主要原因。传统的保护电器可以对相间短路故障和接地短路故障时进行有效的保护,是针对大电流的过流保护。而串联故障电弧的特点是电流小,温度高,持续时间短,当发生故障电弧时,传统的过流保护并不能有效的对线路进行保护。因此,对串联故障电弧进行有效的检测和识别具有重要的意义。目前,国内对低压故障电弧的研究主要停留在论文阶段,几乎没有产品出现,而国外对于故障电弧的研究要比国内成熟的多,国外已经有AFCI的相关产品并且已经得到了应用。国内对于故障电弧的主要研究方法主要是通过对故障电弧电流、电压波形的特征分析,采用小波分析、傅里叶分析、自回归参数模型以及独立分量分析等方法分别在时域和频域上对故障电弧进行分析。然而由于建筑电气中具有负载多样性的特点,有些负载正常运行时的电流特性和故障电弧电流的特性很相似。此外,很多学者对故障电弧电流特性的研究大多采用的单一负载,对混合型负载研究的较少。因此本文根据国内外的研究成果搭建了故障电弧实验平台,采集了多种负载正常运行时和发生故障电弧时的电流波形。通过对采集的负载电流波形进行分析,观察负载电流正常运行时和发生故障电弧时的电流波形的不同点。然后采用MATLAB软件对电流波形进行傅里叶分析,提取出负载中正常运行时和发生故障电弧时的电流波形的谐波分量,对两种情况下的谐波分量进行对比分析,找到其发生故障电弧时的特征量,然后通过spss软件分别对其进行单参数的t-分布检验和k-s检验,确定特征量的分布类型为对数正态分布,然后确定故障电弧判据的特征量的阈值范围。由此本文提出了一种基于5次谐波分量的故障电弧识别方法,为故障电弧的检测提供了可靠的参考依据。
[Abstract]:Electricity has been widely used since the invention of electricity in the 19 th century. Nowadays, the household appliances have been widely used in people's lives, and there are many new types of household appliances, and the frequency of electric fires is higher and higher, and the harm caused is more and more serious. The fault arc is the main cause of the fire. The traditional protective apparatus can effectively protect the interphase short circuit fault and the ground short circuit fault, which is aimed at the overcurrent protection of high current. The characteristics of series fault arc are low current, high temperature and short duration. When fault arc occurs, the traditional over-current protection can not effectively protect the line. Therefore, it is of great significance to detect and identify the series fault arc effectively. At present, the domestic research on low-voltage fault arc mainly stays in the paper stage, almost no products appear, but the foreign research on the fault arc is much more mature than at home. There have been related products of AFCI in foreign countries and have been applied. The main research methods of fault arc in China are the characteristic analysis of fault arc current and voltage waveform, wavelet analysis and Fourier analysis. The autoregressive parameter model and independent component analysis are used to analyze the fault arc in time domain and frequency domain respectively. However, due to the diversity of loads in building electricity, the current characteristics of some loads during normal operation are similar to those of fault arc currents. In addition, many scholars study the characteristics of fault arc current mostly using a single load, and less research on hybrid load. Therefore, based on the domestic and foreign research results, a fault arc experimental platform is built, and the current waveforms of various loads during normal operation and fault arc are collected. Through the analysis of the collected load current waveform, the differences between the load current waveform and the fault arc wave are observed. Then using MATLAB software to carry on the Fourier analysis to the current waveform, extract the harmonic component of the current waveform in the normal operation of the load and the fault arc, and carry on the contrast analysis to the harmonic component of the two kinds of cases. The characteristic quantity of the fault arc is found, and the single parameter t-distribution test and k-s test are carried out by spss software, and the distribution type of the characteristic quantity is determined as logarithmic normal distribution. Then the threshold range of the characteristic quantity of the fault arc criterion is determined. In this paper, a method of fault arc identification based on the fifth harmonic component is proposed, which provides a reliable reference for fault arc detection.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM501.2;TU85

【参考文献】