基于变分模态分解与排列熵的输电线路故障诊断

发布时间：2018-05-24 06:54

本文选题：电力系统 + 输电线路　；参考：《安徽理工大学》2017年硕士论文

【摘要】：输电线路的安全稳定对电力系统可靠运行至关重要,随着电力系统向智能化、复杂化方向发展,输电线路的故障诊断与健康管理正逐步受到重视。输电线以及配电线在电网中遍布在系统的每个角落并处于露天的状况下,故障发生率相比其他电气设备高很多,因而传统的人工巡线方式查找故障类型已无法满足当前电力系统发展的要求。因此,基于输电线路状态信号的快速、可靠的输电线路故障分析方法对电力系统的安全稳定运行及科学健康管理具有重要的意义。输电线路发生故障时,故障信息隐含于故障信号中,可通过适当的故障信号分析方法以实现其有效诊断。现有电力系统故障信号分析方法种类繁多,其中傅里叶变换、小波变换、经验模态分解等应用最为广泛。傅里叶变换在分析线性信号的全局性具有良好的优势,但不能有效的给出非线性信号的局部特征信息。小波分析具有对被分析故障信号可进行任意的放大平移并对其特征进行提取的优点,但该方法存在选取不同小波基和分解尺度造成故障信号特征的遗失及分析故障信号无法反映信号的本质特征的缺点。经验模态分解一种新的信号分析方法,它克服小波分析的缺点,自动地选择最佳基函数对信号进行分解,确定信号在不同频带的分辨率,避免了选取小波基与分解尺度的困难。但该方法属于递归式模态分解易出现模态混叠造成某个分量频带过宽、噪声干扰过多而掩盖故障信号微弱的特征信息。由于输电线路发生故障时,其故障波形因时间、故障点、故障过渡电阻及系统工作情况的不同而有所差异,具有突变性及随机性,在故障信号中含有大量的非周期性分量和大量的谐波并且这些分量随着时间会逐渐衰减,因此在上述三种传统方法分析输电线路故障信号时,会遗失信号的部分有效特征信息,并影响最终的故障诊断结果。本文鉴于以上原因,提出基于变分模态分解和排列熵结合的方法来诊断输电线路故障,以此解决上述存在的问题。本文输电线路故障诊断主要步骤分为:第一步骤,利用小波变换与排列熵的方案对输电线路故障进行处理分析;第二步骤,利用变分模态分解与排列熵的方案对输电线路故障进行处理分析。最后将两种方案的最终结果进行比较。针对上述算法,本文采用PSCAD和MATLAB对算法进行仿真。
[Abstract]:The safety and stability of transmission line is very important to the reliable operation of power system. With the development of intelligent and complicated power system, fault diagnosis and health management of transmission line are being paid more and more attention. Transmission lines and distribution lines are scattered throughout the system and in open air conditions, and the incidence of failures is much higher than that of other electrical equipment. Therefore, the traditional manual inspection method can not meet the requirements of the current power system development. Therefore, the fast and reliable fault analysis method based on the transmission line state signal is of great significance to the safe and stable operation and scientific health management of power system. The fault information is hidden in the fault signal when the fault occurs in the transmission line, and it can be effectively diagnosed by the appropriate fault signal analysis method. There are many kinds of power system fault signal analysis methods, among which Fourier transform, wavelet transform and empirical mode decomposition are the most widely used. Fourier transform has a good advantage in analyzing the global character of linear signal, but it can not give the local characteristic information of nonlinear signal effectively. Wavelet analysis has the advantages of arbitrary amplification and translation of the analyzed fault signal and extraction of its features. However, this method has the disadvantage of selecting different wavelet bases and decomposing scale, which results in the loss of fault signal features and the analysis of fault signal can not reflect the essential characteristics of the signal. Empirical mode decomposition (EMD) is a new signal analysis method, which overcomes the shortcoming of wavelet analysis, automatically selects the best basis function to decompose the signal, determines the resolution of the signal in different frequency bands, and avoids the difficulty of selecting wavelet basis and decomposition scale. But this method belongs to the recursive mode decomposition, which is prone to appear mode aliasing, which leads to a component frequency band too wide, too much noise interference, and masking the weak characteristic information of the fault signal. Because the fault waveform of transmission line is different because of different time, fault point, fault transition resistance and system working condition, it has the character of mutation and randomness. There are a lot of aperiodic components and lots of harmonics in the fault signal and these components will attenuate gradually with time. Therefore, some effective characteristic information of the transmission line fault signal will be lost when the three traditional methods mentioned above are used to analyze the fault signal. And affect the final fault diagnosis results. In view of the above reasons, a method based on variational mode decomposition and permutation entropy is proposed to diagnose the fault of transmission lines and solve the problems mentioned above. In this paper, the main steps of transmission line fault diagnosis are as follows: the first step, using wavelet transform and permutation entropy to deal with the transmission line fault, the second step, The scheme of variational mode decomposition and permutation entropy is used to deal with transmission line faults. Finally, the final results of the two schemes are compared. In this paper, PSCAD and MATLAB are used to simulate the algorithm.
【学位授予单位】：安徽理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM755

【参考文献】