基于小波变换和曲线拟合的电网畸变信号检测方法研究

发布时间：2018-05-08 04:04

本文选题：电网畸变信号 + 畸变信号检测　；参考：《哈尔滨理工大学》2017年硕士论文

【摘要】：随着电网中非线性负载的增加,电网基波信号上会叠加畸变信号形成电网畸变信号,畸变信号的检测精度直接决定了电网畸变信号条件下电能计量的准确性。电网畸变信号由于具有时变、频率复杂特性,使电能计量准确性降低,让国家电能计量部门利益受损。部分电网畸变信号研究都是在已知畸变信号频率能量分布基础上进行分析,有利于选取合适的小波变换分解重构层数。当小波变换分解层数过多或过少时,会降低电网畸变信号的畸变信号检测精度。解决以上问题需要有能够在小波变换分解层数不合适时,仍能够准确分析检测电网畸变信号的方法。本文提出了一种基于小波变换与正弦曲线拟合的电网畸变信号检测方法。首先利用小波变换对电网畸变信号进行分解与重构,得到基波信号,再通过傅里叶级数将重构得到的基波信号进行正弦拟合,最后在电网原始畸变信号中减去正弦信号,得到畸变信号。本文以稳态电网畸变信号模型和非稳态电网畸变信号模型为例对本文方法进行验证,具体研究内容如下:1.研究电网畸变信号条件下畸变信号的数学模型,通过建立稳态与非稳态电网畸变信号的数学模型,分析电网信号中畸变信号对电网电能质量的影响以及已有电网畸变信号检测方法的局限性。2.寻找适用于电网畸变信号频率及能量分布未知的电网畸变信号检测方法,首先将小波变换应用于电网畸变信号分解与重构中,然后通过曲线拟合方法与归一化将重构后基波信号拟合成标准基波信号,最后得到准确度高的畸变信号。3.对几种叠加不同类型畸变信号的典型电网畸变信号进行数学建模,通过本文方法对以上电网畸变信号进行仿真实验,并与单一小波变换方法所得结果进行比较,分析本文方法相对于传统方法的优势。实验结果为:在电网畸变信号分析检测领域,通过本文方法重构信号与原始信号的相关系数稳定在0.95以上,即两者的一致程度在95%以上。在相同分解层数下,本文方法都可以提高单独小波变换方法重构所得基波信号和畸变信号与原始信号的一致程度;而在不同分解层数,本文方法检测结果没有因为小波变换方法分解层数改变而有较大变化,畸变信号检测稳定性高。
[Abstract]:With the increase of nonlinear load in the power network, the distortion signal will be superimposed on the fundamental wave signal to form the distortion signal. The accuracy of the power measurement under the condition of the distortion signal is directly determined by the detection accuracy of the distortion signal. Because of its time-varying and complex frequency characteristics, the power grid distortion signal reduces the accuracy of electric energy metering and impairs the interests of the national electric energy metering department. Based on the known frequency energy distribution of the distortion signal, some researches on the distortion signal in the power network are analyzed, which is helpful to select the appropriate number of wavelet transform decomposition and reconstruction layers. When the number of decomposition layers of wavelet transform is too large or too small, the detection accuracy of distortion signal of power network distortion signal will be reduced. To solve the above problem, we need to have a method that can accurately analyze and detect the distortion signal of power network when the number of decomposition layers is not suitable. In this paper, a method of detecting power grid distortion signal based on wavelet transform and sinusoidal curve fitting is proposed. The fundamental signal is obtained by decomposing and reconstructing the distortion signal of power network by wavelet transform, then the reconstructed fundamental signal is fitted by Fourier series, and finally the sinusoidal signal is subtracted from the original distortion signal. The distortion signal is obtained. In this paper, the distortion signal model of steady-state power network and the distortion signal model of unsteady power network are taken as examples to verify the method in this paper. The detailed research contents are as follows: 1. This paper studies the mathematical model of the distortion signal under the condition of power grid distortion signal, and establishes the mathematical model of the steady-state and non-steady-state power grid distortion signal. This paper analyzes the influence of distortion signal on power quality and the limitation of existing methods. In order to find the detection method of the distortion signal which is suitable for the unknown frequency and energy distribution of the distortion signal in the power network, the wavelet transform is first applied to the decomposition and reconstruction of the distortion signal of the power network. Then the reconstructed fundamental signal is quasi-synthesized into the standard fundamental signal by curve fitting method and normalization. Finally, the distortion signal with high accuracy is obtained. Several typical power grid distortion signals superimposed with different types of distortion signals are modeled mathematically. The simulation experiments of the above power grid distortion signals are carried out by using this method, and the results are compared with the results obtained by the single wavelet transform method. The advantage of this method over the traditional method is analyzed. The experimental results are as follows: in the field of power grid distortion signal analysis and detection, the correlation coefficient between the reconstructed signal and the original signal is stable above 0.95, that is, the consistency of the two signals is over 95%. Under the same number of decomposition layers, the method in this paper can improve the consistency between the fundamental signal and the distortion signal reconstructed by the single wavelet transform method and the original signal, but in different decomposition layers, The detection results of this method have not changed greatly because of the decomposition layer number of wavelet transform method, and the detection stability of distortion signal is high.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911.23;TM933.4

【相似文献】