基于BP神经网络和全相位快速傅里叶变换的电力系统谐波检测技术研究
2017-02-16曹英丽尹希哲
曹英丽+尹希哲


摘 要: 在分析现有谐波检测方法的基础上,研究了基于全相位快速傅里叶变换和人工神经网络的谐波和间谐波检测算法。针对谐波检测精度不高的问题,提出了一种基于全相位快速傅里叶变换和BP神经网络的谐波检测方法,同时采用全相位快速傅里叶变换和自适应神经网络的谐波检测方法,进一步提高了谐波检测的精度。通过虚拟仪器软件开发平台LabWindows/CVI,设计了基于全相位快速傅里叶变换和自适应神经网络的谐波检测软件。软件实现了谐波幅值和相位的检测以及总谐波畸变率的计算,并能够在总谐波畸变率超标的情况下给出警报。
关键词: 谐波; 间谐波; 全相位快速傅里叶变换; 人工神经网络; 虚拟仪器
中图分类号: TN711?34; TM417 文献标识码: A 文章编号: 1004?373X(2017)01?0125?04
Abstract: On the basis of analyzing the available harmonic detection methods, the harmonic and interharmonic detection method based on all?phase fast Fourier transform and artificial neural network is studied. A new harmonic detection method based on all?phase fast Fourier transform and BP neural network is proposed to solve the problem of low harmonic detection precision. And a harmonic detection method based on all?phase fast Fourier transform and adaptive neural network is used to further improve the accuracy of harmonic detection. A harmonic detection software based on all?phase fast Fourier transform and adaptive neural network was designed on virtual instrument software development platform LabWindows/CVI. The software can realize the detection of harmonic amplitude and phase and calculation of total harmonic distortion, and give an alarm when the total harmonic distortion is out of limit.
Keywords: harmonic; interharmonic; all?phase fast Fourier transform; artificial neural network; virtual instrument
在理想情况下,电力系统的电能应该是具有单一频率、单一波形和若干电压等级的正弦电压信号。但是实际生产生活中由于一些原因,电网中的电能很难保持理想的波形,实际的波形总是存在偏差和形变,这种波形畸变称为谐波畸变[1]。造成谐波畸变的原因是电网中存在大量的电力系统谐波。随着谐波污染问题愈加严重,其产生的危害也越来越广泛。因此,谐波检测问题具有十分重要的研究价值和意义[2]。
1 基于全相位快速傅里叶变换和BP神经网络
的谐波检测
1.1 谐波相角检测
全相位快速傅里叶变换具有相位不变性。利用该性质对电网电压信号的采样值进行全相位快速傅里叶变换谱分析,获得高精度的谐波相位值[3]。……
