摘要
随着大功率器件使用,造成电网中有大量谐波,威胁设备的安全。提出运用小波神经网络(Wave Neural Network,WNN)算法来检测谐波。首先,针对神经网络初始值设置不当导致的网络收敛慢甚至不收敛的问题,提出了网络初始参数自相关修正的优化方法,提高了网络的性能。其次,运用附加动量项的训练算法平滑了权值学习路径,有效避免了网络训练陷入局部最小,提高了谐波检测精度。最后,经过与其它检测方法的仿真对比,证明了所述方法具有收敛速度快,检测精度高的优点。
With the use of high power devices,resulting in a large number of harmonics in power grid,which threat the safety of equipment. This paper proposes the use of wavelet neural network( WNN) algorithm to detect harmonics. Firstly,the initial value of neural network convergence set due to improper slow or even non-convergence problems,a method of optimal initial parameters correlation correction is put forward to improve the network performance. Secondly,smoothing training algorithm with additional momentum item weights learning path is adopted to avoid network training into local minimum,thus improving the precision of harmonic detection. Finally,it is proved that the method presented in this paper has advantages of fast convergence speed,detection high precision through simulation and comparison with other detection methods.
作者
李圣清
王飞刚
朱晓青
Li Shengqing;Wang Feigang;Zhu Xiaoqing(School of Electrical and Information Engineering, Hunan University of Technology, Zhuzhou 412007, Hurnan, China)
出处
《电测与仪表》
北大核心
2019年第10期118-121,共4页
Electrical Measurement & Instrumentation
基金
国家自然科学基金资助项目(61673165)
湖南省自然科学基金资助项目(2017JJ4024)
湖南省教育厅开放基金(15k036)
湖南省重点实验室项目(2016TP1018)
关键词
谐波
小波神经网路
神经网络
自相关
收敛
优化
harmonic
wavelet neural network
neural network
autocorrelation
convergence
optimization
