摘要
提出一种基于变分模态分解与改进小波阈值的水轮机组振动信号降噪方法。首先,利用VMD将信号分解为若干个本征模态函数(IMF),并根据IMF的能量和频率特征进行筛选,去除无效模态和噪声模态;然后,对保留的有效模态进行小波变换,通过自适应阈值法对小波系数进行阈值处理;最后,对处理后的小波系数进行逆变换,再利用各个有效模态重构有效降噪信号。经MATLAB验证分析可知,联合VMD-改进小波阈值函数降噪法能有效降低信号均方误差,显著提升水轮机组振动信号的降噪效果。
The signals processed by traditional wavelet threshold have constant deviations compared with the original signals,and the components generated by empirical mode decomposition(EMD)will have the problem of mode aliasing.Therefore,this paper proposes a method of noise reduction of the vibration signals of the water turbine unit based on variational mode decomposition(VMD)and improved wavelet threshold.First,the signals are decomposed into several intrinsic mode functions(IMF)by means of the VMD,and they are screened according to the energy and frequency characteristics of the IMF to remove the invalid modes and noise modes.Then,the retained effective modes are transformed by the wavelet,and the wavelet coefficients are processed by self-adaptive threshold method.Finally,the processed wavelet coefficients are inverted and the effective noise reduction signals are reconstructed by using each effective mode.The MATLAB verification analysis shows that the method of noise reduction based on the combination of the VMD and improved wavelet threshold function can effectively reduce the mean square errors of the signals and significantly improve the effect of noise reduction of the vibration signals of the water turbine unit.
作者
田维坤
胡峰
喻潇
彭海龙
蒋东荣
TIAN Weikun;HU Feng;YU Xiao;PENG Hailong;JIANG Dongrong(Ahai Power Generation Branch of Yunnan Huadian Jinsha River Midstream Hydropower Development Co.,Ltd.,Lijiang Yunnan 674100,P.R.China;Jiangsu Xinnong Microelectronic Technology Co.,Ltd.,Nantong Jiangsu 226361,P.R.China;School of Electrical and Electronic Engineering of Chongqing University of Technology,Chongqing 400045,P.R.China)
出处
《重庆电力高等专科学校学报》
2023年第5期5-9,38,共6页
Journal of Chongqing Electric Power College
基金
2023年重庆市科学技术委员会自然科学基金面上项目(CSTB2023NSCQ-MSX0279)。
关键词
水轮机组
小波降噪
VMD
信号处理
water turbine unit
wavelet noise reduction
VMD
signal processing
