摘要
准确有效识别出水电站厂房振动信号的各个振源,对于水电站长期安全稳定运行有重要指导意义,盲源分离(blind source separation,BSS)是进行信号分解与振源识别的一种有效方法。为了实现水电站厂房复杂环境下振动信号的盲源分离,建立一种基于鲸鱼算法变分模态分解(whale optimization algorithm and variational mode decomposition,WOA-VMD)降噪改进的信号盲源分离模型。采用WOA-VMD和相关法对观测信号进行降噪处理,确保盲源分离结果的准确性;求解多维降噪信号的协方差矩阵并进行奇异值分解,采用优势特征值法进行源数估计;最后对降噪信号进行中心化、白化前处理,通过联合近似对角化算法得到分离矩阵,实现观测信号的盲源分离。仿真结果表明:相较于传统盲源分离模型,改进模型将分离信号与源信号的相关系数分别提升了9.1%,7.1%,8.3%,分离信号主频误差也均有所降低。将该方法运用到水电站厂房振动工程实例,也取得了较好的分离效果。
Accurately and effectively identifying various vibration sources of hydropower plant vibration signals has important guiding significance for long-term safe and stable operation of hydropower station.Blind source separation(BSS)is an effective method for signal decomposition and vibration source identification.Here,to realize BSS of vibration signals in complex environment of hydropower plant,an improved BSS model based on whale optimization algorithm and variational modal decomposition(WOA-VMD)was proposed and established.WOA-VMD and the correlation method were used to do denoising for observed signals,and ensure the accuracy of BSS results.Covariance matrix of multi-dimensional denoised signals was solved and singular value decomposition was performed for it,and the dominant eigenvalue method was used to estimate source number.Finally,de-noised signals were pretreated with centralization and whitening,and the separation matrix was obtained with the joint approximate diagonalization algorithm to realize BSS of observed signals.The simulation results showed that compared with the traditional BSS model,correlation coefficients between separated signals and original signals are improved by 9.1%,7.1%and 8.3%,respectively with the proposed model,and main frequency errors of separated signals are also reduced;the proposed method is applied in vibration engineering examples of hydropower plant,and better separation effect can be obtained.
作者
王海军
郝志豪
刘通
胡航
梁超
WANG Haijun;HAO Zhihao;LIU Tong;HU Hang;LIANG Chao(State Laboratory of Hydraulic Engineering Simulation and Safety,Tianjian University,Tianjin 300350,China;School of Civil Engineering,Tianjin University,Tianjin 30050,China;Ertan Hydropower Plant,Yalong River Hydropower Development Company,Ltd.,Panzhihua 617000,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第9期222-229,共8页
Journal of Vibration and Shock
基金
国家自然科学基金青年项目(51909185)。
关键词
水电站厂房
盲源分离(BSS)
鲸鱼算法
源数估计
联合近似对角化
hydropower plant
blind source separation(BSS)
whale optimization algorithm(WOA)
source number estimation
joint approximate diagonalization
