摘要
针对水工结构在泄流激励作用下动力响应信号被噪声淹没的实际问题,提出一种基于自适应噪声的完整集成经验模态分解(CEEMDAN)和改进小波阈值函数联合降噪的方法。对水工结构振动信号先进行CEEMDAN分解,将信号分解成一系列固有模态(IMF)分量,运用自相关函数和互相关系数剔除无效IMF分量,实现振动信号的初次降噪;再利用改进小波阈值函数对含有主要振动信息的IMF分量进行降噪处理,实现含噪信号的二次降噪;最后,将包含主要结构振动信息的IMF分量和残量进行信号重构,得到降噪后的振动信号。通过数值仿真分析,发现CEEMDAN-改进小波阈值联合降噪后信号(10%噪声水平)的信噪比从20.21提升到23.54,均方根误差从0.19降低到0.09,验证了联合降噪具有更高的信噪比和更低的均方根误差,更为接近纯净信号。结合国内某水电站导墙实测数据,运用该方法进行降噪分析,结果表明该方法具有较好的降噪性能和应用前景,可为水工结构流激振动安全诊断提供有效依据。
To solve the problem that the dynamic response signals of hydraulic structures were drowned by noise under discharge excitation,a joint denoising method based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and improved wavelet threshold function was proposed.The vibration signal of the hydraulic structure was decomposed into a series of intrinsic mode functions(IMF)components by CEEMDAN decomposition,and the invalid IMF components were eliminated by using autocorrelation function graph method and cross-correlation coefficient method to achieve the initial noise reduction of the vibration signals.Then the IMF component containing the main vibration information was denoised by the improved wavelet threshold function to achieve the secondary denoising of the noisy signal.Finally,the IMF component and residual containing the main structural vibration information were reconstructed to obtain the vibration signal after noise reduction.Through numerical simulation analysis,the signal-to-noise ratio of CEEMDAN-improved wavelet threshold combined denoising(10%noise level)was improved from 20.21 to 23.54,and the root mean square error was reduced from 0.19 to 0.09,which proved that the combined denoising has higher signal-to-noise ratio and lower root mean square error,and was closer to pure signal.Combined with the measured data of the guide wall of a hydropower station in China,this method was used to analyze the noise reduction.The results showed that this method has good noise reduction performance and application prospect,and provided an effective basis for the safety diagnosis of flow induced vibration of hydraulic structure.
作者
李火坤
邬鹏贞
黄伟
刘晗玥
LI Huokun;WU Pengzhen;HUANG Wei;LIU Hanyue(School of Infrastructure Engineering,Nanchang University,Nanchang 330031,China)
出处
《南昌大学学报（工科版）》
CAS
2022年第1期1-8,共8页
Journal of Nanchang University(Engineering & Technology)
基金
国家自然科学基金资助项目(51879126,52079061,51909115)
江西省自然科学基金项目(20192BAB216038)
江西省水利厅科技项目(课题)(202123YBKT04)。
关键词
水工结构
自适应噪声的完整集成经验模态分解
改进小波阈值函数
信号降噪
hydraulic structure
complete ensemble empirical mode decomposition with adaptive noise
improved wavelet threshold function
signal noise reduction
作者简介
李火坤(1981—),男,教授,博士,研究方向为水工水力学与工程安全检测;通信作者:黄伟(1990—),男,讲师,博士,研究方向为水工水力学方面研究。E-mail:huangwei0214@ncu.edu.cn。
