摘要
振动信号的时频特征提取是分析大型工程结构物健康状态、评估结构设计参数的有效方法,也是及时诊断与预警结构异常的重要前提。针对现有特征提取方法不能准确提取完整时频特征的问题,提出一种结合变分模态分解(variational mode decomposition,VMD)与希尔伯特-黄变换(Hilbert-Huang transform,HHT)的时间-频率-能量多维度特征提取的新方法(VMD-HHT)。首先利用VMD提取振动事件的频率分量,滤除噪声分量;然后对提取出的振动信号进行HHT变换,获取振动事件的时间-频率-能量谱;最后为求取振动位移量,在提取的特征频率基础上,通过对加速度数据进行频域积分获取振动形变量。通过仿真分析和海洋石油平台上振动监测实测发现,所提方法能够精确提取时间-频率-能量多维特征信息,VMD-HHT模型辅助的频域积分方法求取的振动位移具有较高的精度和可靠性。
Objectives:Time and frequencies are important for investigating dynamic responses of offshore oil platforms.Time-frequency feature extraction is a key issue in structural health monitoring of large civil engineering structures.Current popular methods for signal analysis and processing cannot extract the characteristics of dynamic responses accurately and completely.Therefore,a new approach combining variational mode decomposition(VMD)and Hilbert-Huang transform(HHT)is proposed to extract multi-dimensional vibration characteristics of time,frequency and energy.Methods:Firstly,VMD is used to extract the vibration frequency component of events and the noise component could thus be removed.Then,HHT is applied to extract multi-dimensional vibration characteristics of time,frequency and energy using the cleaned accelerometer data.Finally,the frequency domain integration approach is introduced to compute the vibration displacements based on cleaned data processed by VMD.A series of simulation tests and ship impact test on offshore oil platform are performed using accelerometers in the MEMS IMU.Results:Experimental results show that the extraction method based on VMD-HHT can obtain the multi-dimensional time-frequencyenergy characteristics accurately.The frequency domain integration approach can produce highly accurate and reliable vibration displacements using accelerometer data.Conclusions:VMD-HHT feature extraction model can not only extract the vibration frequency more completely and accurately,but also clearly see the time of vibration events,the corresponding frequency range and vibration intensity.VMD-HHT aided frequency domain integration approach can effectively remove the low-frequency noise and determine cut-off frequency quickly.
作者
柳絮
王坚
李文
LIU Xu;WANG Jian;LI Wen(School of Geomatics and Urban Spatial Informatics,Beijing University of Civil Engineering and Architecture,Beijing 102616,China;Research Center for Urban Big Data Applications,Beijing University of Civil Engineering and Architecture,Beijing 100044,China)
出处
《武汉大学学报(信息科学版)》
EI
CAS
CSCD
北大核心
2021年第11期1686-1692,共7页
Geomatics and Information Science of Wuhan University
基金
国家自然科学基金(41874029)
北京建筑大学研究生创新项目(PG2020065)
关键词
振动监测
加速度计
变分模态分解
希尔伯特-黄变换
频域积分
vibration monitoring
accelerometer
variational mode decomposition
Hilbert-Huang transform
frequency domain integration