摘要
针对近断层地震速度脉冲信号非线性非平稳的特点,建立了以互补集合经验模态分解(Coplementary Ensemble Empirical Mode Decomposition,CEEMD)为基础的一种降噪算法。该算法首先对仿真信号进行CEEMD操作,获得从高频到低频的固有模态函数(Intrinsic Mode Function,IMF),然后对IMF筛选叠加,得到多组含不同阶数的重构信号;通过算法相关度和逼近度组成综合评价指标,对多组重构信号进行筛选,得到全局最优重构组合Rec3,实现信号的有效去噪。然后采用以所选最优重构组合为基础的降噪算法分析实测速度脉冲信号。结果表明:该降噪算法具有很好的降噪效果,所得结果曲线较原信号曲线光滑平整,主脉冲信号清晰可辨,且由于算法本身没有人工加窗这类操作,使得其自适应良好,具有一定的实用性和可靠性。
Aiming at the nonlinear and non-stationary characteristics of near-fault pulse-like earthquake signals,a denoise algorithm based on Complementary Ensemble Empirical Mode Decomposition (CEEMD) was established.Firstly,the algorithm performs CEEMD on the simulated signal to obtain the Intrinsic Mode Function (IMF) components ranging from high frequency to low frequency.The IMFs are superimposed to obtain multiple sets of reconstructed signals with different orders.A comprehensive evaluation index composed of algorithm correlation and approximation degree is used to screen multiple sets of reconstructed signals,and the global optimal reconstruction combination Rec3 is obtained to achieve effective denoising of signals.The selected excellent noise reduction algorithm is applied to analyze the measured velocity pulse signal.The results show that the noise reduction algorithm has a good noise reduction effect,and the obtained result curve is smoother than the original signal curve.The main pulse signal is also clearly identif iable .The algorithm itself has no manual windowing operation,contributing better self-adaptability.Therefore,the algorithm is practicable and reliable.
作者
刘欣悦
单德山
谭康熹
LIU Xinyue;SHAN Deshan;TAN Kangxi(School of Civil Engineering,Southwest Jiaotong University,Chengdu Sichuan 610031,China)
出处
《铁道建筑》
北大核心
2019年第5期59-63,共5页
Railway Engineering
基金
国家重点研发计划(2016YFC0802202)
国家自然科学基金(51678489)
四川省科技计划(2016JY0130)
云南省交通运输厅科技计划(2017(A)03)
中电建路桥集团资助科研项目(SCMQ-201728-ZB)
关键词
公路桥梁
信号降噪
数值计算
互补集合经验模态
速度脉冲
Highway bridge
Signal denoise
Numerical calculation
Complementary ensemble empirical mode
Velocity pulse
