Purpose The paper addresses the adverse effects of near-field pulse-like ground motion on long-period and seismic isolation structures, and solves the problems of overestimating/underestimating periods and relying on pulse model matching in traditional pulse period identification methods. A ground motion identification method combining time–frequency domain dual parameters is proposed to improve the accuracy of pulse period characterization and the reliability of structural nonlinear response evaluation. Method Based on 122 pulse-like and 10 non-pulse ground motion data in the NGA West2 database, the time-domain parameters-half wavelength width and its energy-weighted value <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(D_{{\text{w}}}\)</EquationSource> </InlineEquation>, as well as the frequency domain parameters-the energy-weighted average period <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(T_{{\text{H}}}\)</EquationSource> </InlineEquation> combining the Modified Ensemble Empirical Mode Decomposition (MEEMD) and Hilbert-Huang marginal spectrum are proposed. The modal aliasing was eliminated through MEEMD, and the effective components were screened by using permutation entropy. Combining half-wavelength energy weighting and marginal spectral energy distribution, a dual-parameter identification system of <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(D_{{\text{w}}}\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(T_{{\text{H}}}\)</EquationSource> </InlineEquation> was constructed, and the validity of the method was verified by the Bouc-Wen-Baber-Noori (BWBN) model. Results The <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(D_{{\text{w}}}\)</EquationSource> </InlineEquation> shows a strong positive correlation with seismic magnitude and traditional pulse period parameter Tp. <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(T_{{\text{H}}}\)</EquationSource> </InlineEquation> stabilizes the pulse period range to 1.14–3.98 s (while the traditional method is 0.4–13.12 s), solving the problem of engineering applicability. Based on the threshold screening of <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(D_{{\text{w}}}\)</EquationSource> </InlineEquation> < 0.51 s and <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\(T_{{\text{H}}}\)</EquationSource> </InlineEquation> < 1.31 s, the mean nonlinear displacement ratio of the 10 pulse-like ground motions in the original database within the period range of 0–6 s is lower than that of the 10 non-pulse round motions. It indicates that the new method can identify the pulse-like ground motion that is unfavorable to the structure more accurately. Conclusion The half wavelength width and energy-weighted marginal spectral period can effectively characterize the characteristics of low-frequency ground motion pulses. The time–frequency dual parameter joint screening method significantly improves the reliability of pulse-like ground motion identification, providing theoretical support for performance-based seismic design. Research has confirmed that the concentration of long-period pulse energy is a key inducement for the nonlinear displacement response of structures. The new method lays the foundation for the optimization of seismic analysis of near-fault engineering.

Identification of Pulse-Like Ground Motion Based on Pulse Waveform Anatomy

期刊论文
发表日期：

2025年10月31日

查看全文

Dengjia Fang,
Shengkui Di,
Yu Zhou

摘要

Purpose

The paper addresses the adverse effects of near-field pulse-like ground motion on long-period and seismic isolation structures, and solves the problems of overestimating/underestimating periods and relying on pulse model matching in traditional pulse period identification methods. A ground motion identification method combining time–frequency domain dual parameters is proposed to improve the accuracy of pulse period characterization and the reliability of structural nonlinear response evaluation.

Method

Based on 122 pulse-like and 10 non-pulse ground motion data in the NGA West2 database, the time-domain parameters-half wavelength width and its energy-weighted value \(D_{{\text{w}}}\) , as well as the frequency domain parameters-the energy-weighted average period \(T_{{\text{H}}}\) combining the Modified Ensemble Empirical Mode Decomposition (MEEMD) and Hilbert-Huang marginal spectrum are proposed. The modal aliasing was eliminated through MEEMD, and the effective components were screened by using permutation entropy. Combining half-wavelength energy weighting and marginal spectral energy distribution, a dual-parameter identification system of \(D_{{\text{w}}}\) and \(T_{{\text{H}}}\) was constructed, and the validity of the method was verified by the Bouc-Wen-Baber-Noori (BWBN) model.

Results

The \(D_{{\text{w}}}\) shows a strong positive correlation with seismic magnitude and traditional pulse period parameter T_p. \(T_{{\text{H}}}\) stabilizes the pulse period range to 1.14–3.98 s (while the traditional method is 0.4–13.12 s), solving the problem of engineering applicability. Based on the threshold screening of \(D_{{\text{w}}}\) < 0.51 s and \(T_{{\text{H}}}\) < 1.31 s, the mean nonlinear displacement ratio of the 10 pulse-like ground motions in the original database within the period range of 0–6 s is lower than that of the 10 non-pulse round motions. It indicates that the new method can identify the pulse-like ground motion that is unfavorable to the structure more accurately.

Conclusion

The half wavelength width and energy-weighted marginal spectral period can effectively characterize the characteristics of low-frequency ground motion pulses. The time–frequency dual parameter joint screening method significantly improves the reliability of pulse-like ground motion identification, providing theoretical support for performance-based seismic design. Research has confirmed that the concentration of long-period pulse energy is a key inducement for the nonlinear displacement response of structures. The new method lays the foundation for the optimization of seismic analysis of near-fault engineering.