基于变分模态分解方法的近断层速度脉冲参数化研究OA

Parameterization of the velocity pulses in near-fault ground motions is crucial for refined seismic in-put and quantitative assessment of the dynamic response in earthquake engineering.Most previous studies fo-cus on the quantitative identification of pulse-like features,with limited attention paid on the parameterization of the velocity pulses of interest.In this study,we propose a progressive iterative method based on the varia-tional mode decomposition(VMD),enabling a fully automated parameterization of the velocity pulse.Firstly,the original velocity time history is decomposed using the VMD,and a filtered signal is obtained by signal reconstruction using the intrinsic mode functions.Subsequently,a pulse model is employed to perform least-squares fitting on the filtered signal,progressively determining the physical characteristic parameters of the velocity pulse,e.g.,the pulse period,the number of cycles,and the pulse phase.Finally,the proposed method was applied to analyze the ground motion data from the 2023 Jishishan earthquake in Gansu Province.Four velocity pulse-like ground motions with significant pulse attributes were successfully identified.The pulse periods ranged from 0.3 s to 1.1 s,the number of cycles ranged from 1.0 to 1.8,and the phase angles varied between-95.15° and 84.80°,revealing the diversity of velocity pulse waveforms in this earthquake event.The pulse periods identified by the proposed method show strong agreement with those obtained using Baker's method,demonstrating its effectiveness.Notably,the pulse energy ratio at station GS_N0052 reaches as high as 0.91,further highlighting the prominence of pulse-like effects in ground motions within the near-fault region.The proposed method enables accurate characterization of the physical attributes of velocity pulses,providing essential data support for the analysis of near-fault geohazards and the seismic response of en-gineering structures.