An efficient method for simulating uni- and bi-directional spectrum-compatible ground motions using particle swarm optimization

Qinhao Gao*, Giorgio Monti, Zhiwang Chang*, Zhi Li, Fabrizio Mollaioli

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

Spectrum-compatible seismic records are widely recommended in modern seismic codes for assessing the seismic performance of both new constructions and pre-existing buildings. This study introduces a novel method to generate uni- and bi- directional spectrum-compatible ground motions utilizing particle swarm optimization (PSO). Initially, seed records are transformed via continuous wavelet transform (CWT) to extract wavelet coefficients. Subsequently, the optimal wavelet coefficients for producing spectrum-compatible components will be obtained through PSO with a dynamic solution space determined according to the physical relationship between the time-frequency characteristics of each ground motion and the response spectrum. Moreover, the normalized radial spectral acceleration pattern (RadSAP) effectively captures the orientation variation characteristics of bidirectional records. The challenge of maintaining RadSAPs while generating spectrum-compatible ground motions adds complexity to this research. Consequently, this paper also details the methodology for generating bidirectional spectrum-compatible ground motions while preserving the RadSAPs. The effectiveness of the proposed approach is demonstrated through the designed spectrum, site-specific target spectrum and additional metrics (e.g. intensity, frequency and duration).

Original languageEnglish
Article number109399
JournalSoil Dynamics and Earthquake Engineering
Volume195
DOIs
Publication statusPublished - Aug 2025

Keywords

  • Continuous wavelet transform
  • Particle swarm optimization
  • Simulation of ground motions
  • Spectrum-compatible
  • uni- and bi-directional records

Fingerprint

Dive into the research topics of 'An efficient method for simulating uni- and bi-directional spectrum-compatible ground motions using particle swarm optimization'. Together they form a unique fingerprint.

Cite this