Scholarlink

<p>This study introduced a novel integration of particle swarm optimization (PSO) with finite element modelling to optimize the design of uncased stone columns embedded in soft clay. The originality of the research lay in combining optimization and numerical modelling to systematically determine the most influential design parameters, namely column spacing and modulus of elasticity, in minimizing vertical displacement. A half-embankment model with a height of 2.6&#xa0;m was simulated over a 12&#xa0;m-thick clay layer reinforced with stone columns of varying stiffness and spacing, and subjected to incremental loading. The results showed that PSO identified a closer spacing combined with the highest modulus of elasticity as the optimal configuration, resulting in a 20.4% reduction in settlement. Column spacing emerged as the dominant factor influencing settlement behaviour. A multiple linear regression model was developed to generalize settlement prediction, demonstrating strong performance indices and good predictive capability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing Stone Column Spacing and Stiffness using Particle Swarm Optimization (PSO) for Finite Element Method (FEM) Based Settlement Prediction

  • 期刊论文

  • 发表日期：

    2026年4月2日

查看全文
  • Masyitah Md Nujid,
  • Duratul Ain Tholibon,
  • Muhammad Mukhlisin,
  • Yang Ruijuan,
  • Moh Muntaha,
  • Fitria Wahyuni

摘要

This study introduced a novel integration of particle swarm optimization (PSO) with finite element modelling to optimize the design of uncased stone columns embedded in soft clay. The originality of the research lay in combining optimization and numerical modelling to systematically determine the most influential design parameters, namely column spacing and modulus of elasticity, in minimizing vertical displacement. A half-embankment model with a height of 2.6 m was simulated over a 12 m-thick clay layer reinforced with stone columns of varying stiffness and spacing, and subjected to incremental loading. The results showed that PSO identified a closer spacing combined with the highest modulus of elasticity as the optimal configuration, resulting in a 20.4% reduction in settlement. Column spacing emerged as the dominant factor influencing settlement behaviour. A multiple linear regression model was developed to generalize settlement prediction, demonstrating strong performance indices and good predictive capability.