Scholarlink

<p>This paper deals with a comprehensive study of the suitability of optimized machine learning models via particle swarm optimization (PSO) to predict the shear strength of ultra-high-performance fiber-reinforced concrete (UHPFRCs) at various applied strain rates. Firstly, a dataset comprising 144 experimental records with nine input variables is considered to train and test model. The hybrid machine learning model, namely RF-PSO, was proposed for anticipating the shear strength of UHPFRCs. The results showed that the hybrid RF-PSO model demonstrated highest precision with coefficients of determination (R<sup>2</sup>) reaching 0.992 and 0.990 at both training and testing, respectively. Moreover, the effect of input variables on the shear strength of UHPFRCs was explored via sensitivity analysis. It was found that the fiber index is most important variable affected to the shear strength of UHPFRCs. The critical value of fiber index is revealed at 100%. The shear strength increased with an increasing fiber index and strain rates.</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

Harnessing hybrid machine learning technique for shear resistance prediction of ultra-high-performance fiber-reinforced concrete at high strain rates

  • 期刊论文

  • 发表日期：

    2025年10月30日

查看全文
  • Nguyen Thien Phuc

摘要

This paper deals with a comprehensive study of the suitability of optimized machine learning models via particle swarm optimization (PSO) to predict the shear strength of ultra-high-performance fiber-reinforced concrete (UHPFRCs) at various applied strain rates. Firstly, a dataset comprising 144 experimental records with nine input variables is considered to train and test model. The hybrid machine learning model, namely RF-PSO, was proposed for anticipating the shear strength of UHPFRCs. The results showed that the hybrid RF-PSO model demonstrated highest precision with coefficients of determination (R2) reaching 0.992 and 0.990 at both training and testing, respectively. Moreover, the effect of input variables on the shear strength of UHPFRCs was explored via sensitivity analysis. It was found that the fiber index is most important variable affected to the shear strength of UHPFRCs. The critical value of fiber index is revealed at 100%. The shear strength increased with an increasing fiber index and strain rates.