Scholarlink

<p>Anti-slide piles (ASP) are critical for slope reinforcement, with design parameters significantly influencing stability. A three-dimensional numerical model for ASP-reinforced layered slopes was developed and validated against model test results. Numerical simulations generated a dataset, and a predictive model was constructed using Back Propagation Neural Network and Multi-expression Programming. A Genetic Algorithm optimized the ASP design by maximizing the safety factor, and the results were verified numerically. The Back Propagation Neural Network achieved 96.72% prediction accuracy, demonstrating its effectiveness in assessing reinforced slope stability. The optimized ASP parameters yielded a safety factor of 1.67, outperforming other designs with only a 3.45% deviation from numerical calculations. The combined Back Propagation Neural Network and Genetic Algorithm approach proved efficient for ASP design optimization.</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

Optimization Design of Anti-slide Pile Reinforcement for Bedding Slopes Using BP Neural Network and Genetic Algorithm

  • 期刊论文

  • 发表日期：

    2025年7月10日

查看全文
  • Shi Wang,
  • Xinyu Wang,
  • Yufang Zhang,
  • Junjie Wang

摘要

Anti-slide piles (ASP) are critical for slope reinforcement, with design parameters significantly influencing stability. A three-dimensional numerical model for ASP-reinforced layered slopes was developed and validated against model test results. Numerical simulations generated a dataset, and a predictive model was constructed using Back Propagation Neural Network and Multi-expression Programming. A Genetic Algorithm optimized the ASP design by maximizing the safety factor, and the results were verified numerically. The Back Propagation Neural Network achieved 96.72% prediction accuracy, demonstrating its effectiveness in assessing reinforced slope stability. The optimized ASP parameters yielded a safety factor of 1.67, outperforming other designs with only a 3.45% deviation from numerical calculations. The combined Back Propagation Neural Network and Genetic Algorithm approach proved efficient for ASP design optimization.