Scholarlink

<p>Construction project planning inherently involves complex decision-making due to the presence of multiple conflicting objectives, including project duration, cost efficiency, quality performance, and safety risk. This study presents an artificial intelligence (AI)-based optimization framework for solving many-objective construction time–cost–quality–safety risk trade-off optimization problems (TCQSRTP) using a 37-activity project as a case study. Each activity is modeled with multiple execution modes, characterized by distinct durations, costs, quality levels, and safety risk indices, thereby forming a large-scale combinatorial optimization problem. To efficiently explore the search space and enhance solution quality, an improved metaheuristic algorithm based on the Opposition-Based Learning Teaching–Learning-Based Optimization (OBL–TLBO) algorithm is proposed. The TLBO algorithm is a parameter-less optimization technique that updates candidate solutions through teacher and learner phases, ensuring a balance between exploration and exploitation. To further improve convergence speed and population diversity, opposition-based learning (OBL) is incorporated, enabling simultaneous consideration of candidate solutions and their corresponding opposite solutions within the search space. This mechanism increases the probability of locating regions closer to the global optimum and helps avoid premature convergence. The proposed framework integrates the Critical Path Method (CPM) for accurate project scheduling and adopts a Pareto-based multi-objective optimization approach to identify non-dominated solutions. The model simultaneously minimizes project duration, total cost, and safety risk while maximizing overall quality performance. The results demonstrate that the proposed OBL–TLBO approach produces a well-distributed and diverse Pareto front, offering a wide range of optimal trade-off solutions for decision-makers. Comparative analysis reveals that the proposed method outperforms conventional algorithms in terms of convergence speed, solution diversity, and robustness. The findings highlight the effectiveness of integrating opposition-based learning with TLBO as a powerful decision-support tool for enhancing sustainability, safety, and overall performance in complex construction project planning.</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

Utilizing artificial intelligence for solving many-objective construction time–cost–quality–safety risk trade-off optimization problems

  • 期刊论文

  • 发表日期：

    2026年4月16日

查看全文
  • Vishnu Lakkamraju,
  • Naveen Tewari,
  • Gyana Ranjana Panigrahi,
  • Suvendu Kumar Nayak,
  • Akanksha Kulkarni,
  • Mohammad Zaki Mortazavi

摘要

Construction project planning inherently involves complex decision-making due to the presence of multiple conflicting objectives, including project duration, cost efficiency, quality performance, and safety risk. This study presents an artificial intelligence (AI)-based optimization framework for solving many-objective construction time–cost–quality–safety risk trade-off optimization problems (TCQSRTP) using a 37-activity project as a case study. Each activity is modeled with multiple execution modes, characterized by distinct durations, costs, quality levels, and safety risk indices, thereby forming a large-scale combinatorial optimization problem. To efficiently explore the search space and enhance solution quality, an improved metaheuristic algorithm based on the Opposition-Based Learning Teaching–Learning-Based Optimization (OBL–TLBO) algorithm is proposed. The TLBO algorithm is a parameter-less optimization technique that updates candidate solutions through teacher and learner phases, ensuring a balance between exploration and exploitation. To further improve convergence speed and population diversity, opposition-based learning (OBL) is incorporated, enabling simultaneous consideration of candidate solutions and their corresponding opposite solutions within the search space. This mechanism increases the probability of locating regions closer to the global optimum and helps avoid premature convergence. The proposed framework integrates the Critical Path Method (CPM) for accurate project scheduling and adopts a Pareto-based multi-objective optimization approach to identify non-dominated solutions. The model simultaneously minimizes project duration, total cost, and safety risk while maximizing overall quality performance. The results demonstrate that the proposed OBL–TLBO approach produces a well-distributed and diverse Pareto front, offering a wide range of optimal trade-off solutions for decision-makers. Comparative analysis reveals that the proposed method outperforms conventional algorithms in terms of convergence speed, solution diversity, and robustness. The findings highlight the effectiveness of integrating opposition-based learning with TLBO as a powerful decision-support tool for enhancing sustainability, safety, and overall performance in complex construction project planning.