Scholarlink

<p>This research delivers a comprehensive analysis of Artificial Intelligence (AI) applications for solving the Multi-Mode Resource-Constrained Multi-Project Scheduling Problem (MRCMPSP), one of the most complex optimization challenges in construction management. Using a unified framework, seven state-of-the-art multi-objective metaheuristic algorithms and the proposed Tournament Selection Multi-Objective Giant Pacific Octopus Optimizer (MOGPOO-TS)—were applied to multi construction projects to determine optimal start–finish times while balancing stringent resource, time, cost, and quality constraints. MOGPOO-TS was specifically designed to explore vast, high-dimensional solution spaces, maintain diversity along the Pareto front, and converge quickly to high-quality schedules. Across all evaluation criteria, MOGPOO consistently outperformed the other six algorithms, demonstrating superior trade-offs among time, cost and quality as well as remarkable stability under complex multi-project constraints. This highlights MOGPOO-TS’s pivotal role as an AI-enhanced benchmark for accelerating and refining multi-project construction scheduling, bridging the gap between single-project and enterprise-wide optimization and setting a new standard for practical operations research in the construction sector.</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

Multi-project scheduling optimization with artificial intelligence: a novel metaheuristic framework

  • 期刊论文

  • 发表日期：

    2025年12月29日

查看全文
  • Pham Vu Hong Son,
  • Luu Ngoc Quynh Khoi,
  • Luu Xuan Loc

摘要

This research delivers a comprehensive analysis of Artificial Intelligence (AI) applications for solving the Multi-Mode Resource-Constrained Multi-Project Scheduling Problem (MRCMPSP), one of the most complex optimization challenges in construction management. Using a unified framework, seven state-of-the-art multi-objective metaheuristic algorithms and the proposed Tournament Selection Multi-Objective Giant Pacific Octopus Optimizer (MOGPOO-TS)—were applied to multi construction projects to determine optimal start–finish times while balancing stringent resource, time, cost, and quality constraints. MOGPOO-TS was specifically designed to explore vast, high-dimensional solution spaces, maintain diversity along the Pareto front, and converge quickly to high-quality schedules. Across all evaluation criteria, MOGPOO consistently outperformed the other six algorithms, demonstrating superior trade-offs among time, cost and quality as well as remarkable stability under complex multi-project constraints. This highlights MOGPOO-TS’s pivotal role as an AI-enhanced benchmark for accelerating and refining multi-project construction scheduling, bridging the gap between single-project and enterprise-wide optimization and setting a new standard for practical operations research in the construction sector.