This study addresses the multi-project scheduling problem with resource transfer (MPSP-RT), a complex NP-hard problem involving inter-project dependencies, shared resources, local constraints, and transfer-related time–cost interactions. It aims to develop a robust and scalable multi-objective optimization framework for improving scheduling efficiency and decision-making in multi-project environments.
An enhanced multi-objective Giant Pacific Octopus Optimizer (MOGPOO) is proposed, integrating a strength Pareto-based archive management mechanism to improve dominance evaluation and maintain solution diversity. Transfer-aware encoding and constraint-handling strategies are incorporated to effectively coordinate resource allocation and inter-project labor transfers while optimizing makespan, cost, and delay-related indicators.
Results from computational experiments demonstrate that MOGPOO significantly outperforms NSGA-II and MOSMA across key performance metrics. The proposed method reduces project completion time, lowers overall costs, and ensures more stable resource allocation. Delay indicators are also substantially improved, reflecting enhanced schedule reliability. Statistical tests confirm that these improvements are significant at 95% confidence level, indicating strong robustness and consistency. These results emphasize the effectiveness of integrating transfer-aware optimization with advanced Pareto-based mechanisms.
This study contributes to the advancement of multi-objective optimization by introducing a structurally enhanced MOGPOO framework that improves convergence stability, scalability, and decision quality in resource-transfer-aware environments. From a practical perspective, the proposed model provides a reliable decision-support tool for optimizing resource coordination, reducing delays, and improving overall performance in complex construction project portfolios.
