Escalonamento de projetos com restrições de recursos: modelagem matemática e aplicações práticas na gestão de projetos

Abstract

Project scheduling under resource constraints is a recurrent challenge in contemporary project management, demanding effective methods to simultaneously optimize time, cost, and quality in complex environments. Traditional techniques such as PERT (Program Evaluation and Review Technique) and CPM (Critical Path Method) show significant limitations in handling multiple execution modes and dynamic variations in resource availability. In this context, this study investigates how mathematical modeling combined with advanced metaheuristics can enhance project scheduling, enabling a more efficient and adaptive allocation of available resources. The methodology adopted was a systematic literature review, covering publications from the past ten years in prominent scientific databases such as CAPES, SciELO, and IEEE Xplore. Relevant articles addressing the classical Resource- Constrained Project Scheduling Problem (RCPSP) were analyzed, with special attention to heuristic and metaheuristic techniques, including genetic algorithms, tabu search, and particle swarm optimization (PSO). The findings indicate that hybrid approaches, particularly those integrating evolutionary algorithms with local search techniques, consistently outperform traditional methods, especially regarding flexibility and adaptability to unforeseen project changes. Additionally, emerging trends identified in the literature suggest that integrating machine learning techniques could potentially reduce computational complexity and enhance the effectiveness of the employed models. The study concludes that combining mathematical modeling with advanced heuristics represents a significant advancement, not only in optimizing resource allocation but also in fostering innovative computational strategies, particularly applicable to sectors such as construction, manufacturing, and information technology.