Research on hybrid algorithms for optimizing construction site supply chain management: Experimental application for high-rise buildings in Vietnam

Abstract

Effective inventory management on construction sites significantly influences both project costs and timelines. However, optimizing material ordering strategies remains challenging due to the conflict between cost and computational complexity. This study addresses inventory optimization in a high-rise building project in Hanoi, Vietnam, using three approaches: the exact method (Exact) with mixed integer linear programming (MILP), genetic algorithm (GA), and a hybrid method (Hybrid) combining GA and MILP. These methods are applied to a 25-story building with a 24-week superstructure construction schedule, considering key materials like cement, steel, bricks, sand, and stone. The results indicate that the Exact method minimizes inventory costs but requires a lengthy computation time, making it impractical for larger projects. While the GA algorithm is faster, it produces less optimal results. The Hybrid method, combining GA and MILP, strikes an effective balance between speed and solution quality, making it highly suitable for construction management where time constraints exist. This study highlights the practical value of the Hybrid method in inventory management and paves the way for future research into real-time decision-making tools for project management.