Cost optimization of repetitive project scheduling through a constraint programming-based relax-and-solve algorithm

Abstract

This paper focuses on the cost minimization of the multi-mode resource-constrained repetitive project scheduling problem with multiple crews, crew interruptions, and soft logic. The resource allocation of each crew is considered. To explore the impact of different construction strategies on project costs, mixed-integer linear programming (MILP) and constraint programming (CP) models are developed representing different construction scenarios. A relax-and-solve (R&S) algorithm, incorporating a rolling horizon and constraint programming, is proposed to obtain near-optimal solutions within reasonable time limits. The case study reveals that considering crew resource allocation and adopting more flexible construction strategies can contribute to reducing total project costs. The findings provide construction managers with practical strategies to improve scheduling, resource management, and cost control. Meanwhile, the proposed algorithm performs competitively compared with MILP and CP models, which inspires future research to apply this algorithm to other repetitive project scheduling problems.