A Twofold Constructive Genetic Algorithm for Resource-Constrained Multi-Project Scheduling Problem (RCMPSP)

Abstract

Resource-constrained multi-project scheduling problem (RCMPSP) arises in many project-based organizations, including construction and civil engineering companies. Numerous heuristic and meta-heuristic approaches have been proposed for a project scheduling problem with limited resources. In this paper, a twofold constructive genetic algorithm is proposed for the resource-constraint multiple projects scheduling problem, which benefits from a number of priority and several auxiliary rules which are fed into a serial schedule generation scheme (SGS), where auxiliary rules are used to break the tie situations where several activities have equal priority values. Numerical standard problems in different sizes are retrieved from the multi-project scheduling problem LIBRARY (MPSPLIB) website, and the numerical results are analyzed in different scenarios. Then, the genetic algorithm is used to improve the results where its parameters are tuned via Taguchi design of experiments (DOE). The results of this study showed that the performance of the proposed approach has significantly improved the solution of several problem instances and registered in the MPSPLIB.