Genetic Algorithms to Solve Resource-Constrained Project Scheduling Problems with Variable Activity Durations

Resource-Constrained Project Scheduling Problem (RCPSP) is one of the most popular problems in the scheduling phase of any project. This paper tackles the RCPSP in which activity durations can vary within their certain ranges such as RCPSP with variable activity durations. In this paper, we have attempted to find the most suitable hybridization of GA variants to solve the mentioned problem. For this reason, three GA variants (Standard GA, Stud GA and Jumping Gene) were utilized for first GA, and two GA variants (Standard GA, Stud GA) for the second one, and their hybridizations were compared. For this purpose, several comparisons of the following hybridizations of GAs are performed: Standard-Standard GA, Standard-Stud GA, Stud-Standard GA, Stud-Stud GA, Jumping Gene-Standard GA, and Jumping Gene-Stud GA. Simulation results show that implementing Stud-Stud GA hybridization to solve this problem will cause convergence on the minimum project makespan, faster and more accurate than other hybrids. The robustness of the Stud GA in solving the well-known benchmarking RCPSP problems with deterministic activity durations is also analyzed.