Multi-Operator based Genetic Algorithm for Resource Constrained Project Scheduling

Abstract

Solving Resource constrained project scheduling problem (RCPSP) is a significant research topic because of its importance in theory and practice. Over the last few decades, many different approaches have been proposed for solving RCPSPs. Among them, evolutionary computation based approaches are popular. However, these approaches do not perform consistently over all types of problems because the algorithms are usually designed targeting certain type of problems and the choices of algorithmic parameters are difficult. T o address these issues, we propose a multi-operator based Genetic Algorithm (GA) for solving RCPSPs. Here, in selecting the operators, we develop a self-adaptive mechanism that helps to apply the best performing operator with a higher probability. A local search is applied to refine t he solution, a nd a n automatic restart strategy i s used to diversify the population as needed. The performance of the proposed algorithm is evaluated by solving a wide variety of test problems. The experimental results show that the proposed method delivers high-quality solutions on a lower computational budget than the existing algorithms.