A genetic algorithm with resource buffers for the resource-constrained multi-project scheduling problem

In this study, we compose a new metaheuristic algorithm for solving the resource-constrained multi-project scheduling problem. Our approach is based on a general metaheuristic strategy which incorporates two resource-buffered scheduling tactics. We build on the most effective evolutionary operators and other well-known scheduling methods to create a novel genetic algorithm with resource buffers. We test our algorithm on a large benchmark dataset and compare its performance to ten existing metaheuristic algorithms. Our results show that our algorithm can generate new best-known solutions for about 20% of the test instances, depending on the optimisation criterion and due date. In some cases, our algorithm outperforms all other available methods combined. Finally, we introduce a new schedule metric that can quantitatively measure the dominant structure of a solution, and use it to analyse the differences between the best solutions for different objectives, due dates, and instance parameters.