Metaheuristic optimization for the Single-Item Dynamic Lot Sizing problem with returns and remanufacturing

The use of metaheuristics for solving the Single-Item Dynamic Lot Sizing problem with returns and remanufacturing has increasingly gained research interest. Recently, preliminary experiments with Particle Swarm Optimization revealed that population-based algorithms can be competitive with existing state-of-the-art approaches. In the current work, we thoroughly investigate the performance of a very popular population-based algorithm, namely Differential Evolution (DE), on the specific problem. The most promising variant of the algorithm is experimentally identified and properly modified to further enhance its performance. Also, necessary modifications in the formulation of the corresponding optimization problem are introduced. The algorithm is applied on an abundant test suite employed in previous studies. Its performance is analyzed and compared with a state-of-the-art approach as well as with a previously investigated metaheuristic algorithm. The results suggest that specific DE variants can be placed among the most efficient approaches, thereby enriching the available algorithmic artillery for tackling the specific type of problems.