Application of metaheuristics algorithm on a multi-objective container loading problem considering container's utilization and vehicle's balance

R.J. Kuo,Pei-Cheng Ho,Ferani E. Zulvia
DOI: https://doi.org/10.1016/j.asoc.2023.110417
2023-05-20
Applied Soft Computing Journal
Abstract:The container loading problem (CLP) intends to arrange a set of boxes into a container by considering different aspects, such as the dimension and weight of the box, capacity of the container, etc. In the multi-destination delivery, box arrangement should also consider routing, or delivery order. In addition, it is crucial to balance the boxes for safety in order to avoid traffic accident. Therefore, in order to overcome the aforementioned problems, this study proposes a mathematical model with two objectives, maximizing container's utilization and minimizing truck's unbalance force penalty. The proposed model also considers some practical constraints including geometry, maximum weight, and delivery order, or loading priority. Two metaheuristics, non-dominated sorting genetic algorithm II (NSGA II) and multi-objective particle swarm optimization algorithm (MOPSO) are employed to solve the proposed model and verified by using seven benchmark datasets in terms of hypervolume and spacing. The experimental results show that both algorithms have their own merits.
What problem does this paper attempt to address?