A hybrid heuristic for the multi-product straight pipeline scheduling problem
Résumé
In this paper, we propose a hybrid approach for the multi-product straight pipeline scheduling problem. The straight pipeline is shared for the transport of the different quality of phosphate slurries from the pumping station to the distribution center. The objective is to schedule a sequence of batches that ensures all daily product demand with a minimum total cost, while respecting the different constraints. The proposed approach is composed of a construction heuristic and a MILP model as post-optimization procedure. The construction method uses a strategy maximizing the volume of batches to favor generating a solution that satisfies all of daily demands. A MILP model is used as post-optimization procedure to adjust the volume of batches. The approach was tested on an instance from an operational pipeline on a one-month time horizon.