Matheuristics for a Parallel Machine Scheduling Problem with Non-Anticipatory Family Setup Times: Application in the Offshore Oil and Gas Industry

TL;DR

This paper introduces new matheuristics for a complex parallel machine scheduling problem with family setup times, applied to offshore oil logistics, achieving better solutions than existing methods on benchmark instances.

Contribution

It develops ILS and GRASP matheuristics with MIP-based neighborhoods for a novel offshore oil logistics scheduling problem, outperforming current approaches.

Findings

Achieved over 10% improvement in objective function for large instances.

Provided best solutions for all medium and large instances.

Demonstrated effectiveness of matheuristics over existing methods.

Abstract

In this paper, we address a variant of a batch scheduling problem with identical parallel machines and non-anticipatory family setup times to minimize the total weighted completion time. We developed an ILS and a GRASP matheuristics to solve the problem using a constructive heuristic and two MIP-based neighborhood searches, considering two batch scheduling mathematical formulations. The problem derives from a ship scheduling problem related to offshore oil & gas logistics, the Pipe Laying Support Vessel Scheduling Problem (PLSVSP). The developed methods overcome the current solution approaches in the PLSVSP literature, according to experiments carried out on a benchmark of 72 instances, with different sizes and characteristics, in terms of computational time and solution quality. New best solutions are provided for all medium and large-sized instances, achieving a reduction of more than 10% in the objective function of the best case.