Modeling and solving the joint berth allocation and vessel sequencing problem with speed optimization in a busy seaport

Vessel sequencing, speed optimization, and berth allocation comprise the primary interventions for servicing calling vessels in a busy seaport. The objectives are to minimize vessel completion time and reduce carbon emissions, thus balancing port service efficiency with environmental sustainability. Despite interdependent, these challenges have often been addressed in isolation, leading to sub-optimal or even infeasible solutions for vessel services. In this paper, we propose a bi-objective mixed-integer linear programming model that jointly optimizes the allocation of vessels to berths, as well as the sequencing and sailing speeds of vessels within the channel. To solve this model, we develop a tailored non-dominated sorting genetic algorithm incorporating reinforcement learning. Several efficient methods are presented to improve the performance of the developed algorithm. We also introduce a new relative distance-based metric to evaluate Pareto solutions. Extensive computational experiments on Jingtang Port, China, show that our algorithm outperforms the benchmark algorithms from the literature, yielding far superior solutions in shorter computational times. Various Pareto solutions are provided, based on which trade-offs between service efficiency and environmental sustainability are analyzed and some managerial insights are outlined.