A GRASP+VND Hybrid Heuristic for the Multi-Load Berth Allocation Problem

Dry Bulk Terminals (DBT) handle and store dry bulk cargo, transported unpackaged in large quantities. Used worldwide, they facilitate the export of materials such as iron ore and grains. This study is motivated by a real case of one of the world's largest DBT, located in Brazil. We study the Berth Allocation Problem (BAP) in DBT, where berths operate different types of cargo at varying rates, with times depending on both the berth and the cargo. The BAP involves assigning a set of vessels to a berth layout within a defined time horizon. We present a Greedy Randomized Adaptive Search Procedure (GRASP) heuristic  and its hybridization with the Variable Neighborhood Descent (VND) algorithm, GRASP+VND. They solve instances generated from real data of the Port of Tubarão Complex. The results are compared against those generated by a mathematical model fot the problem, demonstrating the algorithms provide high-quality solutions.