The Seaport Traffic Scheduling Problem: Formulations and a Column-Row Generation Algorithm

Abstract

Traffic congestion in seaports can lead to uncontrolled execution of berth plans, causing severe vessel delays and long waiting times. In this paper, we schedule the vessel traffic of a seaport by optimizing the utilization of the navigation channel and the anchorage areas in the terminal basin. We model the traffic scheduling problem as a mixed integer linear program (MILP), which takes as inputs the physical layout of a seaport and the berth plans devised by the terminal operators, to minimize the berthing and departure delays of vessels and the number of vessels that cannot berth or depart successfully. We reformulate the MILP into a more compact set-partitioning formulation, develop a column generation algorithm for solving the problem, and propose three performance enhancement strategies, namely, a row generation method, a pricing problem reduction method, and a heuristic pricing method, to accelerate the convergence of the column generation algorithm. Computational performance of the proposed algorithm is evaluated on benchmark instances that are generated from the physical layout and operational data of the Yangshan Deep-water Port in Shanghai.