Exact methods for the quay crane scheduling problem when tasks are modeled at the single container level

• Published in: Computers & operations research Vol. 99; pp. 218 - 233
• Main Authors: Msakni, Mohamed Kais, Diabat, Ali, Rabadi, Ghaith, Al-Salem, Mohamed, Kotachi, Mariam
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 01.11.2018; Pergamon Press Inc
• Subjects: Central processing units; Constraint modelling; Container terminals; Containers; CPUs; Cranes; Integer programming; Mathematical models; Mixed-integer programming; Operations research; Quay crane scheduling problem; Safety margins; Scheduling; Search algorithms; Studies; Task scheduling; Transport buildings, stations and terminals; Quay crane scheduling problem; Scheduling; Container terminals; Mixed-integer programming
• ISSN: 0305-0548; 1873-765X; 0305-0548
• DOI: 10.1016/j.cor.2018.07.005

•The scheduling problem of quay cranes in container terminal is studied.•Different technical constraints related to the quay cranes are considered.•Two exact methods are proposed to provide optimal schedules.•A construction heuristic is developed to provide near-optimal solutions.•Computational experiments show the efficiency of the proposed methods. The scheduling of quay cranes (QCs) to minimize the handling time of a berthed vessel is one of the most important operations in container terminals as it impacts the terminal’s overall productivity. In this paper, we propose two exact methods to solve the quay crane scheduling problem (QCSP) where a task is defined as handling a single container and subject to different technical constraints including QCs’ safety margin, non-crossing, initial position, and nonzero traveling time. The first method is based on two versions of a compact mixed-integer programming formulation that can solve large problem instances using a general purpose solver. The second is a combination of some constraints of the proposed mathematical model and the binary search algorithm to reduce the CPU time, and solve more efficiently large-sized problems. Unlike existing studies, the computational study demonstrates that both methods can reach optimal solutions for large-sized instances and validates their dominance compared to an exact model proposed in the literature which finds solutions only for small problems.