A Rolling Horizon-Based Constraint Programming Method for Continuous Berth Allocation Problem

As a specific area or facility for berthing and loading and unloading cargo, berths have an important impact on the throughput of a terminal. Efficient Continuous Berth Allocation (CBA) makes optimal use of berth space to reduce the port staying time of vessels. In this paper a rolling horizon-based Constraint Programming (CP) approach is proposed to solve the CBA problem. To comprehensively model the process of allocating berth to vessels, the declarative modeling method, constraint programming is employed to formulate the constraints in the CBA problem. Additionally, a rolling horizon-based scheduling framework is proposed to tackle the uncertainty caused by the dynamic arrival of vessels. Benefiting from the solving efficiency of the CP method, the subproblems obtained through rolling horizon decomposition can be solved in seconds. As the experimental results show, the proposed method can find the optimal result for each subproblem and eligible scheduling for a one-month CBA problem. Compared with the historical operation of a real terminal, the staying time of vessels is remarkably reduced.