Multi-population for Multi-objective Genetic Algorithm with Adaptive Information Sharing Strategy for Berth Allocation and Quay Crane Assignment Problems

For the freight transportation industry, the method of allocating container terminal berths and shore bridges is particularly important. Existing berth and bridge allocation methods mainly rely on manual methods, which are usually difficult to deal with large-scale schedule problems and face high cost and inefficiency problems, so there is an urgent need to design an efficient berth and bridge allocation method. Berths and quay cranes are scarce resources for container terminals and a sound scheduling plan is conducive to improving port mobility. This paper focuses on the integrated Berth and Quay Crane Assignment (BQCA) problem under continuous and dynamic conditions, to minimize the total waiting time and deviation distances from the preferred berth of ships. A Multi-Population for Multi-Objective Genetic Algorithm with Adaptive Information Sharing (MPMOGA-AIS) is proposed for solving the BQCA problem, which includes a Hybrid Heuristic Initialization (HHI) strategy and an Adaptive Information Sharing (AIS) strategy utilizing local and global information. The proposed MPMOGA-AIS is tested on the latest dataset from Huawei, and the experimental results are compared with the port result and NSGAII, which shows that our MPMOGA-AIS algorithm provides better feasible solutions for BQCA problems.