Simulation-based optimization of yard slot allocation in U-shaped container terminals

The yard slot allocation problem (SAP), which concerns locating containers in the storage yard, could critically affect the performance of ports. The optimization of this problem is challenging due to the complex operational conditions and real-time decision requirement in practice. As a new type of layout, the U-shaped layout offers external and internal trucks (ETs and ITs) novel combinations of travel routes and container handover points that may result in unique characteristics for the SAP. This study addresses the SAP under the U-shaped layout to minimize the delay time of ITs and ETs. A novel simulation-based evaluation method considering multiple criteria is proposed to allocate slots for arriving containers. In this method, an evolving neural decision network (ENDN) is developed to explore the influence of real-time information on the weights of these criteria. We develop an efficient genetic algorithm tailored to optimize the parameters of the ENDN. A simulation model is developed to evaluate the algorithm’s performance under realistic operational uncertainties that may promote the practical implementation of the ENDN. The experimental results demonstrate that our method can determine slot allocations of shorter total vehicle delay time compared with existing methods.