Digital Modeling of Yard Container Characteristics Using 3D-CNN for Enhanced Terminal Operation

With the rapid development of container transportation worldwide, container terminals have become critical nodes connecting international logistics networks. However, the increasing volume of containers at these terminals has led to numerous challenges in traditional port operations. One significant issue is the inability to verify the quality of container intake during yard operations. Additionally, central control dispatch struggles to fully grasp the characteristics of containers in the yard, hindering effective operational guidance. To address these challenges and enhance operational efficiency, this paper proposes a data modeling approach for yard characteristics. Through a data-driven method, complex container features in the yard are abstracted into a matrix, establishing a digital model of yard container characteristics. Subsequently, a Three-Dimensional Convolutional Neural Network (3D-CNN) is employed to predict the loading efficiency of ships. Numerical experiments demonstrate that the model can deeply analyze container features, effectively predict ship loading efficiency, and provide guidance for terminal control and scheduling operations, ultimately enhancing terminal efficiency.