A digital twin-based approach for optimizing operation energy consumption at automated container terminals

The sustainable development of port operation management is strongly related to the energy consumption of production at automated container terminals (ACTs). This paper focuses on the production activities at a container yard, which is the primary facility of ACTs. A digital twin-based approach is proposed to optimize the operation of an automatic stacking crane (ASC) handling containers in terms of energy consumption. A virtual container yard that syncs with a physical container yard in the ACT digital twin system for observation and validation is developed. A mathematical model is established to minimize the total energy consumption of completing all tasks. Then, the Q-learning algorithm is adapted to optimize a solution based on the operating data from the ACT digital twin system. Numerical experiments are conducted to demonstrate the effectiveness of the proposed approach by comparing it with two other solution algorithms, viz., genetic algorithm (GA) and particle swarm optimization (PSO). The total energy consumption of two operation strategies (i.e., centralized and decentralized) are also compared using the proposed digital twin-based approach. With digital twin, the operational environment and energy consumption are visualized to support optimization and management of ASCs. Managers and operators can choose an appropriate strategy according to the designated sustainable goal.