Integrated green scheduling of quay cranes, internal vehicles, and yard cranes with uncertainties via learning-assisted evolutionary algorithms

Nowadays, maritime trade becomes the mainstream mode of commerce and transportation, and port operation plays important roles in improving the efficiency of global trades. In recent years, many studies focus on scheduling handling equipment in container terminals. Nevertheless, rare attention is given to achieving both economic and environment sustainability in coping with integrated scheduling of handling equipment in uncertain environments. This work aims to address an integrated green scheduling of quay cranes, internal vehicles and yard cranes with random operation times and movement speeds to minimize maximum completion time and total energy consumption. Firstly, a stochastic programming model is formulated to define the problem mathematically. Secondly, a reinforcement learning-assisted evolutionary algorithm is developed to solve the formulized model. In particular, crossover, mutation and local search methods in consideration of the model’s features are specially designed, and they are formed as three search combinations which are intelligently chosen for performing by reinforcement learning methods. Finally, the developed algorithm is verified by making comparisons with four renowned meta-heuristics on a set of test instances. The results demonstrate that the designed methods have obvious competitiveness in comparison with the rivals for handling the investigated problem.