集装箱码头堆场龙门吊调度强化学习法

his paper studies the problem of scheduling a yard crane to perform a given set of loading/unloading jobs with different ready times. The objective is to minimize the sum of job waiting times. Meanwhile a Q-learning based algorithm is designed to develop a decision-making policy on selecting the appropriate dispatching rule at different states. Also, the settings used to apply Q-learning algorithms to the yard crane scheduling, such as states, action, learning step size, and discount rate etc are presented. Lastly, simulation studies are performed to investigate the effect of the proposed approaches. The results show that the Q-learning algorithm can performs better than other dispatching rules such as FCFS(first come first served), UN(uni-directional travel), and NT(nearest truck first served) when the time between yard trailer arrivals to the container yard is long and the distribution of containers’ arrival locations deviates from a uniform distribution.