针对停机位分配问题展开系统研究，目标是最小化远机位分配数量以及近机位空闲时间，针对其多目标多约束特性，提出以最小远机位分配数量和最小近机位空闲时间为目标的多目标数学模型，该模型考虑了航班进出港实际时间、机型类别及停机位间相互关系等参数。结合深度强化学习方法，特别是深度确定性策略梯度算法（deep deterministic policy gradient, DDPG），对停机位分配过程进行优化。为提升算法的寻优能力与性能，设计了改进后的DDPG算法，融入优先级经验回放和多策略融合探索机制。通过对比实验，表明改进后的算法更优，显著减少了最小远机位分配数量并优化了近机位空闲时间，且收敛更快、全局寻优能力更强，充分证实了其有效性。

A systematic study was conducted on the issue of gate assignment, with the goal of minimizing the number of remote gate assignments and the idle time of near gates. A multi-objective mathematical model was proposed to address the multi-objective and multi-constraint characteristics of the problem. The model was designed to minimize the number of remote gate assignments and the idle time of near gates while taking into account parameters such as actual flight arrival and departure times, aircraft types, and the interrelationships among gates. The gate assignment process was optimized using the deep reinforcement learning method, specifically the deep deterministic policy gradient(DDPG) algorithm. To enhance the optimization ability and performance of the algorithm, an improved DDPG algorithm was developed by incorporating prioritized experience replay and multi-strategy exploration mechanisms. Comparative experiments were conducted, and the results show that the improved algorithm significantly reduces the number of remote gate assignments and optimized time utilization. The algorithm also achieves faster convergence and stronger global optimization capabilities, confirming its effectiveness.