影响因子：6.5
DOI码：10.1016/j.cie.2026.112006
所属单位：哈工程
发表刊物：Computers & Industrial Engineering
刊物所在地：美国
项目来源：Key Research and Development Program of Hainan Province (ZDYF2023GXJS017); National Natural Science
关键字：Berth and marine experiment allocation; Uncertainty; Stochastic programming; Reinforcement learning; Weighted average algorithm
摘要：The cargo and research port (CR_Port) is a multifunctional facility that integrates cargo transportation, research vessel docking, and marine experiments. The inherent complexities and potential uncertainties associated with these diverse operations present significant challenges for port scheduling. To address these challenges, this study introduces differentiated operational zones and designated time windows for marine experiments. Furthermore, a stochastic programming model for berth and experiment allocation (SPBAE) is developed, aiming to minimize vessel waiting times and departure delays, while accommodating the constraints related to vessels, experiments, and berths. To improve the tractability of the SPBAE model, a sample average approximation (SAA) method is applied, transforming the stochastic model into a deterministic version (DBAE). Subsequently, an enhanced weighted average algorithm (PCBWAA) is proposed to solve the DBAE model. This algorithm integrates crossover strategies, the bottom-left algorithm, and proximal policy optimization techniques. Following this, a novel solution approach, SPBAE_PCBWAA, is introduced by combining the SPBAE model with the PCBWAA algorithm to address the scheduling problem. Numerical experiments using data from a CR_Port in southern China are conducted to assess the performance of the proposed model and solution method. The results indicate that, across 60 test cases of varying scales, the SPBAE model outperforms the DBAE model, achieving an average improvement of 26.13%. Furthermore, as port congestion and uncertainty levels increase, the performance improvement of the SPBAE model relative to the DBAE model becomes more significant. Finally, the PCBWAA algorithm demonstrates superior robustness and greater suitability for solving the SPBAE model compared to other comparison algorithms selected in this paper.
备注：Key Research and Development Program of Hainan Province (ZDYF2023GXJS017); National Natural Science Foundation of China (No. 52371315); and Natural Science Foundation of Hainan Province (525MS110).
合写作者：李明伟,洪维强
第一作者：李向阳
论文类型：期刊论文
通讯作者：杨忠仪
论文编号：112006
学科门类：工学
文献类型：J
卷号：217
是否译文：否
发表时间：2026
收录刊物：SCI
发布期刊链接：https://www.sciencedirect.com/science/article/pii/S036083522600207X?via%3Dihub