考虑汽车流冲突的汽车滚装码头泊位分配研究  

Berth allocation research on automobile Ro-Ro terminals considering vehicle flow conflicts

作　　者：隋毅 金建钢 SUI Yi;JIN Jiangang(School of Ocean and Civil Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]上海交通大学,船舶海洋与建筑工程学院,上海200240

出　　处：《交通运输工程与信息学报》2024年第3期118-133,共16页Journal of Transportation Engineering and Information

基　　金：国家自然科学基金优秀青年基金项目(72122014)。

摘　　要：本文提出了考虑汽车流冲突的汽车滚装码头泊位分配问题。首先,定义汽车滚装码头前沿的汽车流冲突情况,引入冲突惩罚时间系数,量化汽车码头前沿汽车流冲突惩罚时间,建立允许发生汽车流冲突和不允许发生汽车流冲突的两种以最小化船舶总工作时间为目标函数的混合整数规划模型。其次,根据滚装船和汽车流的对应关系,设计滚装船-汽车流多层对应编码的遗传算法和基于规则策略的解码算法。然后,生成4组不同规模的测试算例进行数值实验,根据实验结果确定遗传算法的相关参数。在小规模算例下,遗传算法的最优解与CPLEX精确解之间的差距小于1%;在大规模算例下,遗传算法的最优解明显优于基于经验排班的策略解,进而验证本文提出的模型正确性和遗传算法的有效性。最后,在两种冲突策略的场景下对冲突惩罚系数进行敏感性分析,结果表明当冲突惩罚系数低于0.5时,两种策略的目标函数值有明显偏差,汽车滚装码头制定泊位分配计划时应该考虑汽车流之间冲突情况的影响。This paper addresses the berth allocation problem in automobile rollon/rolloff(Ro-Ro)terminals,considering conflicts in the flow of automobiles.Initially,the study defines conflicts in the forefront vehicle flow at Ro-Ro terminals,introducing a conflict penalty time coefficient to quantify the associated penalty time.We then proposed two mixed-integer programming models:one allowing automobile flow conflicts and the other prohibiting them.Both models aim to minimize the total working time of the ships.Subsequently,we design a genetic algorithm(GA)incorporating a multi layer correspondence encoding for Ro-Ro ships and automobile flows.Additionally,we implement a decoding algorithm based on a rule strategy.A total of four sets of test cases of varying scales are generated for numerical experiments,and the parameters of the GA are determined based on experimental results.In small-scale cases,the optimal solution of genetic algorithm is within 1%of the exact CPLEX solution.In large-scale cases,the GA outperforms the strategy-based solution derived from empirical scheduling,thus validating the correctness of the model and the effectiveness of the GA.Finally, this study conducts sensitivity analysis of conflict penalty coefficients under two conflict strategies. The results indicate that when the coefficient is below 0.5, a noticeable deviation is observed in the objective function values of the two strategies. Hence, when formulating berth allocation plans for automobile Ro-Ro terminals, the impact of conflicts between automobile flows should be considered, particularly when the conflict penalty coefficient is low.

关 键 词：水路运输 泊位分配 遗传算法 滚装码头 汽车流 

分 类 号：U691.3[交通运输工程—港口、海岸及近海工程]