Abstract: With the vigorous development of the automobile import and export business in recent years, the RO-RO terminal has become an important logistics node of the automobile trade. In 2021 the RO-RO export volume accounted for more than 50% of the national export volume, compared with air transport and other methods in the automobile export business. RO-RO transport has gradually become the mainstream mode of transport. However, with an increase in the annual throughput, the limited parking space of the yard and the characteristic ( the vehicle cannot be stacked) affect the collection and distribution operation efficiency of the RO-RO terminal and restrict the further development of the RO-RO transportation business, therefore it is necessary to study scientific and efficient scheduling optimization strategies based on the operation characteristics of the RO-RO terminal.
In order to improve the utilization rate of parking space in the storage yard, based on the existing literature on the scheduling optimization problem of the RO-RO terminal this paper puts forward the optimization problem of the collection cycle of the RO-RO terminal, analyzes the operation and scheduling characteristics of the RO-RO terminal and the export operation process of the finished vehicle, considers the service ability of the terminal yard and the storage cost of the finished vehicle, and arranges the collection cycle reasonably for the port operation of the automobile manufacturer. In order to minimize the actual daily work difference in yard workers and the demurrage cost of the parking space occupied by the finished vehicle, an integer programming model is established, by taking the finished vehicle demand of the RO-RO vessel as the constraint, the collection cycle and the actual daily finished vehicle collection number of the automobile manufacturer in the RO-RO terminal are decided.
Considering that the optimization problem of collection cycle is a problem requiring long-term decision, and the column generation algorithm is usually used to solve large-scale linear programming problems, the branch pricing algorithm based on column generation is chosen to solve the problem. The problem is first reconstructed and then decomposed into a limited main problem of selecting the combination of the optimization scheme of the collection period and a pricing sub-problem of the optimization of the collection period with demand constraints.Through the continuous iteration of the column generation algorithm, when the subproblem cannot produce the solution with the objective function less than 0, the current solution is the optimal solution of the linear relaxation problem of the main problem, the optimal integer solution of the problem can be solved by branch and bound algorithm, and then random examples are generated based on the actual problem size for experiment.
In order to verify the accuracy of the model, this paper first conducts small-scale experiments to solve the model M2 by using the branch pricing algorithm programmed by IBM ILOG CPLEX 12.10 and Python3.6 respectively. The gap value (the gap value is equal to the difference between the objective function value under branch pricing algorithm and the objective function value under CPLEX divided by the objective function value under CPLEX) in the experimental results is compared to verify the algorithm. The branch pricing algorithm is used to solve the large-scale experiment with different daily service capacity and ship demand, and it is verified that the branch pricing algorithm has certain application value in solving the optimization problem of the collection cycle of the vehicle RO-RO terminal. Finally, through the sensitivity analysis of the parameter daily service capacity and free period of the yard, it is found that if the inventory cost of the yard is considered, increasing the daily service capacity of the yard can better improve the utilization rate of the parking space in the storage yard.
Future research can be considered from the following aspects: further refining the collection process of the vehicle RO-RO terminal, and combining the vehicle ex-factory combined transport with additional services of the RO-RO terminal. Overall vehicle export process of automobile RO-RO terminal, and a series of plans for the vehicle parking space distribution-drivers-collection of ship loading, thus to enhance the overall efficiency of the automobile RO-RO terminal.

Key words: automobile logistics, RO-RO terminals, collection cycle, column generation, branch and price algorithm

摘要： 随着近年来汽车进出口业务的蓬勃发展,滚装码头成为汽车贸易的关键物流节点,为了提高其堆场车位的利用率,本文提出汽车滚装码头集车周期优化问题,通过分析汽车滚装码头的整车出口作业流程及运营调度特点,以最小化堆场工作人员实际每日工作差和船舶的需求整车占用堆场车位的滞期成本为目标建立了优化集车周期问题的整数规划模型。在此基础上重构模型,将原问题分解为限制主问题和定价子问题,提出了基于列生成的分支定价算法。基于实际问题规模生成随机算例实验验证了所提出模型和算法的准确性,并对堆场日均服务能力和免堆期参数进行了灵敏度分析。实验结果表明,在求解该问题时分支定价算法在求解质量和时间上均优于CPLEX求解原整数规划模型,所提出的模型和算法可以为汽车滚装码头的集车周期优化问题提供决策支持。

关键词: 整车物流, 滚装码头, 集车周期, 列生成, 分支定价算法