Container slot allocation for liner shipping services is to allocate the limited container slots of ships to different segments of demands in order to maximize the total revenue over a shipping network. This study focuses on a planning-level container slot allocation problem with uncertain demand, which is essential in container shipping revenue management. Due to the challenge of calibrating/formulating a specific probability distribution of uncertain container slot demand, we can rely on its fundamental descriptive statistics, namely, mean, upper/lower bounds as well as variance to tackle the container slot allocation problem. We, therefore, develop a robust optimization model using the fundamental descriptive statistics, which is approximated by a solvable second-order cone programming model. A numerical example based on a real shipping network demonstrates that the optimal solution from the second-order cone programming model outperforms the models using the expectation of uncertain demand data and the normally distributed demand. The numerical results indicate that the robust optimization model can well deal with the large fluctuations of uncertain container slot demand.

班轮运输服务的集装箱舱位分配是将船舶有限的集装箱舱位分配给不同的需求细分，以最大化航运网络的总收入。本研究侧重于需求不确定的规划级集装箱舱位分配问题，这在集装箱航运收入管理中至关重要。由于校准/制定不确定集装箱槽位需求的特定概率分布的挑战，我们可以依靠其基本描述统计，即均值、上/下界以及方差来解决集装箱槽位分配问题。因此，我们使用基本描述统计开发了一个稳健的优化模型，该模型由可解的二阶锥规划模型近似。一个基于真实航运网络的数值例子表明，二阶锥规划模型的最优解优于使用不确定需求数据的期望和正态分布需求的模型。数值结果表明，鲁棒优化模型能够很好地处理不确定的集装箱槽位需求的大波动。