中文摘要：

将网络舱位控制的资源分解法与作者提出的单航段超订下舱位控制的R—MDP模型结合，提出超订下网络舱位控制的稳健模型，给出了9种稳健联合策略．并对9种稳健联合策略的总收益进行了比较．模拟分析表明，当存在需求预测误差时，网络稳健收益高于网络名义收益，随着剩余座位数减少或误差水平增大，稳健联合策略对于期望总收益的提高幅度将逐渐增大．

英文摘要：

Combining the single-leg R-MDP model appeared in reference [9] with the resource decomposition approach of network capacity control, we build up the robust model of capacity control with overbooking for the airline hub-and-spoke network, and design nine types of robust joint strategies. We give a numerical example to show that the robust revenue obtained from the optimal robust joint strategy is higher than the corresponding nominal revenue obtained from the optimal nominal joint strategy. Moreover, with the existence of demand estimation errors, the less the remaining capacity or the larger the error level is, the deeper the extent of the increasing revenue by using the robust joint strategy. At last, we analyze two kinds of robust bid price strategies and compare the total revenue of the proposed nine robust joint strategies for the numerical example.