【摘要】 为深入优化物流企业的自有及外包运力配置，综合考虑运输需求不确定性和运输车型异质性，以全年运输利润最大化为目标，构建了多车型参与下运力配置鲁棒优化模型.依据决策变量的层次特性，设计了一种内外双层结构的嵌套式自适应遗传算法(NAGA).依托上海某物流企业的实际数据，进行了需求确定及不确定情形下的实例分析.结果表明，所构建的鲁棒优化模型兼顾经济性与鲁棒性，NAGA能对该模型进行高效求解.在需求确定情形下，优化后的运输利润提高了6.2%,且NAGA在求解质量及稳定性方面表现优于典型嵌套式启发算法；在需求不确定情形下，采用所提出的运力配置方法，决策者可根据市场波动水平及风险偏好调整鲁棒优化模型参数取值，以灵活获取合适的自有及外包运力配置方案.更多还原

【Abstract】 To optimize the owned and outsourced capacity allocation for logistics companies, considering the demand uncertainty and the vehicle heterogeneity, a robust optimization model for logistic capacity allocation with multiple vehicle types was constructed to maximize annual transport profit. A nested adaptive genetic algorithm(NAGA) with an inner and outer double-layer structure was designed based on the hierarchical characteristics of decision variables. Using actual data from a logistics company in Shanghai, case studies under certain and uncertain demand scenarios were conducted. Results show that the robust optimization model effectively balances economy and robustness, and the NAGA solves it efficiently. Under certain demand, optimized transport profit increases by 6.2%, with the NAGA outperforming typical nested heuristic algorithms in solution quality and stability. Under uncertain demand, the proposed method allows decision-makers to adjust the parameters of the robust optimization model according to the market volatility and risk preference, in order to flexibly obtain suitable owned and outsourced capacity allocation schemes.更多还原