Abstract: Electric car-sharing contributes to alleviating traffic congestion and environmental pollution. To improve the efficiency of the system and support its sustainability, this paper proposes the method of making planning decisions for one-way electric car-sharing system based on uncertain demand. The multiple scenario method is applied to deal with the uncertainty of the demand, electric car and personnel are aggregated by stations and their status. The one-way electric car-sharing system planning model is then developed with consideration of the interaction between demand and trip price. According to the feature that demand is integer, an outer-approximation scheme based on secant is designed and the corresponding outer-approximation algorithm is proposed to solve the models with three types of the demandprice function. The numerical examples are used to verify the necessity of considering uncertainty of demand and analyze the effects of demand variation, demand elasticity and the demand on profit. The results show the profit increases with the increase of car-sharing demand and variation; the profit decreases with the increase of the absolute value of demand elasticity.

Key words: urban traffic, planning decision, mix-integer non-linear programming, electric car-sharing, outerapproximation algorithm

摘要： 电动汽车共享有助于解决交通拥堵和环境污染问题。为提高电动汽车共享系统效率，保 障系统的可持续性，本文考虑共享需求的不确定性，提出单向电动汽车共享系统最优配置方法。 首先，以多情景处理共享需求的不确定性，以车站和状态集计车辆和调度员，考虑需求价格弹性， 构建共享电动汽车系统配置模型。然后，针对3类共享出行需求价格弹性函数，根据需求为整数 的特点，设计基于割线的外逼策略，提出用于模型求解的外点法。最后，算例证实了考虑共享需 求不确定性的必要性，分析了共享需求方差、弹性以及需求量对共享电动汽车系统利润的影响。 研究结果表明，共享电动汽车系统利润随需求量和需求方差的增加而增加，随需求弹性绝对值的 增加而减少。

关键词: 城市交通, 规划决策, 混合整数非线性规划, 共享电动汽车, 外点法