Abstract: According to the theory of “car sharing”, self-service public taxi is a means of transportation which is suitable for the short-time bi-directional travel in the city. The investigation on the intention of self-service public taxi and demand analysis was carried out. Under the limit of the total scale which was solved by grey forecast method, the alternative scheme of network layout based on the balance between supply and demand was obtained; on this basis, considering the total social investment and users’ interests, the bi-level programming model was established and the optimal scheme of network layout was obtained by the application of simulated annealing algorithm. At last, the empirical analysis was carried out in Changsha; the calculation scheme was compared with the schemes calculated out by other methods. The research results show that the simulated annealing algorithm is an effective algorithm, and bi-level programming model can be a good application to the network layout optimization. The economic benefit analysis shows that the network layout scheme meets the needs of the community and economic benefits.

Key words: traffic engineering, bi-level programming model, simulated annealing algorithm, network layout, self-service public taxi

摘要： 基于“汽车 共享”的理念，提出了一种适合城市内部短时双向出行的交通方式——自助式公共出租车。通过自助式公共出租车乘车意向调查和需求分析，在灰色预测 法预测出的总规模限制下，先基于供需平衡得到备选网点布局方案，再综合考虑社会总投资和各用户利益，建立双层规划模型，用模拟退火算法得到网点 布局优化方案。最后，以长沙市为例进行实例分析，将运算方案与其他方法计算出的方案进行了比较。研究结果表明：模拟退火算法是一个有效的算法， 双层规划模型可以很好的应用到网点布局优化。经济效益分析得出网点布局方案符合社会需求、满足经济效益。

关键词: 交通工程, 双层规划模型, 模拟退火算法, 网点布局, 自助式公共出租车