基于交叉巢式Logit的多用户多模式随机用户均衡模型

为了完善现有城市交通分配模型，建立了一个基于交叉巢式Logit的多用户多模式随机用户网络均衡模型。该网络具有以下特征：网络中存在多类用户，不同类别用户的出行时间价值不同；用户可使用多种交通模式，不同模式相互干扰，且模式间路段阻抗影响满足对称条件。路径的选择采用克服多项式Logit模型的不相关备选方案独立属性的交叉巢式Logit模型来表示，并建立了一个等价数学规划公式来表示这一多用户多模式混合均衡状态，同时使用一个简单算例对基于交叉巢式Logit和基于传统的多项式Logit的模型进行了比较。结果表明：数学规划公式与基于交叉巢式Logit的随机用户均衡条件具有等价性，证明了模型最优解的存在性和唯一性；相比传统的离散选择模型，所提出的模型能更好地在混合状态下进行路径选择。

Multi-user and Multi-mode Stochastic User Equilibrium Model Based on Cross-nested Logit

In order to improve current urban traffic assignment models,a multi-user and multi-mode stochastic user equilibrium model based on cross-nested Logit was established.There were many different users in this network and different users had different values of time.Users could use many modes.Different modes interfered with each other and had symmetry effect on each other.A cross-nested Logit model,which could overcome the shortcoming of independence and identical distribution of Logit model,was used to model route choice.An equivalent mathematical programming formula was used to show the mixed equilibrium status.A simple illustration was used to compare the model based on cross-nested Logit with ones based conventional multinomial Logit.The results prove the equivalence of the mathematical programming formula to stochastic user equilibrium conditions and the existence and uniqueness of the optimum solution.The established model can better perform route choice under mixed state compared with traditional discrete choice models.