智慧公路多源数据下的交通出行演化模型

为探索智慧公路对出行者路径选择行为的影响，将网络可靠性和经验学习理论引入出行者的路径选择过程. 运用交通流理论统一化由智慧公路采集的多源数据类型，进一步采用最小方差加权平均方法融合多源数据. 智慧公路上的出行者根据融合信息实时决策出行路径；普通公路上的出行者根据前一天的路网交通状况和历史出行经验选择出行路径. 采用不动点理论证明模型解的等价性、存在性和稳定性条件. 算例结果表明：道路流量行为系数和感知时间误差的增大均会导致模型进入不稳定状态；在模型解稳定性方面，具有冒险倾向的出行者显著优于具有风险规避倾向的出行者，且多源融合数据演化的路网交通流量更具鲁棒性.

Traffic evolution model with multi-source data of intelligent highway

The network reliability and experimential learning theory was introduced into travelers’ route choice decision process, to explore the influence of intelligent highway on path choice behavior. The traffic flow theory was used to transform the multi-source data of intelligent highway to a uniform format. Moreover, the method of minimum variance weighted average was proposed to fuse the multi-source data. Based on the fuse information, the travelers on intelligent highway chose their travel paths. While, the travelers on the ordinary road made their travel decision through considering the road network traffic conditions on the last day and the historical travel experiences. The fixed point theory was adopted to prove the equivalency, existence, and stability of the solutions of the built model. The numerical examples demonstrate that both the increase of road traffic flow behavior coefficient and perception travel time error cause the model to enter an unsteady state; in the aspect of stability of model solutions, the risk-prone travelers are significantly better than the risk-averse travelers; the evolved network traffic flows with fused multi-source data have better robustness.