多用户动态交通流分配模型及算法研究

Multi-user dynamic traffic assignment model and relevant algorithm

通过对动态交通网络中瞬时反应型和预测型出行者行为假设的分析,提出了将两者结合并增加固定路线出行行为假设的多用户动态交通流分配模型.同时通过对动态交通网络出行者行为的相互制约特性分析,给出了求解上述模型的迭代算法.本模型拓展了以往构建模型中出行者行为的单一性假设.本模型及算法也可直接应用于先进的出行者信息系统,并为路径导行系统中网络的效益分析提供理论依据.

Through analyzing behavioral assumptions based on instantaneous and real-time travel time, a multi-user dynamic traffic assignment model the is proposed. In the new model the three behavioral assumptions, i. e, instantaneous preference, real-time preference and constant routes preference are integrated. Taking advantages of the interplay among three behaviors in the dynamic traffic network, an iterative solution algorithm is put forward. The new model expands the conventional single behavioral assumption and can be applied to the advanced traveler information systems（ATIS） directly. The multiuser dynamic traffic assignment model and relevant algorithm provide a basis for the utility analysis of dynamic vehicle route guidance system.