论文部分内容阅读
为了分析时间标准以小时或天为单位的网络交通流演化问题,利用投影动态系统理论建立了相关问题的演化投影动态系统。从网络组成基本元素入手,首先分析了出行者途中对出行路段的选择行为和对局部网络节点最短行程时间的重估问题,得到了基于网络局部信息的出行者路线调整策略。对上述策略进行时间上的演化建模,得到了网络交通流演化的投影动态系统。通过分析与该投影动态系统有相同稳态点的变分不等式,揭示了该系统与基于用户均衡的交通流分配模型的一致性。利用建模过程的分析,设计了启发式的投影动态系统离散演化算法。通过设计阶梯式算法收敛指标,可以加速演化得到网络的稳态。算例分析验证了模型算法的有效性。研究表明基于投影动态系统的交通流演化模型不仅能较好地刻画交通流演化特征,而且可应用于交通管理措施的短期交通流影响分析。
In order to analyze the hourly or day-based network traffic flow evolution problem, the dynamic evolutionary projection system of related problems is established by using projection dynamic system theory. Starting from the basic elements of network composition, this paper firstly analyzes the choice behavior of the traveler on the way of travel and the re-estimation of the shortest travel time of the local network node, and obtains the route adjustment strategy of the traveler based on the local information of the network. The temporal evolution of the above strategy is modeled and the projection dynamic system of network traffic flow evolution is obtained. By analyzing the variational inequality which has the same steady-state point with the projection dynamic system, the consistency between the system and the traffic flow distribution model based on user equilibrium is revealed. Based on the analysis of modeling process, a heuristic algorithm for discrete evolution of projection dynamic system is designed. By designing the convergence index of stepwise algorithm, the steady state of the network can be accelerated and evolved. The example analysis verifies the validity of the model algorithm. The research shows that the evolution model of traffic flow based on projection dynamic system can not only characterize the evolution of traffic flow well, but also be applied to the short-term traffic flow impact analysis of traffic management measures.