考虑多峰的系统最优动态路径流量分配图解方法研究

动态交通分配是交通科学研究的难点，尤其是如何得到系统最优解.Munoz和Laval (2006)介绍了一类图解方法，得到了并行网络动态路径流量分配系统最优解.在此基础上，本文扩展其图解方法，研究存在多个高峰情形的并行网络动态路径流量分配系统最优问题.基于累计到达曲线已知的假设，即出行用户的出发时刻选择给定，并借助于瓶颈模型点排队假设，以两条并行路径为例，考虑其中一条路径瓶颈处的容量为常数，另一条路径瓶颈处容量为常数和无穷大两种情形，通过变分法描绘满足动态最优性条件的系统最优流量分配曲线，得到动态路径流量分配系统最优解.本文的研究有助于加深对交通流量时空分布规律的理解.

Graphical Solution for System Optimum Dynamic Traffic Assignment Problem with Multiple-peak Demand

It is difficult to study dynamic traffic assignment (DTA) problem in transportation science area, especially how to acquire the system optimal solutions. One kind of graphical solution method is introduced by Munoz and Laval (2006) to get the dynamic system optimal solutions. Based on this, this paper extends their method and investigates the system optimum dynamic traffic assignment problem in parallel network with multipeak demand. It is assumed that the cumulative arrival curve is known and point-queue bottleneck, and the traffic assignment lines are drawn by calculating the sum of variations to make sure that the dynamical optimality conditions are satisfied. Two cases are considered: the bottleneck capacity of one link is constant and that of another link are, respectively, constant and unlimited. The conclusions of this study are helpful to understand the spatial and temporal distribution of traffic flow deeply.