先验时间特性约束下的动态OD矩阵估计

Dynamic OD Matrix Estimation under Prior Temporal Characteristic Constraints

利用路段流量和先验信息,提出了先验时间特性约束下的动态OD矩阵估计模型.该模型最主要是在经典的Cascetta提出的模型(传统模型)基础之上,通过应用Toeplitz约束矩阵捕获OD矩阵的时间平滑特性以进一步提高模型精度.在交通分配过程中,采用Toledo提出的交通分配矩阵线性近似方法,不需要假设OD分配矩阵是一个常量,而是具有时变性,算法基于交通仿真软件Mezzo实现.以青岛市市南区路网为实例进行实验,结果表明:本文提出的先验时间特性约束下的动态OD矩阵估计模型与传统模型相比效果更好.

Estimating dynamic origin-destination ( OD ) demand matrices is very important in transportation demand modeling. In this paper, we mainly use the traffic flow and prior information to propose a dynamic OD matrix estimation model under the constraint of prior temporal characteristic. Based on the classical model by Cascetta, the proposed model captures the temporal smoothing characteristic of OD matrix by using Toeplitz constraint matrix to further improve the accuracy of the model. In the process of traffic assignment, Toledo s linear approximation method of traffic assignment matrix is adopted. We do not need to assume that OD assignment matrix is a constant, but with time-varying. The algorithms are implemented in a tool that uses the traffic simulation model Mezzo to conduct network loadings. A case study network in the Shinan District of Qingdao City, is used to test the proposed model, and to compare its performance with traditional models. The results show that the proposed dynamic OD matrix estimation model with prior temporal characteristic constraints achieves better results.