城市道路网络动态OD矩阵预测模型

实时动态OD(Origin Destination)矩阵是动态交通分配系统的重要输入. 运用状态空间模型的有关概念，考虑交叉口进口道观测数据建立了相应的状态空间模型，提出了一步预测卡尔曼滤波方程实现估计和预测. 案例研究表明，将交叉口观测数据纳入测量方程能更好地对OD矩阵做出较精确的估计和预测，且基于仿真的分配矩阵估计方法能更准确地反映OD流量在检测设施上的比例分配.

Dynamic OD Matrix Estimation Model of Urban Network

Real-time dynamic OD matrix is an important input for dynamic traffic assignment system.This paper built a new state-space model whose measurement equation will take the interaction detection data into account,and proposed a one-step prediction Kalman-filtering model to achieve this target.The case study shows that including the interaction data into the measurement equation will give a more accurate estimation and prediction and a simulation-based assignment matrix will give a more accurate reflection on how many volumes are detected by certain facility.