Estimation of OD Patterns on Dynamic Networks

TB26.1 An Adaptive Kalman Filtering Algorithm for the Dynamic Estimation & Prediction of Freeway OD Matrices ShouRen Hu, Samer M. Madanat --- Purdue Univ., Sch. of Civ. Eng., West Lafayette, IN 47907 ,

The problem of dynamic estimation and prediction of freeway OD¨ matrices has received increasing attention in view of its¨ applicability to on-line traffic management systems. We present an¨ adaptive Kalman filtering algorithm in which the state vector to be¨ estimated includes time-varying OD flows and model parameters. By¨ capturing the time-varying nature of traffic variables, the proposed¨ algorithm provides, in a statistical sense, more accurate estimates¨ of dynamic OD flows...

TB26.2 Estimation of Dynamic OD Matrices for Signalized Urban Networks Gag-Len Chang, Xingding Tao --- Univ. of MD, Dept. of Civ. Eng., College Park, MD 20742 ,

This study has developed a 2-stage estimation model for dynamic O-D¨ matrices in signalized urban networks. The first stage involves a¨ parallel computation of dynamic O-D at an intersection level. The¨ results from the first stage, considering the impacts of signal, are¨ then integrated in a network-wide estimation.

TB26.3 Application of Microscopic Simulation to Evaluate Real-Time Traffic Control Systems Moshe E. Ben-Akiva, A. Hotz, H. Koutsopoulos, R. Mishalani, Q. Yang --- MIT, Room 1-181, Dept. of Env. & Civ. Eng., Cambridge, MA 02139 , (mba@mit.edu)

A simulation laboratory for evaluating real-time traffic control¨ systems is presented. The laboratory includes 4 integrated¨ components: traffic control, control devices, traffic flow and¨ surveillance. Application to Boston's Central Artery/Tunnel control¨ system is discussed.