Study of real-time traffic state estimation and short-term prediction of signalized arterial network considering heterogeneous information sources

Abstract

Compared with a freeway network, real-time traffic state estimation and prediction of a signalized arterial network is a challenging yet under-studied field. Starting from discussing the arterial traffic flow dynamics, this study proposes a novel framework for real-time traffic state estimation and short-term prediction for signalized corridors. Particle filter techniques are used to integrate field measurements from different sources to improve the accuracy and robustness of the model. Several comprehensive numerical studies based on both real world and simulated datasets showed that the proposed model can generate reliable estimation and short-term prediction of different traffic states including queue length, flow density, speed and travel time with a high degree of accuracy. The proposed model can serve as the key component in both ATIS (Advanced Traveler's Information System) and proactive traffic control systems