Real-time traffic state predictor based on dynamic traveller behaviour

Real-time traffic state prediction is a crucial component of the practical applications of and existing theoretical research on active traffic management (ATM) systems. Dynamic traveller behaviour is a key factor of the complexity of traffic state prediction under information provision. The aim of this study is to consider the influence of travellers' route-switching behaviour (TRSB) on traffic state prediction for accurate network traffic prediction under information provision. A realistic TRSB model and dynamic traveller behaviour combined with a macroscopic dynamic traffic flow model framework for information-based ATM are proposed. Accordingly, this work establishes an integrated realistic TRSB with traffic state prediction model under the state-space model framework and an extended Kalman filtering solving algorithm. The experimental results on the road network of the traffic guidance demonstration projects reveal that the proposed model can accurately estimate the drivers' responses to traffic information and the impact of realistic TRSB on traffic prediction and improve the accuracy of traffic state prediction under information provision. Therefore, considering the influence of dynamic travel behaviour on network traffic state prediction under guidance information is necessary to ensure the sustainability of the information-based ATM system.