Active group route guidance based on DODE: A novel modeling framework integrating dynamic driver behavior

Abstract

This work aims to provide effective and reliable traffic guidance strategies for improving the system performance and travel reliability, wherein the drivers’ route diversion behavior is the key determinant in these strategies. This study presents a novel modeling framework that can incorporate the dynamic driver behavior into a real-time group route guidance model based on dynamic origin–destination demand estimation and prediction (DODE) for information-based active traffic management. Experiments are conducted to test the effectiveness of the proposed model on the basis of the traffic dataset of the Route Guidance Pilot Project. Experimental results show that the effect of route diversion on DODE under information provision, which can improve the accuracy of DODE, must be considered. Compared with the traditional guidance model, the proposed model considers the system objective and the actual route diversion behavior and can provide better performance and ensure system sustainability.