Title: Designing of intersection driving behaviours based on reward points in congestion-induced emergency situations

Authors: Dexian Zeng; Wen-Long Shang; Huibo Bi

Addresses: College of Metropolitan Transportation, Beijing University of Technology, Beijing, China ' Beijing Key Laboratory of Traffic Engineering, College of Metropolitan Transportation, Beijing University of Technology, Beijing, China ' Beijing Key Laboratory of Traffic Engineering, College of Metropolitan Transportation, Beijing University of Technology, Beijing, China

Abstract: With the ubiquity of portable smart devices, the perception, calculation, and communication capabilities of drivers have been greatly improved, and they have gradually changed from the pure service object of the transportation system to co-decision-makers. Therefore, this paper proposes an intersection-targeted driving behaviour optimisation mechanism by encouraging drivers to conduct the traffic flow ratio changing tasks in exchange for reward points. We first formalise compliance rate-reward value functions by using a number of questionnaire-based surveys. Then we employ a reinforcement learning model to optimise the traffic flow ratios at intersections of a road network and publish reward gaining tasks to dynamically optimise the traffic flow ratios based on the current traffic flow ratios and the number of vehicles at the intersections. The experimental results show that the introduction of points-based rewards can effectively improve the traffic efficiency at intersections.

Keywords: points-based reward; driving behaviour optimisation; Simulation of Urban MObility; SUMO; reinforcement learning.

DOI: 10.1504/IJCCM.2024.140445

International Journal of Chinese Culture and Management, 2024 Vol.5 No.4, pp.293 - 309

Received: 29 May 2021
Accepted: 30 Nov 2021
Published online: 19 Aug 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article