An Optimal Control-Based Vehicle Speed Guidance Strategy to Improve Traffic Safety and Efficiency Against Freeway Jam Waves
Yu Han,Hao Yu,Zhibin Li,Chengcheng Xu,Yanjie Ji,Pan Liu
DOI: https://doi.org/10.1016/j.aap.2021.106429
IF: 5.9
2021-01-01
Accident Analysis & Prevention
Abstract:Freeway jam waves create many problems, including capacity reduction, travel delays, and safety risks. The development of cooperative vehicle infrastructure system (CVIS) has prompted numerous new strategies, which can resolve jam waves by implementing microscopic car-following control actions to individual vehicles. However, most of those strategies aimed at eliminating freeway jam waves without considering the safety risks induced by the car-following control. This paper proposes an optimal control-based vehicle speed guidance strategy to improve both traffic efficiency and safety against jam waves. The optimal controller is developed based on a discrete first-order traffic flow model formulated in Lagrangian coordinates. The optimization of vehicles' driving speed is formulated as a linear programming problem, where the constraints concerning threshold safety measures are imposed. The proposed vehicle speed guidance strategy is tested using a modified Intelligent Driving Model (IDM+), which represents real traffic dynamics in CVIS environment. The proposed speed guidance strategy is compared with a state-of-the-art jam-absorption driving strategy, which also aimed to eliminate freeway jam waves. Simulation results show that the proposed strategy outperforms that strategy in terms of both total time spent saving and surrogate safety measures' reduction. The time exposed time-tocollision (TET) is reduced by 31%, and the time integrated time-to-collision (TIT) is reduced by 9.5% on average. Furthermore, the computation time of the linear optimization is only a few seconds, which is fast enough for the online application of the proposed strategy.