Modeling Traffic Flow Mixed with Automated Vehicles Considering Dynamic Safety Distance

Abstract

Autonomous driving technology is hailed as a pivotal solution for impending transportation challenges. Current research primarily analyzes the impact of autonomous driving technology through numerical simulation to better develop autonomous driving application scenarios. To elaborate on the driving behavior of various vehicle types, a concept of dynamic safety distance has been proposed in this paper for a detailed analysis of the impact of road traffic conditions on vehicles during driving. By revising car-following and lane-changing rules, a mixed traffic flow analysis model for autonomous and manual driving vehicles has been established. Through a small-scale cellular automaton model, congestion ratio (CR) and lane-changing ratio (LCR) are defined to discuss the impact of parameters such as response time and the penetration rates of autonomous vehicles on the lane traffic flow. The numerical simulations show that the traffic flow and average velocity of vehicles on the lane have significantly improved as the penetration rates of autonomous vehicles increase. This is because autonomous vehicles excel in controlling response time and distance factors compared with traditional manual driving vehicles. As seen from the vehicle-following response, autonomous vehicles maintain a shorter headway distance with the preceding vehicle and also improve lane utilization by changing lanes. When the penetration rate of autonomous vehicles increases from 0.2 to 0.6 in the 60% density lane, the congestion situation of the lane can be reduced by approximately 25%.