Smoothing Traffic Flow Through Automated Vehicle Control With Optimal Parameter Selection

Abstract

Stop-and-go traffic waves reduce transportation system efficiency by increasing traffic oscillations and energy consumption. This study develops an approach to synthesize a class of additive feedback controllers for automated vehicles (AVs) to smooth nonlinear mixed traffic flow, which includes both AVs and human-driven vehicles (HVs). Unlike recent AV controllers that rely on strict assumptions, such as vehicle-to-vehicle communication or equilibrium traffic speed, our proposed controller utilizes only local traffic information, including inter-vehicle spacing and relative speed, which is readily available through onboard AV sensors. By tracking a smoothed version of the perturbed speed profile of its preceding vehicle, the controlled AV facilitates smoother traffic flow. Additionally, we present a method for selecting optimal control parameter values to enhance traffic-smoothing efficiency. These features significantly improve the implementability of the proposed AV controller. We evaluate its effectiveness through simulations of three distinct traffic scenarios with varying levels of oscillation. The results show that AVs using the proposed controller can effectively reduce traffic oscillations and decrease fuel consumption by up to 46.96% and 2.74%, respectively, across the traffic scenarios considered. The traffic-smoothing effect is more pronounced at higher AV penetration rates. While the proposed approach exhibits slightly lower performance than the most recent additive AV controller, which relies on stricter assumptions, it offers greater implementability and provides a systematic method for selecting optimal control parameters.