Combined Variable Speed Limit and Lane Change Control for Highway Traffic

Abstract

Variable speed limit (VSL) control of highway traffic is expected to improve traffic mobility, safety, and environment, especially during incidents. However, most existing VSL controllers show significant benefits in macroscopic analysis but little improvement in microscopic simulations in terms of traffic mobility. We demonstrate that the lack of improvement for travel time in many incident cases is due to lane changes that are taking place close to the bottleneck leading to severe capacity drop, which is not adequately captured by most macroscopic models. In this paper, we develop a combined lane change and VSL control scheme, which generates consistent improvements both with macroscopic and microscopic models. The lane change controller generates lane change recommendations upstream the incident or bottleneck in order to reduce the effect of the capacity drop. The VSL controller is developed using a feedback linearization approach based on the cell transmission macroscopic model and is shown analytically to guarantee exponential convergence to the optimum equilibrium point. Microscopic Monte Carlo simulations of traffic on the I-710 freeway were used to demonstrate that this combined control strategy is able to generate consistent improvements with respect to travel time, safety, and environmental impact under different traffic conditions and incident scenarios.