Integrated traffic control for mixed urban and freeway networks: A model predictive control approach

Abstract

We develop a control method for networks containing both urban roads and freeways. These two road types are closely connected: congestion on the freeway often causes spill-back leading to urban queues, slowing down the urban traffic, and vice versa. As a consequence, control measures taken in one of the two areas can have a significant influence on the other area. We first develop a model that describes the evolution of the traffic flows in mixed networks. Next, we propose the control method that is used for the integrated control. This approach is based on model predictive control, in which the optimal control inputs are determined on-line using numerical optimisation and a prediction model in combination with a receding horizon approach. We also compare our newly developed control method with systems that are similar to existing dynamic traffic control systems like SCOOT and UTOPIA/SPOT, in a qualitative as well as in a quantitative way via a simple case study. The results illustrate the potential benefits of the proposed approach and motivate further development and improvement of the proposed control method.