LAUSR

Uncertainties and disturbances in traffic flows will result in performance degradation and even the critical instability of automated freeway systems. To handle these issues, a multivariable model predictive controller is proposed in this article, taking the speed and density of traffic flows as the control objectives and accounting for the uncertainty caused by ramps. Specifically, a state-space macroscopic traffic flow model is established based on a macroscopic traffic flow model. Then, a multivariable model predictive controller is developed based on the state-space model, and a sequence quadratic program algorithm is designed to find the optimal inputs. Finally, extensive simulations and comparisons are conducted. The results from the simulations verify that the proposed controller can make the density and speed of traffic flows approach a desired value more quickly and without chattering. In addition, the controller can effectively avoid traffic congestion and has better stability in the presence of disturbances.