Simulation-based assessment of natural robustness of freeway traffic systems controlled via MPC

Model Predictive Control schemes have been often studied for defining suitable traffic control laws for freeway systems. If, as known, MPC allows to update the control law in real time depending on the measurements on the system state, its main drawback stands in the high computational effort required to solve a finite-horizon optimal control problem at each time step. In order to reduce the overall control computational effort and then to increase the applicability of MPC in freeway traffic systems, a control law precalculated off line on the basis of a nominal system could be applied. Of course this is effective only if the system, that is characterized by uncertainties such as the variations of the mainstream and on-ramp demands with respect to the nominal ones, results to be naturally robust. The objective of this paper is to evaluate if the MPC law obtained by considering the nominal system for prediction remains effective when applied to the system affected by disturbances. To this purpose, the property of natural robustness is formally introduced and is studied via simulation.