Approximate NMPC for vehicle stability: Design, implementation and SIL testing

In this paper, Nonlinear Model Predictive Control is used to improve vehicle stability and handling by means of a rear active differential. In order to allow on-line control computations within the required sampling time, a Set Membership approximation of the designed controller is employed. The real applicability and effectiveness of such a technique, as well as the improvement over an existing control approach based on Internal Model Control, is shown through the implementation on a commercial embedded device with limited computational capacity and the testing via software-in-the-loop simulations of demanding maneuvers, using an accurate nonlinear vehicle model.