The PhD work investigated the control problem arising in complex powertrains through the use of unified model predictive control approach. In order to meet the tight real-time requirements of automotive control, the Constrained Parameterized Approach (CPA) is used. The control application concerns two different powertrains: a mild-hybrid powertrain equipped with an automated manual transmission, and a high downsized powertrain equipped with a manual transmission. In the first application, the proposed strategy totally controls the powertrain dynamic (engine and transmission), while ensuring transparency w.r.t the driver demand. Whereas, in the second application that concerns a manual transmission, the proposed control ensures the "driving assistance", by assuming an advanced control of the engine torque. As part of the thesis, a phenomenological clutch model has been developed to be used by the control algorithm.
