PhD Alumni

De Vito Daniele

Present position:

Thesis title:  Hierarchical and Multiobjective Model Predictive Control
Advisor:  Riccardo Scattolini
Research area:  Control Systems
Thesis abstract:  
The Thesis deals with the development of Model Predictive Control (MPC) algorithms, with guaranteed closed-loop stability and performance properties, for the solution of a number of important control problems in industrial applications. In particular, the first issue discussed concerns multiobjective control problems for nominal and perturbed discrete-time dynamical systems, where multiple, even competitive, performance goals need to be simultaneously fulfilled. Then, the investigation of hierarchical two-level systems, involving an upper and a number of lower controllers, all designed with MPC and working at different time scales, follows. Finally, two-level hierarchical controllers, where the high layer regulator manages a set of low level already controlled actuators and has reconfigurable capabilities in the face of actuators’ addition/replacement events, close the manuscript. In all such cases, the adoption of the MPC paradigm is justified by its main features, like the time domain definition of the control problem, its constraints handling capability or its skills in guaranteeing robustness properties, which make MPC particularly suited to cope with the problems that have been studied. Each control method addressed has also been numerically tested by means of several (possibly realistic) simulation examples.