Portale di Ateneo - En.Unibs.it

The first part of the course aims to teach the student to analyse a complex dynamic system and to set up a regulator integrated system, taking into account also uncertainties. The control variables have to be optimized with respect to an objective function, with assigned constraints. Optimal control can be formulated for both continuous-time and discrete-time systems. Pontryagin maximum principle and Dynamic programming approaches are illustrated for linear and non linear systems and examples are presented and discussed. Moreover Kalman filter techniques allow to estimate the state of the system when uncertainty is not negligible. Finally the integrated control system designed has to be reliable and robust with respect to external events.
In the second part of the course Data mining techniques are illustrated with the aim to extract the maximum essential information from the analysis of big amount of data and variables. Multiple regression models, neural networks, principal components analysis, factor analysis, Granger causality analysis are introduced together with examples of real data applications.