Research areas

Parameter estimation in the set-membership context

Parameter estimation in the set-membership context deals with the identification of a set of all possible parameters of a system consistent with the measurements, the assumed model structure and the hypothesis on the error. Differently from the stochastic identification framework, in the set-membership context the statistical description of the noise on the measurements is not available and the only information on such an error is that its amplitude or energy is bounded. In my PhD. thesis, I've focused on the problem of identifying: (i) linear systems when both input and output data are corrupted by bounded noise (bounded errors-in-variables problem); (ii) Hammerstein systems; (iii) Hammerstein-like and Wiener-like structures with backlash; (iv) block-structured nonlinear feedback systems. The considered identification problems have been in terms of nonconvex polynomial optimization and suitable convex-relaxation approaches have been proposed to efficiently compute guaranteed bounds on the parameters of the considered system.

Some related publications:

D. Piga, A convex relaxation approach to set-membership identification, Ph.D. Thesis, Politecnico di Torino, Torino, Italy, 2012.

V. Cerone, D. Piga, D. Regruto, Computational load reduction in bounded error identification of Hammerstein systems, IEEE Transactions on Automatic Control, Vol. 58, No. 5, pp. 1317-1322, 2013.

V. Cerone, D. Piga, D. Regruto, Bounded error identification of Hammerstein systems through sparse polynomial optimization, Automatica, vol. 48, n. 10, pp. 2693-2698, 2012.

V. Cerone, D. Piga, D. Regruto, Set-Membership Error-in-Variables Identification Through Convex Relaxation Techniques , IEEE Transactions on Automatic Control, Vol. 57, No. 2, pp. 517-522, 2012.

V. Cerone, D. Piga, D. Regruto, V. Cerone, D. Piga, D. Regruto, Bounding the parameters of block-structured nonlinear feedback systems, International Journal of Robust and Nonlinear Control, vol. 23, n. 1, pp. 33-47, 2013.

V. Cerone, D. Piga, D. Regruto, Bounded-Error Identification of Linear Systems with Input and Output Backlash , Proc. of the 16th IFAC Symposium on System Identification, Brussels, Belgium, 2012.




Identification of linear-parameter-varying models

Linear-parameter varying (LPV) models belong to the more general class of linear time-varying systems and, roughly speaking, they can be defined as linear systems where, either the matrixes of the state equations or the coefficients of the input-output relation, depend on one or more time varying parameters, whose real-time samples are assumed to be available. We've proposed two different approaches to identify LPV models when both the output data and the scheduling parameters are corrupted by bounded-noise. The proposed identification schemes have been applied to derive an LPV model describing vehicle lateral dynamics (based on experimental data provided by FIAT) and to derive a dynamical model describing the glucose-insulin dynamics in diabetic patients.

Some related publications:

V. Cerone, D. Piga, D. Regruto, V. Cerone, D. Piga, D. Regruto, A convex relaxation approach to Set-membership identication of LPV systems, Automatica, Vol. 49, No. 9, pp. 2853-2859, 2013.

V. Cerone, D. Piga, D. Regruto, Set-membership LPV model identification of vehicle lateral dynamics , Automatica, Vol. 47, No. 8, pp. 1794-1799, 2011.

V. Cerone, D. Piga, D. Regruto, B. Sintayehu, LPV Identification of the Glucose-Insulin Dynamics in Type I Diabetes , Proc. of the 16th IFAC Symposium on System Identification, Brussels, Belgium, 2012.




Enforcing stability constraints in system identification

Although many times the system to be identified is surely known to be stable, most of the identification techniques do not exploit such a prior information in the definition of the assumed model structure. As a result, the identification procedure may give rise to inaccurate models and even instability may arise, especially in the presence of shortage of data, modeling error and high measurement noise. The papers reported below address the problem of enforcing stability constraints in identifying linear-time-invariant (LTI) systems and linear-parameter varying (LPV) .

Some related publications:

V. Cerone, D. Piga, D. Regruto, Enforcing stability constraints in set-membership identification of linear dynamic systems, Automatica, Vol. 47, No. 11, pp. 2488-2494, 2011.

V. Cerone, D. Piga, D. Regruto, R. Tóth, Input-Output LPV Model Identification with Guaranteed Quadratic Stability , Proc. of the 16th IFAC Symposium on System Identification, Brussels, Belgium, 2012.




High-altitude wind power generation

The key idea of high-altitude wind power generation is to capture the high-altitude wind flows using tethered airfoils (e.g. power kites used for surfing or sailing), linked to the ground with one or more cables which are employed to control their flight and to convert the aerodynamical forces into mechanical and electrical power, using suitable rotating mechanisms and electric generators kept at ground level. The airfoils are able to exploit wind flows at higher altitudes than those of wind towers (up to 1000 m), where stronger and more constant wind can be found basically everywhere in the world. Each airfoil is equipped with on-board sensors, and other sensors are installed at ground level to monitor the generated energy and the wind conditions. A nonlinear model predictive control (NMPC) strategy has been proposed to maximize the generated energy while explicitly taking into account the state and input constraints, related to actuator limitations and to the need of preventing the airfoil from falling to the ground and the lines from entangling. The obtained results are reported in the papers below, including theoretical analysis, numerical simulations and experimental tests with a small-scale prototype built at Politecnico di Torino in collaboration with the company Sequoia Automation.

Some related publications:

D. Piga, “Performance analysis of KiteGen system: high-altitude wind power generation , Msc. Thesis (advisor: Prof. M. Milanese), Politecnico di Torino, Torino, Italy, 2008.

L. Fagiano, M. Milanese, D. Piga, High-Altitude Wind Power Generation , IEEE Transactions on Energy Conversion, vol. 25, pp. 168-180, 2010.

.