Phd and Research

PhD and Research

I am a Ph.D. candidate in Information Engineering at the Department of Information Engineering of the University of Padova, Italy. Prof. Sandro Zampieri supervised my graduate studies and my research in the Automatic Control Group.

Here is a short overview of the research topics I'm interested in, with some references to my publications. The first ones in the list are the one in which I'm currently involved, past research interests follow.

Networked Control Systems (NCS) - see the European project FeedNetBack for some details.

Slide for the presentation of my Ph.D. thesis.

Distributed Smart Grid Control

Power distribution networks are becoming smarter (smart-grids) and are being populated by a high number of agents capable of generating power, regulating voltages, sensing the network, and compensating both reactive power and distortion. Microgenerators, wind farms, solar panels, are some examples of this kind of agents. They can be exploited to make the power distribution network to work in a smarter, more energy efficient, and more reliable way. Coordination is needed to achieve sinergy between these agents and to control them in an optimal sense.

Presentation at MIT in November 2009

See Publications [11, 13]

Consensus and clock synchronization

Distributed consensus algorithms for clock synchronization with both offset and skew correction. The algorithm is based on a PI-like structure, and relies on gossip communication between the nodes. Convergence has been proved for the asynchronous, stochastic communication protocol.

See Publications [7]

Distributed Estimation in Wireless Sensor Networks

Distributed consensus algorithms for least-square estimation of channel parameters in Wireless Sensor Networks, and for sensor calibration. A network of wireless sensors is considered, and communication between them is allowed according to a communication graph. The objective is twofold: remove sensor offsets and estimate wireless channel parameters in a completely distributed way: no central unit, no knownedge of the number of nodes, robustness to packet loss, node insertion and removal.

See Publications [5, 8]

Quantum Control

Attractivity and invariance of a quantum subspace can be achieved by designing an appropriate feedback, discrete-time, control law. The control scheme consists in a quantum measurement, which is given, and a coherent control action, which has to be designed. An algorithm has been proposed to check feasibility of the control problem and to return a stabilizing unitary control.

See Publications [9, 10, 12]

Model Predictive Control (MPC)

Stability of Generalized Predictive Control (GPC) and theory of model predictive control with finite / infinite horizon. Some stability analysis tools that are available for Optimal Control (first of all, Riccati equations) are exploited to study stability of MPC. Some smart modification of the Riccati Equations (FARE - Fake Algebraic Riccati Equation) allows to use the same tools for the analysis of finite-horizon controllers.

See Publications [1]

Design and implementation of a Model Predictive Controller for the speed and current control of a permanent magnet synchronous motor (PMSM). Electrical Drives are good testbenches for MPC, as they can be quite well modelized as linear systems with linear constraints (current, voltage, etc.). The application of MPC is therefore promising, both for driving the motor in normal operation, and even more to drive the motor in the flux-weakening operating region. The issue of fast system dynamics and little available computational time is addressed too.

See Publications [4, 6]

Adaptive control of distributed parameter systems

Adaptive control through backstepping of complex valued partial differential equations (PDE) with output feedback and boundary actuation. A parabolic PDE is considered, in which neither the advection coefficient nor the reaction coefficient are known (and they are allowed to be space-variant). The obtained controller is able to stabilize the system to zero measuring only at one end of the spatial domain and acting at the other end of the domain.

See Publications [2, 3]

Teaching

Teaching assistant for the Automatic Control course in 2009.