Abstract EN:

This thesis is devoted to the problem of enlarging the region of attraction of equilibria in power systems; we focus our attention on power systems after a symmetrical 3-phase short circuit fault and propose a new passivity-based controller design methodology for excitation control of synchronous generators. The methodology shapes the total energy function via modification of the energy transfer between the mechanical and electrical components of the system. Applying the procedure it is shown that a, properly tuned, state feedback enlarges both the estimates and the actual domain of attraction, thus increasing critical clearing time for faults. Moreover, we study the problem of identification of the network parameters and the desired equilibrium in applications of excitation control for synchronous generators. Our main contribution in this context is the construction of a new nonlinear identifier that provides asymptotically consistent estimates (with guaranteed transient bounds) of the line impedance and the equilibrium for the classical flux--decay model of a single generator connected to an infinite bus. This model is nonlinear, and nonlinearly parameterized, and the equilibria displays a nonlinearly dependence on the unknown parameters. The proposed estimator can be used, adopting a certainty equivalent approach, to make adaptive any power system stabilizer that relies on the knowledge of these parameters. Finally using the fact that a Hamiltonian system is invariant to a change of coordinates, we propose a new controller that in the new coordinates implies the electrical power as state variable.