Identification and Multivariable State-Space Control of a Permanent Magnet Wind Generator

Abstract

This article presents the results of the identification method used in multivariable systems, and how, through a proper identification of these systems, a competent controller performance is obtained by feedback on the state variables implemented. There exist identification methods that are based on subspaces and have great acceptance due to the attractive features they possess. These methods provide the matrices and the initial state that represent and constitute the linear state space model of the system to control, which is discrete and time-invariant. With the multivariable system identification established, it is necessary to implement controllers that estimate and perform the respective correction of the state variables estimation in light of any disturbance, internal or external, that may affect the system at a given instant of time, and thus reduce the effect of possible uncertainties about the actual dynamics of the plant