A New Flux Observer for Induction Motors with On-Line Identification of Load Torque and Resistances

Abstract

AbstractUnder the assumption that the stator currents and voltages are available from measurements along with the rotor speed, a novel local adaptive flux observer for induction motors is presented which, under persistency of excitation, estimates the motor fluxes and identifies on-line the rotor resistance. A sufficiently slow adaptation for the stator resistance is performed in order to isolate its estimation from the estimation of motor fluxes and rotor resistance. A second-order load torque identifier is also designed to be used in conjunction with the adaptive flux observer and the stator resistance identifier. The advantages of the proposed observer with respect to previous work are: the assumption of bounded stator currents integrals is no longer required; the dynamic order of the adaptive flux observer is reduced to six; a clear physical interpretation of the persistency of excitation condition is given. On the other hand the convergence analysis of the error dynamics holds only locally under suitable identifiability conditions while globally convergent algorithms are available in the literature.