An Adaptive Flux Observer With Online Estimation of DC-Link Voltage and Rotor Resistance For VSI-Based Induction Motors

Abstract

An adaptive observer is proposed for concurrent estimation of rotor fluxes, unknown dc-link voltage, and rotor resistance of induction motor with voltage source inverters. For the reported flux observers and parameter estimators, the implied assumption had been that the input voltages were measured by sensors, or presumed to be known, for example while plug-in converters were utilized. However, this assumption may not be valid due to bus voltage variations, such as for battery powered servo drives, or propulsion system in hybrid electric vehicles. A drift in input voltage measurement results in state and parameter estimation errors. The proposed observer is capable of concurrent flux and dc-link voltage observation with online tuning of rotor resistance, while the phase currents and rotor speed are measured. No terminal voltage measurement is needed for this adaptive observer. The stability of the proposed observer is proved based on a Lyapunov function. Besides, it is shown that injection of a diagnosis signal is necessary to force the estimated parameters converge to the right values regardless of their initial conditions. A complex programmable logic device is implemented for the experimental setup, which controls an intelligent power module including insulated gate bipolar transistors. Extensive simulation and experimental tests verify the asymptotic convergence of the proposed observer.