Design and implementation of a speed adaptive flux observer for the induction motor using Sylvester criterion

Abstract

The paper presents the design and implementation of a flux and speed observer for the induction motor drive. The method employs an observer with linear feedback that yields the motor fluxes and also estimates the speed using an adaptive law. The convergence conditions are formulated based on Lyapunov's nonlinear stability theory. The gains of the observer are designed using Sylvester criterion. The method proposed is reasonably simple and yields a series of inequalities that provide insight into the design choices and stability margins of the observer. The paper shows that the observer is not globally stable - however, with proper design of the adaptation gains, it can be made locally stable. The observer with the corresponding adaptive speed estimate does not perform during low speed regeneration - the speed estimate diverges; to solve this issue, the paper shows a modified adaptation law that yields a stable speed estimate.