Stator Flux Based Model Reference Adaptive Observers for Rotor Position and Speed Estimation in Doubly-Fed Induction Machines

Abstract

Sensorless estimation of speed is preferred for Doubly-Fed Induction Generator (DFIG) used in wind energy conversion systems, in view of cost, maintenance, cabling and reliability issues. This paper proposes two stator flux based Model Reference Adaptive Observers (MRAO) for rotor position and speed estimation in doubly-fed induction machines. The cross product of reference and estimated stator flux linkage vectors is fed as controller input in the existing method, thereby making the observer model non-linear. The first proposed method, which is based on notch filtering, offers a performance comparable to that of the existing method, while reducing the number of voltage and current sensors required. The second method linearizes the observer model, leading to superior dynamic performance over the existing methods. This also simplifies the design of the controller. Simulations and experiments are performed on a 10 HP DFIG coupled to a Squirrel Cage Induction Motor (SCIM).