Parametrically Robust Mutual Inductance Estimation based Adaptive Control Architecture for Doubly Fed Induction Generator (DFIG)

This paper demonstrates a method for adaptive speed sensorless control of DFIG considering machine parameter variations. Proposed architecture overrides the issues related to speed and position sensors by estimating it from measured machine parameters. The adaptive control proposed here is based on system identification using Recursive Least Square identification (RLS) technique and Extended Minimum Variance Controller (EMVC). Addition of terminal voltage control loop considering the rotor current component improves the performance during dynamic system conditions and machine parametric variations. The proposed EMVC based control architecture makes the system parametrically robust along with improving dynamic stability when compared to classical minimum variance control (MVC). The proposed framework is validated both in MATLAB Simulink and Opal RT's Real Time simulation platform for a 1.5MW Doubly Fed Induction Machine setup.