Performance of MRAS Based Speed Estimators for Grid Connected Doubly Fed Induction Machines During Voltage Dips

Abstract

Rotor speed is required for the independent control of active power, reactive power and to achieve maximum power and also for pitch control. Speed estimation is preferred rather than sensing from accuracy, cost and reliability point of view. Different Model Reference Adaptive System (MRAS) based speed estimation techniques are available for grid connected doubly fed induction machines. Performance of all the MRAS speed estimators has been studied under normal grid voltage. As per the new grid codes, wind turbine should remain connected to the grid during severe voltage dips [1]. Hence the performance of these speed estimators should also be studied under voltage dips. This paper compares the performances of different MRAS based estimators during voltage dips up to 50% and 5% of the nominal grid voltage. This paper also shows the effect of resistance variation on speed estimation during above mentioned operating conditions. Simulation has been performed in MATLAB/SIMULINK and the results are presented for various speed estimators.