MRAS-based Sensorless Speed Integral Backstepping Control for Induction Machine, using a Flux Backstepping Observer

Abstract

This paper presents a study concerning a sensorless vector control of an induction machine fed by a voltage source inverter. The aim is to provide a scheme to control the speed and the rotor flux using a sensorless integral backstepping control approach. The rotor speed estimation is done by an observer using the model reference adaptive system (MRAS) technique whereas the nonlinear backstepping observer is used to get the rotor flux. The main objective is to achieve a robust control, adaptive and efficient, which will allow us to test and evaluate the performance of the proposed observer, combined with a sensorless control of the induction machine. Tests and validation are done using numerical simulations with MATLAB/SIMULINK-PSB (Power System Block set) toolbox. The results show good performance in terms of robustness regarding machine parameter variations and show the excellent quality of the control law associated with the observer, despite the observability problems when the machine operates at low speed.