Adaptive control of the three-phase squirrel cage induction motor with load torque estimator

The performances of the adjustable electric drive systems are directly related to the accurate knowledge of the parameters. It is well known that the resistances are temperature dependent and the inductances are function on the magnetising B-H curve of the ferromagnetic material and the saturation effect. In order to obtain a high performance electric drive system, an adaptive structure in direct form (the controller parameters are obtained on-line) is proposed in this paper. The adaptive mechanism of the parameters is obtained by adding two terms: the first one will support a gradient adjustment law (assuring the asymptotic performances) and the second one will comport an adjustment that includes a sigmoid function specific for variable structure control. The last component improves the transient and steady state response by eliminating the small oscillations of the closed loop response around the equilibrium state in order to obtain a zero tracking error. The adaptive mechanism assures the robustness to the external disturbances and to the unmodeled dynamics. Another issue presented in this paper is an original load torque method. The load torque estimator is based on the transversal current stator component and the measured rotor speed. This method presents certain advantages instead of using measured method, which requires a low pass filter: no additional hardware, and faster control response. Matlab/Simulink based simulation results will show the validity of the proposed solution.