A combined speed estimation scheme for indirect vector-controlled induction motors

In this paper a novel speed estimation scheme, combining sliding mode observer (SMO), model reference adaptive system (MRAS), and feedforward control, is proposed for indirect vector-controlled induction motors. Firstly, an intermediate current variable is defined to simplify the $$\Gamma $$Γ-type representation of induction motor. With the definition, a SMO is designed secondly, where the unknown terms in the current equations are replaced with the sliding mode controls. In sliding mode, the dynamics about the equivalent control components are feasible to be derived by solving the sliding mode equations. Following that another set of state equations about the equivalent control components themselves are derived, aiming to form a MRAS with the rotor speed as the adapting parameter. The references of the state variables in the MRAS are provided by filtering out the high-frequency components in the sliding mode functions in the SMO. Meanwhile, a crude value of the rotor speed, calculated directly from the equivalent control components, are fed forward into the speed adaptation mechanism in the MRAS to improve the dynamic performance of the speed estimation. As shown through simulation and experiments, this proposed combined speed observation scheme exhibits better stable and dynamic performance and satisfactory parameter robustness.