Adaptive speed control for induction motors based on a semi-current-fed model

This paper presents a semi-current-fed concept and a virtual desired variable synthesis for induction motor control. First, the traditional ideal current-loop assumption is relaxed. The relaxed assumption, considering more practical situations, allows a bounded stator voltage and a finite absolute-integral of current tracking error. From this novel concept, a semi-current-fed model for induction motors is used for the design. Both unknown rotor resistance and torque load are considered along with immeasurable rotor flux. To cope with the problems, we use a virtual desired variable approach to synthesize the control law. In the absence of a flux sensor and not using a flux observer, the asymptotic speed tracking is ensured without needing the persistent excitation condition. However, if the controlled system is persistently excited, an optimal torque control is additionally achieved. In addition to the theoretical proof, experimental results show the expected tracking performance.