Adaptive control strategy with flux reference optimization for sensorless induction motors

Abstract Avoiding mechanical (speed, torque) sensors in electric motor control entails cost reduction and reliability improvement. Furthermore, sensorless controllers (also referred to output-feedback) are useful, even in the presence of mechanical sensors, to implement fault tolerant control strategies. In this paper, we deal with the problem of output-feedback control for induction motors. The solutions proposed so far have been developed based on the assumption that the machine magnetic circuit characteristic is linear. Ignoring magnetic saturation makes it not possible to meet optimal operation conditions in the presence of wide range speed and load torque variations. Presently, an output-feedback control strategy is developed on the basis of a motor model that accounts for magnetic saturation. The control strategy includes an optimal flux reference generator, designed in order to optimize energy consumption, and an output-feedback designed using the backstepping technique to meet tight speed regulation in the presence of wide range changes in speed reference and load torque. The controller sensorless feature is achieved using an adaptive observer providing the controller with online estimates of the mechanical variables. Adaptation is resorted to cope with the system parameter uncertainty. The controller performances are theoretically analyzed and illustrated by simulation.