A rotor-flux-observer based neural speed controller for high performance induction motor drives

This paper focuses on the control of a highly nonlinear induction motor in which all state variables are not accessible and information required for the generation of the control signal has to be obtained from the observed input/output. An indirect adaptive control system for an induction motor based on a rotor-flux-observer is introduced. The control system consists of two multilayer feedforward neural networks: a tracker identification neural network which captures the nonlinear dynamics of the motor over any arbitrary time interval in its range of operation and estimates the immeasurable components of the rotor flux terms; and a control neural network to provide the necessary control actions to achieve trajectory tracking of the rotor speed. An adaptive learning mechanism is proposed. The two networks are trained in an online mode utilizing the adaptive learning mechanism. This simplifies the learning algorithm in terms of computation time, which is of special importance in real-time implementation