Neural block control with input constraints for induction motors

In this paper we present a novel approach to control induction motor, considering the switching nature of their electrical drivers. Modifying published results for nonlinear identification using dynamic neural networks, we propose a neural network identifier. The controller is derived using the neural block control technique and sliding modes control. Such control law is a discontinuous one, which just takes values from the look-up table given by the inverter circuit connected to the stator windings, This control strategy is motivated by the wearing that suffer the switching electronic devices, when a PWM approach is used at high frequency operation.