Sliding mode control of a three phase induction motor based on RBF neural network

Abstract

A stator-flux-oriented vector controller of induction motor is often used in the controllers due to less depending on parameters of the motors, the parameters of the motors are nonlinear and time-varying solution conditions slip control will be applied by the brilliant advantages of stability control is slipping sustainable and as soon as the system noise. On the other hand, when the parameters of nonlinear objects changes over time, the problems keep constant speed when the load changes are difficult to implement, therefore the neural network is used to identify the speed of machines are needed to increase the stability control system. This article presents a new method of designing sliding mode controller based on radial basic function network (RBF) for three-phase asynchronous motors based a stator-flux-oriented vector controller. Two sliding controllers are designed independently for stator-flux-vector estimation and torque, in which magnetic flux is estimated and the speed of the motor is identified by RBF network, combined seven - level cascade inverter with a reduction common-mode algorithm applied to increase the stability for the controller. Simulations and experiments using Matlab / Simulink for 1-hp induction motor drive, typed 150-rad/s squirrel cage rotor, the results present velocity followed setting values at the frequency change from the lowest 50-rad/s to 150 rad highest/s, the system is still stable when the stator is changed stator resistance and rotor resistance up to 1.5 times the original value.