Application of artificial neural network in the efficient control of three-phase induction motor

This work presents a method for increasing the efficiency of three-phase induction motor drives over the entire operation range. The direct field oriented control of induction motors, including the effects of magnetic saturation is used. The magnetic saturation effect in the machine is modeled by the non-linear magnetization curve of the iron core. Artificial neural networks are used to predict the optimum reference rotor flux to be used in vector control. Details about the chosen neural networks are given. Simulation and experimental results are presented and the motor losses reduction during different load conditions is evaluated.