Sensorless Direct Torque Control of Induction Motor Using Neural Network-Based Duty Ratio Controller

Abstract

Improvements in the sensorless direct torque control of an induction motor by employing fuzzy logic switching controller (FLSC) plus neural network-based duty ratio controller (NNDRC) are explained in this paper. The conventional direct torque control (CDTC) of an induction motor suffers from major drawbacks such as high ripples in motor torque and flux response, poor performance during low speed and starting, and switching frequency variations due to hysteresis bands. Duty ratio controller evaluates the time for which active switching state is applied (δ), and for the remaining time period, zero switching vector is applied. The simulation results show that by using FLSC with NNDRC, considerable reduction in torque and flux ripples and improvement in dynamic response of drive compared to CDTC are achieved.