DIRECT TORQUE CONTROL USING NEURAL NETWORK APPROACH

Direct Torque Control (DTC) is one of the latest technique to control the speed of motor, in this paper, the control technique of DTC is based on when load changes then inverter switch position are changed and supply to the motor is changed, in this paper Proportional Integral (PI), Neural Network (NN) controller and Adaptive motor model is designed this is the heart of the DTC, as we know that DTC doesn’t require any feedback and sensors to measure. The NN structure is to be implemented by input output (nonlinear) mapping models and is constructed with input, output and hidden layers of sigmoid activation functions. It has been introduced as a possible solution to the real multivariate interpolation problem. To improve the performance of DTC with the modern technique using NN approach is implemented, and performance of DTC with PI controller and NN controller is done, hence, the NN approach shows the better performance than conventional PI controller.