Study of Artificial Neural Network-Based Direct Torque Control System

The direct torque control is introduced, and a controller for selecting the voltage space vector was designed with the supervised and fixed-weight neural network, the controller takes full advantage of the parallel computation, learning and fault-tolerant capability of artificial neural network (ANN) , so that it can cope with the time delay caused by the complex calculation required in traditional direct torque (DTC) and simplify the application of hardware. The simulation results show that the speed regulating system has good dynamic performance and the design is feasible.