Speed identification method in direct torque control of asynchronous machine based on neuron network theory

In this paper, a new speed identification method is presented based on a stator magnetic field oriented and neuron theory in direct torque control (DTC) system. Compared with the deficiencies of the previous traditional speed identification method, in the suggested method the speed could be fleetly estimated and transferred backward to adjust controller parameters in order to make the desired speed to approximate the real speed value. The simulation results in the matlab/simulink show that the identified speed could accurately track the actual speed. The system eliminates the speed perturbation of the traditional methods and the stator flux is round, so the DTC system has good dynamic response. In particular, the system has of great robustness when the system parameters change suddenly. The system control performance indicates that the proposed speed identification method is worth reference in engineering application of digital control system.