The Study On The Direct Torque Control System For Speed Sensorless Induction Motor

Recently with the development of electric propulsion technique many works were done on kinds of electric motor control techniques. Among them the Direct Torque Control (DTC) technique attracts more and more attention after vector control theory because of its robust characteristic, simple realization and excellent dynamic response. Sensorless speed observation can increase system redundancy effectively and can reduce the hardware design complexity, which becomes a focus of current research works. The combination of these two works is one of the directions of the future development for motor control technique.In this paper, the direct torque control system for speed senserless induction motor is studied profoundly. Three aspects are mainly studied, which are the improvement of low speed performance, the observation of stator flux and rotor speed in speed sensorless direct torque control system. The main contributions of the paper can be concluded as follows:The theory dependence of DTC is particularly analyzed and DTC method is analyzed from flux self-control and torque self-control, which illuminate the superiority of DTC system in view of the theory. On this base simulation works were done with the tools of Matlab, the results have shown that the applied control scheme gets good performance both on dynamic response and on static response.The traditional stator flux observation models are studied to indicate their advantages and disadvantages by mathematical means, and based on this an improved method is given. The simulation results have shown that the improved method, compared with the traditional one, not only improves effectively the precision of observation, but also enhances itself robust performance, which can implement flux observation in full scope of rotor speed.According to the DTC principle of induction motor and based on the optimized design for conventional Model Reference Adaptive Control (MRAC), two algorithms of speed observation based on Artificial Neural Network (ANN) and Excitation Current Vector (ECV) are given, which can create the favorable condition for lastly implementing high performance speed sensorless DTC system.Using MATLAB toolbox has designed and implemented the simulation experiment of every part of and entire speed sensorless DTC system, and the simulation results are analyzed and explained. The simulation results have shown that the speed sensorless DTC system based on the improved algorithm has good static and dynamic response in low speed scope.