A Speed Sensorless Direct Torque Control System Based On Intelligent Algorithms

Speed sensorless direct torque control system has been a focus in thefield of alternating current drive since it came into being. In direct torquecontrol system, speed sensors are usually used to survey the speed of motor,but speed sensors have many disadvantages, such as inconvenient fixing andmaintenance, inaccurate precision. At present it is an efficient and feasiblemethod to estimate speed by using software instead of hardware. The design ofspeed sensorless optimizes the system structure, reduces the maintenanceexpense, improves the stability and reliability of direct torque control systemand has more extensive application prospect.In the paper, a speed estimator based on artificial neural network isdesigned according to the principle of direct torque control, the speedestimator is optimized by genetic algorithm, and the parameters of PID speedadjuster is also optimized by genetic algorithm. A stator linkage and torqueobserver based on stator resistance estimator using fuzzy neural network isalso designed. Eventually, the intelligent control of direct torque controlsystem is implemented.On the basis of theory research, software and hardware design, the systemexperiment simulation in MATLAB toolbox is given, and the simulation results are analyzed and explained. The simulation results show the system has good static and dynamic performance.The optimized speed estimator by genetic algorithm and the stator linkage observer are applied successfully in direct torque control system. The successful application indicates the intelligent algorithms have better prospect to implement digital and intelligent control in the domain of automation control in the future.