Induction Motor Direct Torque Control Performance At Low Speed Optimization Research

With its relatively simple and direct method as well as the high-performance effect, direct torque control of induction motor has been gaining increasingly more applications in the field of modern AC transmission. However, the characteristics of self-control resulted in such problems as large torque ripple, inaccurate flux observation and unstable and low switching frequency when working at low speed. These flaws compromised the overall controlling performance and consequently hindered its further generalization. It is, therefore, of top priority to work out an optimization strategy that can improve the low-speed performance so as to solve the problems.Based on the mathematical models of induction motor and inverter, this article begins with an analysis of the principle of the voltage space vector, and proceeds to elaborate the basic idea of direct torque control system. A simulation study on the system is then carried out by means of Matlab/Simulink, from which the causes of the low-speed defects of direct torque control are inferred, thus providing a foundation for the optimization.Taking into consideration the complicated, variable and strong coupling performance of the induction motor control system at low speed, fuzzy control method is employed to carry out the optimization. By designing the P-fuzzy self-tuning PID dual stage control strategy of the speed regulator and low-pass filtering the flux observer, the estimation errors during the pure integral phase of the U-I model can be avoided. Besides, fuzzy controllers are devised for flux and torque to replace the conventional two-point regulator, and the voltage space vector in use is fuzzy-selected. The simulation results have validated the effectiveness of this method.In view of the torque adjustment ripple that easily comes up as a result of single voltage vector at low speed, it is decided to use space vector modulation method to synthesize the desired voltage vector. And a system back-stepping controller is worked out, so that the desired voltage vector can be precisely calculated according to the flux and torque errors. This method is effective in reducing the torque ripple, decreasing the stator error and improving the system robustness. Its effectiveness has been proved by simulation.