Research On The Fuzzy Direct Torque Control System

Direct Torque Control is a new control technology, it is no longer need tomake complex coordinate transform to stator current like vector control.SoDirect Torque Control reduces the impact of the motor parameters on the controltechnology. Direct Torque Control outputs stator flux and deviation signal of theelectromagnetic torque by stagnation ring comparator. In the process of selectinginverter switching voltage vector, because of the control range of voltage vectoris too large and single, and the control coefficient of PI controller is fixed, thesystem can not response quickly when rotational speed and electromagnetictorque change suddenly, which leads to the pulse of the amplitude of the startingtorque and the static electromagnetic torque too large, and have a poor control tothe dynamic control of the stator flux.Fuzzy control is applicable to multivariable, nonlinear, robust controlsystem. According to the above problems, this paper introduces the fuzzy controlinto direct torque control system. The fuzzy self-adaptive PI control replaces thePI controller. And the fuzzy self-adaptive PI control takes the dynamicadjustment of the proportional integral parameters depending on theelectromagnetic torque error. The fuzzy controller which subdivides the sector ofthe stator flux angle replaces the hysteresis comparator and grades the deviationof the stator flux and electromagnetic torque. The result input the fuzzycontroller which according to the output of the fuzzy controller to choose themost appropriate inverter switch state.This paper validates the feasibility of the proposed control strategy by thesimulation, and puts forward a simulation model for direct torque control andfuzzy direct torque control system based on the MATLAB/Simulink. Bycomparing of the two control methods, the simulation results proved that thefuzzy direct torque control reduces the starting torque amplitude and steady-statetorque ripple efficiently. The amplitude of the stator flux gets lower while the response speed gets faster. The speed curve become smooth relatively, which ismore better than in the traditional control method. The whole system has betterrobustness.