Neuro-Fuzzy Versus Hysteresis Current Controller for a Vector Control Based Electrical Drive System

The hysteresis current controllers of a vector control based structure for an induction motor are replaced by fuzzy logic controllers, synthesized by ANFIS method. The paper makes consideration on the tuning of such a controller and proves the unexpected ability of the neuro-fuzzy controller to operate in a commutation mode. Various operating conditions are considered, the results being analyzed comparatively with the standard vector control system. The influence of the parameters, load torque, sampling period and operating conditions are taking into account performing various simulations on model. The main goal is to reduce the torque ripple and the sensitivity of the system to the parameters variation.