International Journal of Emerging Trends in Engineering and Development Issue 3, Vol.2 (March 2013) Available online on http://www.rspublication.com/ijeted/ijeted_index.htm ISSN 2249-6149

*MIC College of Technology, Kanchikacherla ,Vijayawada ______________________________________________________________________________________ Abstract: During the last decade a new control method called DTC (Direct torque control) has been developed for electrical machines with conventional PI controller. The conventional PI controller is slow in adapting to speed changes, load disturbances and parameter variations without continuous tuning of its gains. In the conventional approach gains of the PI controller need to be tuned manually according to the load disturbances and parameter variations. For this reason the PI controller is replaced by fuzzy PI controller. In fuzzy PI controller gains are varied automatically according to the predefined set of fuzzy rules. The performance of the fuzzy PI controller & DTC-Artificial Neural Networks (ANN) is tested through MATLAB /SIMULINK. In comparison with conventional PI controller, fuzzy PI controller, show that with the use of ANN-DTC reduced the torque, stator flux and current ripples.