Abstract
This article presents the Adaptive Neuro-Fuzzy Inference System based Direct Torque Control (ANFIS-DTC) for permanent magnet synchronous motor coupled with centrifugal pump. Direct Torque Control (DTC) is an inherent closed loop control with very less complexity, dynamic torque and speed response in comparison with the other vector control techniques. In the variable torque applications like pumps when speed of motor varies, the torque developed by the pump also varies correspondingly. To reduce the ripples in torque and to improve the response of the control algorithm, Adaptive Neuro-Fuzzy Inference System (ANFIS) is implemented along with DTC. In contrast with the conventional DTC and the Fuzzy logic based DTC, the proposed ANFIS based DTC has significantly reduced ripples in flux, torque and stator current. The results of the proposed ANFIS-DTC are validated through the Matlab simulations and the performance of the system is found satisfactory when it is tested with different rotational speeds.