DTC-IM drive using adaptive neuro fuzzy inference strategy with multilevel inverter

Abstract

This paper presents the speed control of direct torque controlled 3ɸ induction motor using adaptive neuro-fuzzy inference strategy (ANFIS). ANFIS controller has been utilized to produce a reference signal for the SVPWM. The gate pulses for the 3ɸ voltage source inverter (VSI) have been obtained from SVPWM. The VSI has finally controlled the induction motor. The Simulink model for this work has been created in MATLAB. The performance exploration of the DTC-IM drive system using ANFIS has been considered, trained, and accomplished in this paper. Simulations have been done for different speeds such as 800, 1000, 1200, and 1400\xa0rpm for both conventional and five-level inverter. The simulation results have revealed that dynamic along with a transient performance of the drive has been improved using ANFIS control strategy. During the sudden variation in load torque, the machine gives good stabilization with admirable learning capability of neural networks by the use of the ANFIS controller. Moreover, the proposed five-level inverter minimizes the total harmonic distortion (THD) in the current and voltage of the inverter compared to the conventional two-level inverter. The same model has been implemented in an experimental prototype to check the feasibility of the proposed configuration.