Robust Modified Structured NFC Integrating with GA for Linearized Induction Motor Drive

Abstract

This paper presents a development of a simplified neuro-fuzzy control (NFC) based on genetic algorithm (GA) for optimal performance of induction motor (IM) drive using feedback linearization (FBL) approach. An intuitive linearization technique based IM is modeled and simulated in the stationary dq reference frame. The proposed simplified NFC with GA (SNFC-GA) incorporated with FBL reduces the torque ripple and improves the speed response of the IM drive. This novel technique also has the benefit of reduced computational burden by improving computational efficiency over conventional NFC and thus, suitable for real-time industrial applications. Moreover, the optimal parameters of the modified NFC are searched by GA in order to ensure the global convergence of tracking error. The effectiveness of the proposed method using linearized IM drive is investigated in simulation as well as in experiment, and it is evident that the system provides optimal dynamic performance and is robust in terms of parameter variations and external load.