Fuzzy membership functions tuning for speed controller of induction motor drive: performance improvement

Abstract

Fuzzy\xa0logic controller (FLC) has gained high interest in the field of speed control of machine drives in both academic and industrial communities. This is due to the features of FLC of handling non-linearity and variations. FLC system consists of three main elements: scaling factors (SFs), membership functions (MFs), and rule-base. Fuzzy MFs can be designed with different types and sizes. For induction motor (IM) speed control, (3x3), (5x5) and (7x7) MFs are the most used MFs sizes, and normally designed based on symmetrical distribution. However, changing the width and peak position of MFs design enhance the performance. In this paper, tuning of MFs of FLC speed control of IM drives is considered. Considering (3x3), (5x5) and (7x7) MFs sizes, the widths and peak positions of these MFs are asymmetrically distributed to improve the performance of IM drive. Based on these MFs sizes, the widths and peak positions are moved toward the origin (zero), negative and positive side that produces a controller less sensitive to the small error variations. Based on simulation and performance evaluations, improvement of 5% in settling time (Ts), 0.5% in rise time and 20% of steady-state improvement achieved with the tuned MFs compared to original\xa0MFs.