ELSEVIER Fuzzy Sets and Systems93 (1998)23-28 FUZ2Y sets and systems A PI-type fuzzy controller with self-tuning scaling factors Hung-Yuan Chung*, Bor-Chin Chen, Jin-Jye Lin Institute of Electrical Engineering, National Central University, Chung-Li 32054, Taiwan, ROC Received December 1994;revisedJuly 1996 Abstract In this paper, a self-tuning fuzzy controller with a smart and easy structure is presented. The tuning schemes allow us to tune the scaling factors by only seven rules. The aim of a controller is to obtain a satisfactory performance, such as the rise time, the overshoot, the steady-state error for the step response. Simulation results of a third order system with time-delay show a better performance and ability of curbing the effect of load disturbance on the control process. © 1998 Elsevier Science B.V. Keywords: Fuzzy; Self-tuning; Scaling factor 1. Introduction During the past several years, the fuzzy control has become one of the most active and fruitful fields for research in the application of fuzzy set theory. A fuzzy system allows a gradual and continuous transition, say, from 0 to 1, rather than a crisp and abrupt change between binary values of 0 and 1. There are certain particular characteristics of fuzzy systems that give better performance for specific applications. In general, fuzzy systems are suitable for uncertain or approximate reasoning, especially for systems whose mathematical models are diffi- cult [5]. A fuzzy controller composed of control rules of conditional linguistic statements on the relation- ship between the input and output variables has *Corresponding author. the advantages of (1) emulating the behavior of a human operator and (2) dealing with model un- certainty. In the design of a fuzzy controller, the definition of membership functions and the estab- lishment of control rules are usually subjective. Besides, there is still no standard method in tuning the scaling factors. It thus follows that a self-tuning fuzzy controller (STFC) is designed by self-adjust- ing the control rules, the membership functions, or the scaling factors or the union. Among them, the control rules and the scaling factors play more important roles I-1-4, 6, 7]. In [2] the authors pro- posed a real-time tuning for both input and output scaling factors. In [7] the authors proposed a switching-type fuzzy controller by switching the scaling factors in each region. In this paper, we propose a new and easy struc- ture to tune the input and output scaling factors. We only use seven tuning rules to tune the scaling factors. In Section 2, a review of PI-type and 0165-0114/98/$19.00 © 1998 ElsevierScienceB.V. All rights reserved PI1 S0165-01 14(96)00215-1