Dynamic Switching of Fuzzy Resolution in Knowledge-Based Self-Tuning Control

Article type: Research Article

Authors: Wu, Q.M. Jonathan | de Silva, Clarence W.

Affiliations: Institute for Sensor and Control Technology, National Research Council, Vancouver, Canada | Department of Mechanical Engineering, University of British Columbia, Vancouver, Canada, e-mail: desilva@mech.ubc.ca

Abstract: Self-tuning of a control system may be effectively carried out by means of a properly designed, fuzzy, decision-making system. Optimization of the operation of such knowledge-based systems requires consideration of several factors including proper selection of fuzzy resolution, shape of the membership functions of fuzzy variables, and the level of overlap among various fuzzy states of a variable. This article studies the effect of fuzzy resolution on the processing speed and response accuracy of a fuzzy logic based control or tuning system. First, the underlying analytical concepts are presented, and on that basis, experiments are carried out to study this effect of fuzzy resolution. Response of the experimental system is studied at different levels of fuzzy resolution. Next, the concept of resolution switching is introduced and incorporated into the experimental system. The results obtained with and without resolution switching are comparatively studied to demonstrate the effectiveness of resolution switching.

DOI: 10.3233/IFS-1996-4106

Journal: Journal of Intelligent and Fuzzy Systems, vol. 4, no. 1, pp. 75-85, 1996