Robotics and Autonomous Systems 33 (2000) 65–88 Systematic design and analysis of fuzzy-logic control and application to robotics, Part I. Modeling M. Reza Emami, Andrew A. Goldenberg, ˙ I. Burhan T ¨ urk¸ sen * Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ont., Canada M5S 3G8 Received 1 January 1999; received in revised form 1 May 1999; accepted 1 December 1999 Abstract A systematic methodology for synthesis and analysis of fuzzy-logic controllers is proposed in this paper (Part I) and its follow up (Part II) [M.R. Emami, et al., Robotics and Autonomous Systems 33 (2000) 89–108]. A robust model-based control structure is suggested that includes a fuzzy-logic inverse dynamics model and several robust fuzzy control rules. The model encapsulates the knowledge of the system dynamics in the form of IF–THEN rules. The paper focuses on how to obtain this knowledge systematically from the input–output data of a complex system; one that is ill-defined or contains complicated phenomena that are difficult to interpret analytically. All practical steps, from data acquisition to model validation, are illustrated using a four degree-of-freedom robot manipulator. Comparing the results with those of a complete analytical model and a heuristic fuzzy modeling technique illustrates the strength of the proposed methodology in terms of capturing effects that are difficult to model. In the follow-up paper, this model is used in the proposed control structure. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Fuzzy-logic control; System identification; Fuzzy systems; Robot manipulator 1. Introduction The goal of developing simple and efficient con- trol strategies for complex systems triggered the emer- gence of a new field of research known as fuzzy-logic control (FLC). The new area originated from the sem- inal work of Zadeh on fuzzy algorithms [2], which introduced the idea of formulating control algorithms using the logical IF–THEN rules. Since the earliest ef- forts by Mamdani and his coworkers [3], most of the related research has focused on practical implemen- * Corresponding author. Tel.: +1-416-978-1278; fax: +1-416-978-3454. E-mail address: turksen@mie.utoronto.ca ( ˙ I.B. T ¨ urk¸ sen). tations of fuzzy controllers. Successful results have been reported in a wide range of applications. Despite the diversity of approaches used in the development of fuzzy controllers, most of them are designed based on “trial-and-error”. Although this could be effective in some cases, it limits the raise of systematic approaches to fuzzy-logic control. One approach to synthesis and analysis of the fuzzy-logic systems is to consider the FLC as a par- ticular class of nonlinear systems, and to apply tools from the classical nonlinear control systems theory. A promising approach in this direction is based on the fact that the FLC is a variable structure system (VSS). Although some of the structures of the VSS may necessarily not be stable, their combination 0921-8890/00/$ – see front matter © 2000 Elsevier Science B.V. All rights reserved. PII:S0921-8890(00)00080-4