G. Sidorov et al. (Eds.): MICAI 2010, Part II, LNAI 6438, pp. 487–499, 2010. © Springer-Verlag Berlin Heidelberg 2010 FPGA Implementation of Fuzzy System with Parametric Membership Functions and Parametric Conjunctions Prometeo Cortés Antonio 1 , Ildar Batyrshin 2 , Heron Molina Lozano 1 , Luis A. Villa Vargas 1 , and Imre Rudas 3 1 National Polytechnic Institute, Mexico acorteo@hotmail.com, hmolina@cic.ipn.mx, lvilla@cic.ipn.mx 2 Mexican Petroleum Institute, Mexico batyr1@gmail.com 3 Obudu University, Hungria rudas@uni-obuda.hu Abstract. A method of FPGA implementation of fuzzy system with parametric membership functions and conjunctions is proposed. The implemented system is based on a Sugeno fuzzy model with two input variables. Fuzzy sets in the premises of the rules are given by parametric triangular membership functions and conjunction operations are defined by parametric (p)-monotone sum of ba- sic t-norms. The paper presents the hardware design of a 8-bit configurable fuzzy system, implemented on the DE2 development board from Altera using VHDL language. Keywords: FPGA, fuzzy system, conjunction, Altera, VHDL. 1 Introduction The great popularity of fuzzy systems in solving everyday problems [1,2] has created the need in hardware implementation of highly reconfigurable fuzzy systems that can be easy adopted to various applications or to change of environment where fuzzy system is operated. Such reconfigurable fuzzy systems can be developed on two lev- els: on the level of the fuzzy model and on the level of hardware implementation. This paper presents a method of hardware implementation of fuzzy systems that recon- figurable on both levels. On the level of fuzzy model we consider fuzzy systems with parametric membership functions and parametric operations. A parameterization of membership functions is a common approach to construction of fuzzy systems [3]. A parameterization of operations does not so common but also used in fuzzy modeling [4-11]. But parameterization of both fuzzy sets and fuzzy operations is a sufficiently new approach in fuzzy modeling [12]. Such parameterization of fuzzy systems gives possibility to construct highly reconfigurable fuzzy systems with high adaptive possi- bilities. On the level of hardware we consider FPGA (Field Programmable Gate Ar- ray) implementation of fuzzy systems, i.e. an easily reprogrammable integrated circuit [13-31] that can be adopted to the change of parameters of fuzzy system and more- over to the change of the structure of fuzzy system.