The Implementation of a Fuzzy Logic Using FPGA

Süleyman Yazici, Yasar Becerikli, Suhap Sahin

Computer Engineering Department, Faculty of Engineering,
Kocaeli University, Izmit, 41040 Turkey
e-mail: {syazici, becer, suhapsahin}@kou.edu.tr

Abstract

Field Programmable Gate Arrays (FPGAs) provide advantages on hardware applications with on-system reconfigurable features by the execution of digital logic function of the designed logic circuits. VHDL is a very high definition hardware description language described in a text file that is used for the documentation, simulation and synthesis of complex digital design. For fuzzy inference systems, it is possible that speed requirements can be provided by FPGAs and flexibility requirements can be met by VHDL. In this study, FPGA based fuzzy simulation is implemented for a two input-one output function. Fuzzy systems have three modules that are fuzzifier, inference mechanism and defuzzifier modules. Fuzzy inference and defuzzifier are realized in the same function. Membership degree of inputs is obtained from a ROM using table method. Center-of-average method is used in defuzzifier module. Input and output values are represented by using 32-bit floating point number.

Keywords: FPGA, VHDL, fuzzy FPGA, floating point numbers, hardware implementation