A Mamdani fuzzy system implementation on Arduino UNO and FPGA

Abstract

The FPGA is the robust platform to implement fast and complex circuits in the industry. Also, nowadays, Arduino microcontroller is very used because has typical preloaded functions that facilities its use. Intuitively it is known that FPGA is faster due its parallelization possibilities. The purpose of this work is to show in a quantitative way, the difference in performance between Arduino UNO and FPGA Xilinx Spartan 3E 100 CP132 when they are used as embedded Fuzzy system. This may bring arguments to decide whether to use one or another platform for this kind of applications. This paper presents a fuzzy system implementation of a Mamdani type to infer the tip given to a restaurant, where it was measured the time spent in implementing and executing; then the results were compared. These last, offered a minimal numeric difference between systems but highlighted the difference in programming time and processing time.