A Real Time Linearization of NTC Thermistor using Hybrid Neuro-Fuzzy Logic based on VLSI Technology

Nonlinear sensors and digital solutions are used in many data acquisition system designs. As the input-output characteristic of most sensors is nonlinear in nature. Hence obtaining data from a nonlinear sensor by using an optimized device has always been a design challenge. Linearization of non linear sensor in digital environment is a vital step in the instrument signal conditioning process.This paper proposes a real time implementation of Hybrid Neuro Fuzzy Logic (HNFL) in Field Programmable Gate Array (FPGA) to linearize the natural non linear characteristics of NTC thermistor. Linearization is achieved by using Adaptive Neuro Fuzzy Inference System (ANFIS) based on Takasi-Sugeno-Kang (TSK) fuzzy inference system (FIS) whose membership functions parameters are adjusted using back propagation and/or gradient algorithms. The network training is carried out in Matlab for obtaining the optimized membership functions parameters for ANFIS implemented on a Waxwing Spartan 6 FPGA Development Board using VHDL. Single precision floating point arithmetic subroutines are developed in IEEE-754 format. Graphical programming language is used for simulation, real time data acquisition and storage.