Linearizing E-Type Thermocouple Output Using Artificial Neural Network and Adaptive Neuro-Fuzzy Inference Systems

A high precision temperature measurement unit is designed in this study. Artificial intelligence and curve fitting techniques are used to linearize output of E-type thermocouple that is preferred due to wide operational range and high emf values. The experimental data that is essential to train and test ANN and ANFIS is obtained using Wavetek 9100 calibration unit. After the successful training completion, the temperature measurement unit is realized and ANN and ANFIS based real time temperature measurement is individually achieved using thermocouple output voltage.