Linearization of Temperature Sensors (K-Type Thermocouple) Using Polynomial Non-Linear Regression Technique and an IoT-Based Data Logger Interface

Thermocouples require Cold Junction Compensation (CJC) and are widely used temperature sensors. These thermocouple types have high nonlinearities and consequently a linearization technique using higher degree polynomial based on nonlinear regression is proposed. A K-type thermocouple is linearized using its inverse mathematical model. The percentage nonlinearity of the thermocouple was reduced from 2.03% to 0.02% Full Scale Span (FSS) in a range of -100 to 1372 degrees C. Thermistors are used for CJC in thermocouples and are highly nonlinear. A constant current is passed through the thermistor, and the drop is measured by an Analog to Digital Converter. The inverse mathematical model obtained from experimental data is cascaded to linearize the thermistor output. The percentage nonlinearity of the thermistor was reduced from 84.63% to 0.27% FSS for a range of 0 to 120 degrees C in experiment. The linearized thermistor was used for CJC of the thermocouple. Maximum absolute error observed in the temperature measurement using Yokogawa calibrator and thermocouple along with CJC were 0.24 degrees C and 0.43 degrees C respectively. A remote data-logger was used to store measured data. Stored data could be viewed and used for further processing from any Internet of Things user interface. The proposed temperature measurement system was found to be smart and reliable.