Nonlinearity Correction of the Thermocouple Based on Neural Network

Thermocouples are widely used in the industrial measurement. However they have serious nonlinear problems in the larger range of measuring temperature, and result in the big error in the measuring results. Taking nickel chromium-Silicon nickel thermocouple (Type K) as an example, BP (Back Propagation) neural network technology was used to conduct corrections on the nonlinear input and output characteristics of the thermocouple. The C program on the nonlinearity correction model of the thermocouple was given and could be used in MCU (Micro Controller Unit).