Correction of second-order adaptive modeling method of infrared temperature measurement

This paper is aimed at building the second-order adaptive model of infrared temperature measurement based on the experimental data of standard blackbody. After temperature measurement of blackbody at different distance, use neural network method to learn the underlying regulations of measuring data adaptively. Set up the functional relationship between the measuring temperature and the current measuring distance and its variation. Then the atmospheric transmittance is calculated online to correct the measured target temperature, so the actual target temperature can be accurately calculated. Simulation results show that the model built in this paper could effectively learn the sample information, reduce the surface temperature measurement errors effectively and improve the accuracy of infrared testing.