Multi-spectral radiation thermometry based on the reconstructed spectral emissivity model

Abstract

With the aim of solving the problem of detecting the true temperature of objects using multi-spectral radiation thermometry, this paper proposes a multi-spectral radiation thermometry method based on the reconstructed spectral emissivity model. Compared with traditional hypothetical spectral emissivity models, this model is a reconstruction emissivity model based on actual measured brightness and known wavelength quantities, which can improve the characterization of the real emissivity state of the measured object and not be affected by any hypothetical conditions or artificial preset parameters. In order to verify the effectiveness of the proposed method for detecting the temperature of real problems, a total of eight typical spectral emissivity models, spectral data of Si3N4 ceramics and ZrO2 ceramics, and measurement data of rocket motor casings are selected for experiments. The experimental results indicate that the true temperature calculated by the proposed method has a relative error of less than 1 %, and the calculation time is kept at the millisecond level, which meets the real-time multi-spectral temperature measurement requirements.