Algorithm Research Of Object Spectral Emissivity Based On The Multispectral Radiation Measurement

Spectral emissivity is an important parameter to study the infrared radiation characteristics of an object, which can more comprehensively reflect the law of the infrared radiation of the object. It plays an important role in industrial production, military defense,aerospace and other fields. This paper uses the advanced Fourier transform infrared spectrum radiometer measurement system to measure the spectral radiance of radiation source to the target. Combined with the infrared transmission theory and based on the background, the reference blackbody, the target source, the measured data is compared and analyzed to research target source spectral radiance of modeling method. The spectral characteristics of the full radiance are obtained in the measurement band to calculate spectral emissivity. The main contents of this paper are as follow.Emissivity measuring system and test scheme are designed. Combined with the infrared theory, this paper uses the energy method through compared radiance of target with the reference blackbody in the same temperature and background to calculate the spectral emissivity of the target. Meanwhile, considering the response of the spectral radiometer to the environmental radiation and the influence of the internal radiation, this paper uses the double temperature calibration method to calibrate the response function of the radiometer and stray radiation of environment. Radiance experimental data of the target is pretreated.The infrared spectral radiance model based on the least squares is established and spectral emissivity is calculated. Combined with knowledge such as infrared transmission atmospheric window, the invalid spectral radiant intensity data is eliminated and the spectral radiance of little effect under the atmospheric window data is selected as a sample. Assuming that there is a high order polynomial relationship between spectral radiance and wavelength,this paper uses the least square algorithm to establish the infrared spectral radiance model of3?m~14?m band, final according to the energy comparison method, calculate the spectral emissivity.The infrared spectral radiance model based on the RBF neural network is established and spectral emissivity is calculated. The wavelength is selected as the input of RBF neural network, the corresponding radiance as output, then the spectral radiance model is established.Ditto, the spectral emissivity of 3?m~14?m band of the target is calculated. Experimental results show that the two kinds of modeling method can reflect the relationship between the spectral radiance and the wavelength. Compared with the theory of emissivity, the calculation precision of RBF modeling method is superior to least square method and improves the measurement precision of the spectral emissivity.