针对基于光谱数据进行温度与发射率分离过程中存在的n个方程包含n+1个未知数这一欠定问题, 提出利用牛顿迭代法来实现材料表面真温及发射率的反演计算, 通过给定温度和发射率初值, 利用泰勒级数的线性项建立迭代公式, 通过迭代得到温度和发射率的近似解。 分别利用理论热辐射谱和腔黑体的实验数据进行验证, 结果表明, 任意给定温度和发射率初值均可获得与真实温度接近的计算温度值, 相对误差小于0.09。 发射率反演结果与真实发射率线形一致, 当温度和发射率初值越接近真实温度和发射率时, 发射率反演结果越精确。 该方法消除了发射率假定模型限制, 有望应用于高温及超高温下各种材料真实温度和光谱发射率研究。

In the separation process of temperature and emissivity from spectra data, there is an underdetermined problem, N equation containing N+1 unknown number. In this paper, Newton’s method is used to inversion calculate the surface temperature and the emissivity of the material. The iterative formula is established with Taylor series linear terms. Given the initial temperature and emissivity, an approximate solution of temperature and emissivity is obtained through iteration. In this paper, the Theoretical value of thermal radiation and the experimental data of cavity blackbody were used to verify ,the results show that the inversion temperature obtained with the given any initial values of temperature and emissivity are close to the true temperature. The relative error is lower than 0.09. For different emissivity models, the lineshape of the inversion emissivity is consistent with that of the true emissivity. The initial values of temperature and emissivity are closer to the true temperature and emissivity, which gives rise to more accurate inversion results. This method eliminates the emissivity assumed model limits. It is expected to be used the inversions of the surface temperature and the emissivity of the material based on the thermic spectra in high temperature and ultra high temperature.