A multi-stage optimization technique for simultaneous reconstruction of infrared optical and thermophysical parameters in semitransparent media

A multi-stage optimization technique is proposed to simultaneously reconstruct the infrared optical and thermophysical parameters in semitransparent media. The coupled radiative-conductive heat transfer in two-dimensional absorbing, scattering and emitting medium is solved by the discrete ordinate method combined with finite volume method. The exit radiative intensity and temperature distribution on the boundary are served as input for the inverse analysis, and the sequential quadratic programming is used as the inverse technique. Since the measurement signals are much more sensitive to the infrared absorption and scattering coefficients than to the thermal conductivity of medium, the thermophysical property cannot be accurately reconstructed by the conventional method. The multi-stage optimization technique is developed to solve the inverse estimation tasks, through which the optical and thermophysical parameters are reconstructed in different stages based on different objective functions. All the retrieval results demonstrate that the multi-stage optimization technique is robust and effective in simultaneous estimation of absorption coefficient, scattering coefficient and thermal conductivity. The optical and thermophysical parameters can be reconstructed accurately even with measurement errors.