Abstract
An inverse radiation analysis for simultaneous estimation of the radiative properties and the surface emissivities for a participating medium in between infinitely long parallel planes, from the knowledge of the measured temperatures and heat fluxes at the boundaries, is presented. The differential discrete ordinate method is employed to solve the radiative transfer equation. The present analysis considers three types of simple scattering phase functions. The inverse problem is solved through minimization of a performance function, which is expressed by the sum of squares of residuals between calculated and observed temperatures and heat fluxes at the boundaries. To check the performance and accuracy in retrieval, a comparison is presented between four retrieval methods, viz. Levenberg-Marquardt algorithm, genetic algorithm, artificial neural network, and the Bayesian algorithm. The results of the present analyses indicate that good precision in retrieval could be achieved by using only temperatures and heat fluxes at the boundaries. The study shows that the radiative properties of medium and surface emissivities can be retrieved even with noisy data using Bayesian retrieval algorithm and artificial neural network. Also, the results demonstrate that genetic algorithms are not efficient but are quite robust. Additionally, it is observed that an increase in the error in measurements significantly deteriorates the retrieval using the Levenberg-Marquardt algorithm.
