Inverse radiation-conduction estimation of temperature-dependent emissivity using a combined method of genetic algorithmand conjugate gradient

The solution of an inverse, conduction-radiation problem in a two-dimensional rectangle is analyzed to determine the temperature-dependent emissivity at the boundary. The medium is gray, absorbing, emitting and isotropically scattering. The bounding surfaces are assumed to be opaque and diffuse. The inverse problem is solved by minimizing the performance function, which is expressed by the sum of square residuals between estimated and exact heat fluxes, using a combined method of genetic algorithm and conjugate gradient. The emissivity is assumed to be represented as a function of boundary temperature with unknown variables. Therefore, the inverse problem is treated by the estimation of these variables. Finally, four examples are presented to show the accuracy of the algorithm. The effect of the measurement errors on the accuracy of the inverse analysis is also investigated. Results show the algorithm can estimate the unknown emissivity when the measurement errors are neglected. Also it is found that increasing the measurement error decreases the accuracy of estimation of temperature-dependent emissivity.