Inverse Estimation for Heat Flux Distribution at the Metal-Mold Interface Using Fuzzy Inference

Abstract

This study provides a new inverse approach based on fuzzy inference for solving the problem of estimating heat flux distribution at the metal-mold interface in the continuous casting process. Measured temperatures acquired with the thermocouples buried in the mold are used to obtain corresponding inference results with the fuzzy inference. Then according to the importance of measured information for estimating the heat flux distribution, inference results are weighted to realize estimation of heat flux distribution at the metal-mold interface. Some numerical tests are presented to discuss the validity of the present approach by using different initial guesses of heat flux distribution, the number of measuring points, and measurement errors. In comparison with the conjugate gradient method, it is concluded that the method based on fuzzy inference is of a good anti-ill-posed characteristic.