Desenvolvimento de uma técnica baseada em funções de Green e observadores dinâmicos para aplicação em problemas inversos

Abstract

This work proposes a new procedure for the study of inverse heat conduction problem. The dynamic observer state technique, used here, is developed to solve not only onedimensional but also three-dimensional heat transfer problem. The inverse heat conduction problem is represented by a classical inverse definition, i.e., an unknown heat flux is imposed at a front surface of a sample. The heat flux is then estimated by using the dynamic observer techniques and temperature data simulated from sensor located at the sample far from the heat source. The derivation of optimal observer equations follows directly from a novel interpretation on inverse heat conduction in the frequency domain: solving the IHCP is viewed as a filter design problem in which the reconstructed heat flux is obtained by low-pass filtering of the true heat flux. The transfer function, crucial for these techniques, is obtained here using the Green function method. This procedure allows a great flexibility to the technique and represents an easy way to apply the observer method to multidimensional problem. Comparisons with the sequential specification method and golden section technique are presented in order to evaluate the technique. The technique shows to be robust and efficient. In addition, the method also has a low processing time.