Parameter Identification Problem Of Heat Conduction Equations

The inverse problem of research field, it is very wide from various backgrounds, belongs to the category of interdisciplinary application, whether in theory or in the practical application of theoretical research in the aspects of meaning are very important. The nonlinearity and ill-posedness of inverse problems make their theory and solution more difficult than direct problems. And their field is broader. This article mainly aims at the solution to system parameters identification of heat conduction equations. The research is mainly on the improved method to the heat transfer coefficient identification problems of one-dimensional heat conduction equations. Numerical experiment results show that the inversion effect of the improved scheme is more close to the real value. The former method and improved method to the heat transfer coefficient identification problems are expanded to two-dimensional heat conduction equations. Numerical experiment results show the inversion effect of the last method is better than the former method. Image denoising also belongs to the inverse problem. The image denoising primal-dual methods based on total variation and image denoising statistical methods based on SVD are improved. By segmentation processing, a new image denoising method is put forward. Numerical experiment results show new method has obvious advantages in image denoising.