부산대학교 도서관

An algorithm for separating overlapping spectral components using the Tikhonov weighted regularization method is proposed. Use of the weighting function allows one to significantly reduce the regularization parameters and separate closely spaced spectral lines. The problem of the appearance of spurious oscillations in a sparse solution is solved by an iterative algorithm for correcting the main matrix. An a posteriori minimum threshold algorithm is used to determine the regularization parameter that provides the maximum resolution of the method. Use of the algorithm fundamentally improves the quality of spectra processing and increases the information content of the spectroscopic methods. The efficiency of the proposed algorithm is shown using processing of model and experimental Mossbauer spectra as examples. Keywords: Mossbauer spectroscopy, Fredholm integral equation of the 1st kind, weighted regularization, iterative algorithm.
UDC 539.1.08,519.6 Introduction. The interpretation of results from physical experiments depends in many instances on the correctness of the mathematical model used to process the experimental data. As a rule, [...]