| When the Purcell and the Bloch found the phenomenon of nuclear magnetic resonance (NMR), it inspires people to research the NMR. The technology of NMR has developed rapidly and widely in every field of life, such as physics, chemistry, material science, life sciences and medicine, etc. At present, NMR spectrum analysis technology becomes the important means which research material molecular structure and identify unknown species. It is one of the main tasks that qualitative discriminate and analysis the unknown chemical compound according to the NMR spectrum corresponding with the peak positions in the chemical displacement.Due to unknown spectrums of the complex diversity, it is not only very slow and difficult to identify an unknown species directly with the eyes. A method of fuzzy recognition for NMR spectra based on maximum principle of membership was brought forward, in order to better simply and rapidly analysis the NMR spectra of the unknown. Firstly, the concept of fuzzy set for NMR spectra to be analyzed was introduced, and membership function was given. Secondly, it avoid false peaks of the interference by making program of the Matlab, making the original spectrum data wavelet transform , feature extraction and reconstructing etc. Finally, Membership function was given based on maximum principle of membership by establishing the standard library of compounds and the algorithm of fuzzy recognition, thus qualitative analysis unknown spectrum. In this paper ,By establishing the standard library of compounds and the algorithm of fuzzy recognition, a quick automatic qualitative analysis was achieved, and 1H NMR spectra of 11 samples mixed from the four compounds of phenol, catechol, resorcinol and hydroquinone were analyzed. The results show that the compounds in the samples can all be recognized accurately under the condition that the peak positions in the spectra are even calculated roughly and that the threshold of average membership takes its value from 0.45 to 0.85. |
PDF Full Text Request