The Fluorescence Spectroscopy Recognition of the Mineral Oil Based on the Multiresolution Orthogonal Multiwavelet Neural Network

摘要

The singular value eigenvectors of the different kinds of the mineral oil stylebooks are obtained by parameterizing the three-dimensional fluorescence spectroscopy. They are complicated and not easy to be recognized by the simple formula. The multiwavelet neural network is introduced to realize the identification of the different kinds of the mineral oil. It was layered. It had the feature of the part study. The prompting function of the network is constructed by the multiscale function and multiwavelet function. The experiment indicates that the network has all the virtue of the wavelet neural network (WNN). It also has the much better approach property than the WNN. It can effectively recognize the fine distinction between the different spectrums and realize the identification of the oil by much fewer train times than the WNN.