Mixture Raman spectrum qualitative analysis method based on sparse non-negative least square

Abstract

The invention discloses a mixture Raman spectrum qualitative detection method based on sparse non-negative least squares, and relates to the field of Raman spectrums. The mixture Raman spectrum qualitative detection method comprises the steps that a to-be-measured Raman spectrum is processed into column vectors with the same dimensionality as Raman spectrums of all pure substances in a Raman spectrum standard library, then suspected materials contained in a to-be-measured mixture are preliminarily screened from the Raman spectrum standard library by solving an objective function established bythe to-be-measured Raman spectrum and the Raman spectrum standard library, then residual errors between the to-be-measured mixture and the suspected materials is calculated through the sparse non-negative least squares, then the significant difference between the adjacent residual errors is calculated sequentially through two-tailed T testing, the pure substances contained in the to-be-measured mixture are secondarily screened from the suspected materials based on the significant difference, qualitative detection of the unknown to-be-measured mixture is realized, the accuracy rate is high, and operation is easy, quick and effective.

Description

technical field [0001] The invention relates to the field of Raman spectroscopy, in particular to a method for qualitative analysis of mixture Raman spectroscopy based on sparse non-negative least squares. Background technique [0002] After more than three decades of development, Raman spectroscopy has become a powerful tool for qualitative and quantitative analysis. The Raman spectroscopy analysis of mixtures is becoming more and more popular due to its simple sampling method, which can be obtained directly from powders, liquids or even Obtained in a transparent container. [0003] In mixture analysis, the amount of information hidden in spectral data consists of hundreds of points. Various stoichiometric, statistical, and numerical methods have been developed to process spectral data in order to obtain basic information about the mixture. The traditional method is to determine the composition of the mixture by matching the peak position and peak intensity of the spectru...

AI Technical Summary

Problems solved by technology

However, with the increase of the spectral complexity of the mixture, overlapping peaks may appear, resulting in a greatly reduced spectral similarity between the spectrum of the mixture and its components, and, with the increase of pure substance samples in the standard library, the detection accuracy of this method will be greatly improved. reduce

Images