Synchronous fluorescence spectroscopy characteristic wavelength screening method based on particle swarm optimization algorithm

Abstract

The invention discloses a synchronous fluorescence spectroscopy characteristic wavelength screening method based on a particle swarm optimization algorithm. The synchronous fluorescence spectroscopy characteristic wavelength screening method comprises the steps of (1) analyzing a three-dimensional synchronous fluorescence spectrum by a parallel factor analysis method so as to determine an optimal wavelength difference delta(lambda) for detecting antibiotics; (2) performing spectral denoising preprocessing and spectral normalization on the synchronous fluorescence spectrum under the optimal wavelength difference delta(lambda); (3) screening out k synchronous fluorescence spectrum characteristic wavelengths by the particle swarm optimization algorithm; and (4) optimizing a kernel function parameter (c, g) of a support vector regression (SVR) model by the particle swarm optimization algorithm so as to construct an SVR forecasting model based on the particle swarm optimization algorithm. The synchronous fluorescence spectroscopy characteristic wavelength screening method based on the particle swarm optimization algorithm is beneficial to improving the precision and the speed of a model for forecasting antibiotic residues in food such as meat and meat products, and a novel method is provided for detecting the antibiotic residues in the food such as the meat and the meat products.

Description

technical field [0001] The invention relates to a method for screening characteristic wavelengths of synchronous fluorescence spectra, in particular to a synchronous method based on particle swarm optimization algorithm for screening antibiotic residues in meat (such as duck, chicken, goose, pork, etc.) and their products. Fluorescence spectrum characteristic wavelength screening method. Background technique [0002] Due to its advantages of high sensitivity, good selectivity, simplified spectrum, narrow band and reduced influence of scattered light on the spectrum, synchronous fluorescence spectroscopy can provide rich spectral data information, and is a better multi-component mixture analysis The method has been used in the fields of food quality detection and analysis such as meat and its products. The traditional method of fluorescence analysis is to establish the relationship between the fluorescence intensity at the optimal wavelength and the concentration of antibiot...

AI Technical Summary

Problems solved by technology

[0003] At present, there are many methods for spectral characteristic wavelength screening and model parameter optimization. Among them, genetic algorithm is a method that is widely used and achieves better results. However, genetic algorithm has disadvantages such as slow convergence speed and easy to fall into local optimal solution.

Images