Wavelength selection based ultraviolet-visible spectrum multi-metal ion detection method

Abstract

The invention discloses an ultraviolet-visible spectrum polymetallic ion detection method based on wavelength screening, which can quickly and efficiently eliminate redundant and interference wavelength information, reduce wavelength screening time, and improve the accuracy and robustness of the model. The method comprises: S1, using ultraviolet-visible spectrophotometry to obtain the mixed absorbance of a solution containing multiple metal ions in the full band of 400-700nm; S2, sampling the obtained absorbance by using a multiple cycle Monte Carlo method, and using exponential decay The function preliminarily selects the wavelength, and uses the non-informative variable elimination method to select the effective wavelength point in the mixed absorbance; S3, establishes the regression model between the concentration and the mixed absorbance based on the PLS method, and separates and calculates the concentration of each ion.

Description

technical field [0001] The invention relates to the field of ultraviolet-visible spectrum analysis, in particular to a detection method for ultraviolet-visible spectrum polymetallic ions based on wavelength screening. Background technique [0002] Ultraviolet-visible spectroscopic analysis technology is a high-tech rapid analysis. Due to its simple pretreatment and convenient detection, it has been more and more widely used in the field of analysis and determination of metal ions. However, when using this method to measure the concentration of polymetallic ions, the full-band wavelength information is generally used for PLS modeling. Theoretically, the absorbance at each wavelength point is the sum of the absorbances of various substances and satisfies the Lambert-Beer theorem, but due to experimental errors, instrument noise and the interaction between the components, each wavelength point The contribution of the model is different, and even some wavelength point informati...

AI Technical Summary

Problems solved by technology

However, through analysis, it is found that the MC-UVE method is mainly used to eliminate noise wavelength points in the spectrum, and has great limitations in removing non-noise points with low contribution to modeling; the CARS method is a new variable proposed in recent years. Screening theory, which removes invalid variables and minimizes the impact of collinear variables on the model

However, due to the different samples selected by Monte Carlo sampling, the regression coefficient has been changing, and its size cannot fully reflect the importance of variables

Images