Method for improving measurement accuracy of uneven sample elements based on LIBS

Abstract

The invention relates to a method for improving measurement accuracy of uneven sample elements based on LIBS. The method comprises the following steps: in a step of sample pretreatment and spectrum detection, carrying out spectral data acquisition on a sample in S1 by utilizing a laser-induced breakdown spectroscopy technology; in a step of spectral correction: preprocessing the original spectraldata obtained in the S2 by utilizing an intensity normalization and multivariate scattering correction method, extracting characteristic spectral peaks of researched elements in a spectrum, and dividing the characteristic spectral peaks into two groups according to a training set and a prediction set; in a step of parameter optimization: optimizing the parameters of the support vector machine by using the training set data and adopting grid search; carrying out model establishment and model prediction. According to the method, the purposes of simplicity or no need of sample pretreatment, rapidness, high accuracy and the like when the LIBS technology is used for detecting the uneven sample are achieved.

Description

technical field [0001] The invention relates to a method for improving the measurement accuracy of uneven sample elements based on LIBS. Background technique [0002] LIBS technology is a spectral detection technology that has developed rapidly in recent years. Due to the advantages of real-time, in-situ, no need for complex sample pretreatment, and multi-component simultaneous detection, this technology has been gradually applied to more and more fields. In the actual detection process, the surface morphology of the sample is an important factor affecting the detection accuracy. Especially when LIBS technology is applied to extreme environments such as Martian rocks, deep-sea hydrothermal chimneys, underground minerals, etc., the uneven surface of the detection object will bring serious matrix effects to LIBS detection, and it is difficult or impossible to realize the detection of samples in these environments. Even if it can be processed quickly, the detection object may ...

AI Technical Summary

Problems solved by technology

[0004] The above-mentioned three types of methods for improving LIBS technology to quantitatively detect elements in different samples of unevenness, among them, for some extreme environments, sample pretreatment may not be applicable, and in the case of real-time monitoring, samples generally cannot be subjected to complex pretreatment

In addition, there are many studies on spectral pretreatment and quantitative analysis methods, which have improved the detection accuracy to a certain extent, but the effect on samples with uneven surfaces is still not ideal.

Images