Genetic Neural NetworkQuantitative analysis method for laser induced breakdown spectroscopy based on gGenetic Neural Networkgenetic neural network

Abstract

The invention belongs to a quantitative analysis method for a laser induced breakdown spectroscopy and particularly relates to the quantitative analysis method for the laser induced breakdown spectroscopy based on a gGenetic Neural Networkgenetic neural network. In an element quantitative detection technology based on the laser induced breakdown spectroscopy and the gGenetic Neural Networkgeneticneural network, laser-induced breakdown spectroscopy (LIBS) data are analyzed by combining an artificial neural network (ANN) with the genetic algorithm (GA for short) so as to realize the quantitative detection of elements. Compared with the traditional LIBS internal calibration method and the traditional BP(Back Propagation)-ANN method, the method has the advantages of little influence from the matrix effect, simple and quick process, avoidance of local convergence, high reliability of quantitative detection and the like.

Description

technical field [0001] The invention belongs to a quantitative analysis method of laser-induced breakdown spectrum, in particular to a quantitative analysis method of laser-induced breakdown spectrum based on a genetic neural network. Background technique [0002] Laser Induced Breakdown Spectroscopy (LIBS for short), as a kind of atomic emission spectroscopy, is a spectral detection technology that has gradually emerged in recent years. LIBS uses a high-intensity pulsed laser to excite the sample to generate plasma, obtains and analyzes atomic and ion emission lines through a spectrometer, and identifies the existence of substances and conducts quantitative analysis based on the position and signal intensity of the analysis lines. LIBS technology has the characteristics of no need for complicated sample pretreatment, fast operation, and the ability to analyze multiple elements at the same time. It is very suitable for real-time detection of elemental components in gases, li...

AI Technical Summary

Problems solved by technology

The BP algorithm uses the gradient descent method to find the optimal solution of the problem in the weight and threshold space, so the convergence speed is slow and it is easy to fall into a local minimum

In addition, different initial weight settings may cause large differences in convergence and generalization errors. However, the selection of initial connection weights and thresholds of neural networks lacks theoretical basis and has great randomness. global starting point

Therefore, the possibility of obtaining the global optimum with BP neural network is small, limiting its role

Images