Laser-induced breakdown spectroscopy spectra analysis with deep learning
Lehtonen, Juuso Mikael (2022)
Lehtonen, Juuso Mikael
2022
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2022100420804
https://urn.fi/URN:NBN:fi:amk-2022100420804
Tiivistelmä
As the prices of equipment suitable for laser-induced breakdown spectroscopy (LIBS) have come down, more research and applications are being implemented with LIBS. One of the improvements that could be made in LIBS is the removal of expertise needed for calibration for analysis of different elements.
As LIBS doesn’t need extensive sample preparation and the results are available almost instantly, it could be useful for different kinds of purposes like analyzing the contents of batteries or online quality control of a production line.
Using a neural network to detect the contents of samples would make the analysis more available and easier to use, once the models are trained.
The aim of this work is to find out if neural networks are suitable for analyzing the LIBS spectra and what kind of network would work well in this task.
As LIBS doesn’t need extensive sample preparation and the results are available almost instantly, it could be useful for different kinds of purposes like analyzing the contents of batteries or online quality control of a production line.
Using a neural network to detect the contents of samples would make the analysis more available and easier to use, once the models are trained.
The aim of this work is to find out if neural networks are suitable for analyzing the LIBS spectra and what kind of network would work well in this task.