Application of Convolutional Neural Networks to Determine Induction Soldering Process Technological Stages
Abstract
Presented study focuses on solving the problem of controlling the technological process of induction soldering of spacecraft waveguide paths in terms of determining the stages of such a technological process based on the analysis of the video image from the soldering zone. The need to solve this problem lies in the features of the used sensors for optical control of the heating temperature of products. The accuracy of the pyrometer readings significantly affects both the quality of preparation of the surfaces of waveguide elements and the features of the soldering process itself. When the appearance of evaporation during the melting of the flux can significantly distort the temperature readings in soldering zone. In this situation, the precisely-set value of the heating process stabilization temperature, at which the solder melts and the joint is formed, may not correspond to the real temperature and cause the appearance of defects in the finished product, associated with insufficient flow of the solder or the appearance of burns. As a means to implement the technology of machine vision, the use of convolutional neural networks is proposed. This study considers the use of one of the popular architectures of such networks - DenseNet. To select the effective values of hyperparameters, a grid search with cross-validation was used. As a result, the model was obtained that allows to determine the stage of solder melting with an accuracy of over 0.99 on test and verification samples.
Copyright information
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License
About this article
Publication Date
27 February 2023
Article Doi
eBook ISBN
978-1-80296-960-3
Publisher
European Publisher
Volume
1
Print ISBN (optional)
-
Edition Number
1st Edition
Pages
1-403
Subjects
Hybrid methods, modeling and optimization, complex systems, mathematical models, data mining, computational intelligence
Cite this article as:
Tynchenko, V., Kurashkin, S., & Kukartsev, V. (2023). Application of Convolutional Neural Networks to Determine Induction Soldering Process Technological Stages. In P. Stanimorovic, A. A. Stupina, E. Semenkin, & I. V. Kovalev (Eds.), Hybrid Methods of Modeling and Optimization in Complex Systems, vol 1. European Proceedings of Computers and Technology (pp. 210-221). European Publisher. https://doi.org/10.15405/epct.23021.26