KFU. Card publication. The possibility of artificial neural network application in prototyping in instrument making industry

KAZAN
FEDERAL UNIVERSITY

- RUS
- 中文
- ES

THE POSSIBILITY OF ARTIFICIAL NEURAL NETWORK APPLICATION IN PROTOTYPING IN INSTRUMENT MAKING INDUSTRY

Form of presentation

Articles in international journals and collections

Year of publication

2023

��

��

Authors, employees of KFU

Koshkina Irina Aleksandrovna, author

Ulengov Ruslan Anatolevich, author

Bibliographic description in the original language

Ovseenko G.A, Kashaev R.S, Koshkina I.A, The possibility of artificial neural network application in prototyping in instrument making industry//Proceedings of the 2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2023. - 2023. - Vol., Is.. - .

Annotation

The article explores the direction of using artificial neural networks to solve problems of classification of defects in the details of the instrument-making industry on the example of cellular panels. An algorithm for constructing and operating principle of a defect classification system based on a multilayer perceptron is described. Studies of the developed system are presented, in the classification of which experimental data obtained during the control of samples of cellular panels by the low-speed impact method were used. The developed neural network made it possible to perform nonlinear separation and classification of objects according to a set of diagnostic features, to identify a complex relationship between the degree of damage to the control object and the values of informative parameters. The disadvantages of the system in training a neural network are shown, which can be attributed to the need to train a multilayer perceptron to the existence of a training sample containing information about possible defects.

Keywords

neural network application, instrument making industry

The name of the journal

Proceedings of the 2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2023

URL

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85154603583&doi=10.1109%2fREEPE57272.2023.10086823&partnerID=40&md5=e031ec579004ed5fea552a81dfd86ae8

Please use this ID to quote from or refer to the card

https://repository.kpfu.ru/eng/?p_id=284466&p_lang=2

Full metadata record

Field DC	Value	Language
dc.contributor.author	Koshkina Irina Aleksandrovna	ru_RU
dc.contributor.author	Ulengov Ruslan Anatolevich	ru_RU
dc.date.accessioned	2023-01-01T00:00:00Z	ru_RU
dc.date.available	2023-01-01T00:00:00Z	ru_RU
dc.date.issued	2023	ru_RU
dc.identifier.citation	Ovseenko G.A, Kashaev R.S, Koshkina I.A, The possibility of artificial neural network application in prototyping in instrument making industry//Proceedings of the 2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2023. - 2023. - Vol., Is.. - .	ru_RU
dc.identifier.uri	https://repository.kpfu.ru/eng/?p_id=284466&p_lang=2	ru_RU
dc.description.abstract	Proceedings of the 2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2023	ru_RU
dc.description.abstract	The article explores the direction of using artificial neural networks to solve problems of classification of defects in the details of the instrument-making industry on the example of cellular panels. An algorithm for constructing and operating principle of a defect classification system based on a multilayer perceptron is described. Studies of the developed system are presented, in the classification of which experimental data obtained during the control of samples of cellular panels by the low-speed impact method were used. The developed neural network made it possible to perform nonlinear separation and classification of objects according to a set of diagnostic features, to identify a complex relationship between the degree of damage to the control object and the values of informative parameters. The disadvantages of the system in training a neural network are shown, which can be attributed to the need to train a multilayer perceptron to the existence of a training sample containing information about possible defects.	ru_RU
dc.language.iso	ru	ru_RU
dc.subject	neural network application	ru_RU
dc.subject	instrument making industry	ru_RU
dc.title	The possibility of artificial neural network application in prototyping in instrument making industry	ru_RU
dc.type	Articles in international journals and collections	ru_RU