Promising directions of application of neural networks in design activity.

Keywords: convolutional neural network, deep neural network, parameters, project, design activity, mechanical engineering, assembly


The article describes promising areas of application of neural networks in design activities. It is emphasized that in today's conditions neural network technologies have found application in economics, medicine, industry, and many other fields of science and technology, capable of solving almost any problem related to modeling, forecasting, optimization. Emphasis is placed on the research issues, emphasizing that production processes are characterized by a huge variety of dynamically interacting parameters and are usually too difficult to create adequate analytical models, and in some cases successful in terms of adequacy of the described process analytical mathematical models fail due to high computing power requirements. Two models of neural networks are proposed: deep neural network and convolutional neural network, the work of which is aimed at use in design activities aimed at designing a car spar. Describes and schematically proposes a block diagram of the reverse design of spar profiles, as well as formed a multilayer architecture of convolutional neural network used in design activities, consisting of convolutional layer, union layer and fully connected layer and formed deep neural network architecture, which is used in design activities aimed at designing a car spar. It is emphasized that, in contrast to the model of convolutional neural network, these loads are considered as a whole, not divided into static and dynamic, and reverse design using a deep neural network is carried out using standard libraries. It is emphasized that neural network technologies can be useful in creating a set of basic software unit models, endowed with certain properties that correspond to certain real processes or phenomena, to further combine them in more complex design systems. Moreover, the most complex part of such a set of modules is the very environment of interaction of such blocks, which in the future can also be built on the basis of neural networks.


Bennis, Fouad & Chedmail, P & Helary, O. (2002). Representation of Design Activities Using Neural Networks. 10.1007/978-94-015-9966-5_4.

Ye, Andre. (2022). Successful Neural Network Architecture Design. 10.1007/978-1-4842-7413-2_6.

Diao, Jie. (2022). BP Neural Network for Design of Hybrid System. 10.1007/978-3-030-89508-2_103.

Sasaki, Hidenori & Hidaka, Yuki & Igarashi, Hironaka. (2021). Explainable Deep Neural Network for Design of Electric Motors. IEEE Transactions on Magnetics. PP. 1-1. 10.1109/TMAG.2021.3063141.

Adithya, D & C, Dinakaran. (2019). Artificial Neural Network Based Design of Governor Controller. 7. 261-267.

Frankreiter, Florian & Breitenreiter, Anselm & Schrape, Oliver & Krstic, Milos. (2021). Power- and Area-optimized Neural Network IC-Design for Academic Education. 1-6. 10.1109/ICECS53924.2021.9665471.

Sossa, Humberto & Virgilio-G, Carlos. (2022). Spiking neural networks and dendrite morphological neural networks: an introduction. 10.1016/B978-0-12-820125-1.00022-1.

Fadhil, Talal & Ahmed, Taher & Mashhadany, Yousif. (2021). Application of Artificial Neural Networks as Design Tool for Hot Mix Asphalt. International Journal of Pavement Research and Technology. 10.1007/s42947-021-00065-7.

Xiao, Yang & Fan, Wuyu & Du, Yuan & Du, Li & Chang, Mau-Chung Frank. (2021). CTT-based Non-Volatile Deep Neural Network Accelerator Design. 258-259. 10.1109/ISOCC53507.2021.9613930.

Guo, Xiaohan & Xu, Xiaopeng & Li, Yu & Huang, Weiping. (2021). Extendable neural network and flexible extendable neural network in nanophotonics. Optics Communications. 508. 127671. 10.1016/j.optcom.2021.127671.

Wang, Dali & Zilouchian, Ali. (2001). Application of Neural Network in Design of Digital Filters. 10.1201/9781420058147.ch5.

Abstract views: 0
PDF Downloads: 0
How to Cite
Koshel А. (2022). Promising directions of application of neural networks in design activity . COMPUTER-INTEGRATED TECHNOLOGIES: EDUCATION, SCIENCE, PRODUCTION, (46), 57-63.
Computer science and computer engineering