<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In this paper, we first use particle swarm optimization to improve the BP neural network algorithm to optimize the neural network connection weights and thresholds and then apply it to the flexible manufacturing system to build a complete, flexible manufacturing process of mass customization. Based on the nonlinear mapping relationship between customer demand and product structure, a three-layer network is used to identify the effective demand of customers as well as the actual risk of the enterprise in order to satisfy the common users who do not care about the product structure and the optimization of industrial structure. After completing the configuration, the system is applied to an automobile company and a manufacturing company for testing purposes. After a long time of use, the average accuracy error is 0.0061. The value of the body module ignored by the customer is as high as 0.129, and the weight of the indicator system of the traveling control module is as low as 0.072, but it has become one of the most important indicators of the potential demand of the customer. Supply chain risk has a maximum difference of 6% when compared to the actual values. The manufacturer’s expectation value of 3.8967 for SCRF11 is “low”.</p>
</abstract>

In this paper, we first use particle swarm optimization to improve the BP neural network algorithm to optimize the neural network connection weights and thresholds and then apply it to the flexible manufacturing system to build a complete, flexible manufacturing process of mass customization. Based on the nonlinear mapping relationship between customer demand and product structure, a three-layer network is used to identify the effective demand of customers as well as the actual risk of the enterprise in order to satisfy the common users who do not care about the product structure and the optimization of industrial structure. After completing the configuration, the system is applied to an automobile company and a manufacturing company for testing purposes. After a long time of use, the average accuracy error is 0.0061. The value of the body module ignored by the customer is as high as 0.129, and the weight of the indicator system of the traveling control module is as low as 0.072, but it has become one of the most important indicators of the potential demand of the customer. Supply chain risk has a maximum difference of 6% when compared to the actual values. The manufacturer’s expectation value of 3.8967 for SCRF11 is “low”.

Research on key technology of mass customization based on flexible manufacturing in the context of deep learning

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

In this paper, we first use particle swarm optimization to improve the BP neural network algorithm to optimize the neural network connection weights and...

Research on key technology of mass customization based on flexible manufacturing in the context of deep learning

Online veröffentlicht: 31. Jan. 2024

Seitenbereich: -

Eingereicht: 15. Dez. 2023

Akzeptiert: 23. Dez. 2023

DOI: https://doi.org/10.2478/amns-2024-0115

Schlüsselwörter
Particle swarm optimization, BP neural network, Deep learning, Flexible manufacturing, Product structure, Mass customization

© 2024 Fuxiang Zeng, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Research on key technology of mass customization based on flexible manufacturing in the context of deep learning

Online veröffentlicht: 31. Jan. 2024

Seitenbereich: -

Eingereicht: 15. Dez. 2023

Akzeptiert: 23. Dez. 2023

DOI: https://doi.org/10.2478/amns-2024-0115

SchlüsselwörterParticle swarm optimization, BP neural network, Deep learning, Flexible manufacturing, Product structure, Mass customization

© 2024 Fuxiang Zeng, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Particle swarm optimization, BP neural network, Deep learning, Flexible manufacturing, Product structure, Mass customization