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

<p>In this paper, for the problems of ceramic sculpture 3D printing process, such as the accuracy is difficult to predict and the process control parameters are difficult to choose, we propose to establish an accuracy prediction model using BP neural network optimized by particle swarm algorithm, and put forward the optimization algorithm of the printing process parameters based on genetic algorithm, and construct the fitness function by using the method of the weight coefficients, to achieve the search for the molding angle under the optimal surface quality and to improve the printing efficiency. In the simulation experiments of printing accuracy prediction, the average errors of dimensional error and surface roughness prediction of the PSO-BP network model proposed in this paper are 5.59% and 5.02%, respectively, and the prediction is accurate and reliable. In the 3D printing parameter optimization experiments, the average deviation of the Y and Z dimensions of the physical size printed with the optimized parameters is 0.0496mm and 0.058mm, respectively, which is smaller than the deviation before optimization. Parameter optimization has also reduced surface roughness in the molded parts of the porcelain sculpture. The extrusion head and hot bed temperature were adjusted to reach the temperature target after maintaining hollow finger precision within 3 degrees.</p>
</abstract>

In this paper, for the problems of ceramic sculpture 3D printing process, such as the accuracy is difficult to predict and the process control parameters are difficult to choose, we propose to establish an accuracy prediction model using BP neural network optimized by particle swarm algorithm, and put forward the optimization algorithm of the printing process parameters based on genetic algorithm, and construct the fitness function by using the method of the weight coefficients, to achieve the search for the molding angle under the optimal surface quality and to improve the printing efficiency. In the simulation experiments of printing accuracy prediction, the average errors of dimensional error and surface roughness prediction of the PSO-BP network model proposed in this paper are 5.59% and 5.02%, respectively, and the prediction is accurate and reliable. In the 3D printing parameter optimization experiments, the average deviation of the Y and Z dimensions of the physical size printed with the optimized parameters is 0.0496mm and 0.058mm, respectively, which is smaller than the deviation before optimization. Parameter optimization has also reduced surface roughness in the molded parts of the porcelain sculpture. The extrusion head and hot bed temperature were adjusted to reach the temperature target after maintaining hollow finger precision within 3 degrees.

Parameter optimization of 3D printing process for ceramic sculptures

Applied Mathematics and Nonlinear Sciences

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

{"article-title":"Parameter optimization of 3D printing process for ceramic sculptures"}

In this paper, for the problems of ceramic sculpture 3D printing process, such as the accuracy is difficult to predict and the process control parameters are...

Parameter optimization of 3D printing process for ceramic sculptures

Data publikacji: 03 cze 2024

Zakres stron: -

Otrzymano: 23 lut 2024

Przyjęty: 03 maj 2024

DOI: https://doi.org/10.2478/amns-2024-1322

Słowa kluczowe
3D printing, Particle swarm algorithm, BP neural network, Genetic algorithm

© 2024 Changzong Shao, published by Sciendo

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

Parameter optimization of 3D printing process for ceramic sculptures

Data publikacji: 03 cze 2024

Zakres stron: -

Otrzymano: 23 lut 2024

Przyjęty: 03 maj 2024

DOI: https://doi.org/10.2478/amns-2024-1322

Słowa kluczowe3D printing, Particle swarm algorithm, BP neural network, Genetic algorithm

© 2024 Changzong Shao, published by Sciendo

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

Słowa kluczowe
3D printing, Particle swarm algorithm, BP neural network, Genetic algorithm