Machine learning-based optimization of deposition line state in concrete 3D printing

doi:10.3785/j.issn.1006-754X.2025.04.183

Chinese Journal of Engineering Design

2025, Vol. 32

Issue (4): 488-498 DOI: 10.3785/j.issn.1006-754X.2025.04.183

Optimization Design

Machine learning-based optimization of deposition line state in concrete 3D printing

Zongfang MA¹(

),Yiyang DAI¹,Jing HE¹(

),Lin SONG¹,Chao LIU²

^1.College of Information and Control Engineering, Xi'an University of Architecture and Technology, Xi'an 710055, China
^2.College of Civil Engineering, Xi'an University of Architecture and Technology, Xi'an 710055, China

Download:

HTML

PDF(3774KB)
Export: BibTeX | EndNote (RIS)

Abstract

Concrete 3D printing, as a rapid and precise additive manufacturing technology, is gradually demonstrating unique advantages in the construction industry. However, concrete 3D printing involves a variety of process parameters (such as printing speed, extrusion speed, layer height and line spacing), and there are complex coupling relationships among these parameters, making it difficult to precisely control the quality of the deposition line state. To address this problem, a machine learning model (BWO-SVM-AdaBoost) based on the beluga whale optimization (BWO) algorithm, support vector machine (SVM) and ensemble learning algorithm AdaBoost was proposed to investigate the relationship between the process parameters of concrete 3D printing and the deposition line state. The BWO-SVM-AdaBoost model was trained and tested based on a dataset consisting of 400 samples, and its prediction accuracy on the training set and the test set reached 99.76% and 95.19%, respectively. Furthermore, the Bayesian optimization algorithm was applied to refine suboptimal printing process parameters and the optimal combination of parameters was generated, thereby effectively enhancing the precision and stability of printed components. The research results provide a novel and effective approach for the systematic optimization of concrete 3D printing process parameters, which can lay a theoretical foundation for achieving high-quality printed structures.

Key words： concrete 3D printing deposition line machine learning Bayesian optimization

Received: 06 December 2024 Published: 01 September 2025

CLC:

TP3-05

Corresponding Authors: Jing HE E-mail: mazf@xauat.edu.cn;hejing0811@xauat.edu.cn

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Zongfang MA
	Yiyang DAI
	Jing HE
	Lin SONG
	Chao LIU

Cite this article:

Zongfang MA,Yiyang DAI,Jing HE,Lin SONG,Chao LIU. Machine learning-based optimization of deposition line state in concrete 3D printing. Chinese Journal of Engineering Design, 2025, 32(4): 488-498.

URL:

https://www.zjujournals.com/gcsjxb/10.3785/j.issn.1006-754X.2025.04.183 OR https://www.zjujournals.com/gcsjxb/Y2025/V32/I4/488

基于机器学习的混凝土3D打印沉积线状态优化

混凝土3D打印作为一种快速、精确的增材制造技术，逐渐在建筑行业中展现出独特优势。然而，混凝土3D打印涉及多种工艺参数（如打印速度、挤出速度、层高和线间距离等），且这些参数之间存在复杂的耦合关系，导致沉积线状态的质量难以精确控制。为解决这一问题，提出了一种基于白鲸优化（beluga whale optimization, BWO）算法、支持向量机（support vector machine, SVM）和集成学习算法AdaBoost的机器学习模型（BWO-SVM-AdaBoost），用于探究混凝土3D打印工艺参数与沉积线状态之间的关系。BWO-SVM-AdaBoost模型以包含400个样本的数据集为基础进行训练与测试，其在训练集和测试集上的预测精度分别达到了99.76%和95.19%。与此同时，利用贝叶斯优化算法对不佳的打印工艺参数进行进一步优化，生成了最优的参数组合，有效地提升了打印构件的精度和稳定性。研究结果为混凝土3D打印工艺参数的系统优化提供了一种新颖且有效的方法，可为实现高质量的打印结构奠定理论基础。

关键词： 混凝土3D打印, 沉积线, 机器学习, 贝叶斯优化

Fig.1 Extrusion process of concrete 3D printing

Fig.2 Schematic of layering and forming process of concrete 3D printing

Fig.3 Overall framework of deposition line state optimization for concrete 3D printing

Table 1 Values of concrete 3D printing process parameters

Fig.4 Schematic of deposition line state observation of dual-line component and experimental platform

Fig.5 Schematic of deposition line state classification

Table 2 Experimental data of concrete 3D printing (partial)

Fig.6 Construction flow of BWO-SVM-AdaBoost model

Fig.7 Hyperplane diagram of SVM classification

Fig.8 Schematic of classifier training process based on AdaBoost

Fig.9 Optimization process of concrete 3D printing process parameters based on Bayesian optimization

Table 3 Number of each deposition line state category of 3D printing

Table 4 Comparison of classification prediction performance of different machine learning models

Fig.10 Classification prediction results of BWO-SVM- AdaBoost model

Fig.11 Optimization results of deposition line state based on Bayesian optimization


[[1]]	刘占省，刘诗楠，赵玉红，等. 智能建造技术发展现状与未来趋势［J］. 建筑技术， 2019， 50（7）： 772-779. LIU Z S， LIU S N， ZHAO Y H， et al. Development status and future trends of intelligent construction technology［J］. Architecture Technology， 2019， 50（7）： 772-779.

[[2]]	CARNEAU P， MESNIL R， BAVEREL O， et al. Layer pressing in concrete extrusion-based 3D-printing： experiments and analysis［J］. Cement and Concrete Research， 2022， 155： 106741.

[[3]]	MARCHON D， KAWASHIMA S， BESSAIES-BEY H， et al. Hydration and rheology control of concrete for digital fabrication： potential admixtures and cement chemistry［J］. Cement and Concrete Research， 2018， 112： 96-110.

[[4]]	YANG L M， SEPASGOZAR S M E， SHIROWZHAN S， et al. Nozzle criteria for enhancing extrudability， buildability and interlayer bonding in 3D printing concrete［J］. Automation in Construction， 2023， 146： 104671.

[[5]]	ZHANG N， SANJAYAN J. Extrusion nozzle design and print parameter selections for 3D concrete printing［J］. Cement and Concrete Composites， 2023， 137： 104939.

[[6]]	LOWKE D， DINI E， PERROT A， et al. Particle-bed 3D printing in concrete construction： possibilities and challenges［J］. Cement and Concrete Research， 2018， 112： 50-65.

[[7]]	ROUSSEL N. Rheological requirements for printable concretes［J］. Cement and Concrete Research， 2018， 112： 76-85.

[[8]]	WOLFS R J M， BOS F P， SALET T A M. Early age mechanical behaviour of 3D printed concrete： numerical modelling and experimental testing［J］. Cement and Concrete Research， 2018， 106： 103-116.

[[9]]	SANJAYAN J G， NEMATOLLAHI B， XIA M， et al. Effect of surface moisture on inter-layer strength of 3D printed concrete［J］. Construction and Building Materials， 2018， 172： 468-475.

[[10]]	HUANG X， YANG W H， SONG F N， et al. Study on the mechanical properties of 3D printing concrete layers and the mechanism of influence of printing parameters［J］. Construction and Building Materials， 2022， 335： 127496.

[[11]]	ZHANG H H， TAN Y K， HAO L C， et al. Intelligent real-time quality control for 3D-printed concrete with near-nozzle secondary mixing［J］. Automation in Construction， 2024， 160： 105325.

[[12]]	TAY Y W D， LI M Y， TAN M J. Effect of printing parameters in 3D concrete printing： printing region and support structures［J］. Journal of Materials Processing Technology， 2019， 271： 261-270.

[[13]]	COMMINAL R， SILVA W R L DA， ANDERSEN T J， et al. Influence of processing parameters on the layer geometry in 3D concrete printing： experiments and modelling［C］//RILEM International Conference on Concrete and Digital Fabrication. Cham： Springer， 2020： 852-862.

[[14]]	LIU Q F， IQBAL M F， YANG J， et al. Prediction of chloride diffusivity in concrete using artificial neural network： modelling and performance evaluation［J］. Construction and Building Materials， 2021， 268： 121082.

[[15]]	KESHAVARZ Z， TORKIAN H. Application of ANN and ANFIS models in determining compressive strength of concrete［J］. Soft Computing in Civil Engineering， 2018， 2（1）： 62-70.

[[16]]	AHMAD A， CHAIYASARN K， FAROOQ F， et al. Compressive strength prediction via gene expression programming （GEP） and artificial neural network （ANN） for concrete containing RCA［J］. Buildings， 2021， 11（8）： 324.

[[17]]	NAZAR S， YANG J， AHMAD A， et al. Comparative study of evolutionary artificial intelligence approaches to predict the rheological properties of fresh concrete［J］. Materials Today Communications， 2022， 32： 103964.

[[18]]	AZIM I， YANG J， IQBAL M F， et al. Semi-analytical model for compressive arch action capacity of RC frame structures［J］. Structures， 2020， 27： 1231-1245.

[[19]]	ALYAMI M， KHAN M， FAWAD M， et al. Predictive modeling for compressive strength of 3D printed fiber-reinforced concrete using machine learning algorithms［J］. Case Studies in Construction Materials， 2024， 20： e02728.

[[20]]	UDDIN M N， YE J H， DENG B Y， et al. Interpretable machine learning for predicting the strength of 3D printed fiber-reinforced concrete （3DP-FRC）［J］. Journal of Building Engineering， 2023， 72： 106648.

[[21]]	BAYAR G， BILIR T. A novel study for the estimation of crack propagation in concrete using machine learning algorithms［J］. Construction and Building Materials， 2019， 215： 670-685.

[[22]]	MÜTEVELLI ÖZKAN İ G， ALDEMIR A. Machine-learning networks to predict the ultimate axial load and displacement capacity of 3D printed concrete walls with different section geometries［J］. Structures， 2024， 66： 106879.

[[23]]	CHANG Z， WAN Z， XU Y D， et al. Convolutional neural network for predicting crack pattern and stress-crack width curve of air-void structure in 3D printed concrete［J］. Engineering Fracture Mechanics， 2022， 271： 108624.

[[24]]	彭云川，杜彦斌，毛恺奕，等. 基于GS-SABO-BPNN和MOPSOCD的激光熔覆工艺参数优化方法［J/OL］. 计算机集成制造系统， 2024： 1-23（2024-09-26）［2024-12-28］. . PENG Y C， DU Y B， MAO K Y， et al. Optimization method of laser cladding process parameters based on GS-SABO-BPNN and MOPSOCD［J/OL］. Computer Integrated Manufacturing Systems， 2024： 1-23（2024-09-26）［2024-12-28］. .

[[25]]	董海，闫思语. 基于GEO-ANN的SLS成型件工艺参数优化与表面质量控制［J/OL］. 计算机集成制造系统， 2024： 1-18（2024-12-02）［2024-12-28］. . DONG H， YAN S Y. Optimization of process parameters and surface quality control of SLS molding parts based on GEO-ANN［J/OL］. Computer Integrated Manufacturing Systems， 2024： 1-18（2024-12-02）［2024-12-28］. .

[[26]]	董文帅，胡彬彬，闫伸，等. 箱式烘烤烟叶高温蒸汽回潮工艺参数优化［J］. 西南农业学报， 2025， 38（1）： 200-208. DONG W S， HU B B， YAN S， et al. Optimization of high temperature steam moisture recovery process parameters for box curing tobacco leaves［J］. Southwest China Journal of Agricultural Sciences， 2025， 38（1）： 200-208.

[[27]]	JIANG J C， YU C L， XU X， et al. Achieving better connections between deposited lines in additive manufacturing via machine learning［J］. Mathematical Biosciences and Engineering， 2020， 17（4）： 3382-3394.

[[28]]	KHAN M S， SANCHEZ F， ZHOU H Y. 3-D printing of concrete： beyond horizons［J］. Cement and Concrete Research， 2020， 133： 106070.

[[29]]	KHOSHNEVIS B. Automated construction by contour crafting： related robotics and information technologies［J］. Automation in Construction， 2004， 13（1）： 5-19.

[[30]]	LABONNOTE N， RØNNQUIST A， MANUM B， et al. Additive construction： state-of-the-art， challenges and opportunities［J］. Automation in Construction， 2016， 72： 347-366.

[[31]]	CHUA C K， LEONG K F. 3D printing and additive manufacturing： principles and applications［M］. Hackensack， New Jersey： World Scientific， 2015.

[[32]]	TAY Y W D， PANDA B， PAUL S C， et al. 3D printing trends in building and construction industry： a review［J］. Virtual and Physical Prototyping， 2017， 12（3）： 261-276.

[[33]]	BUSWELL R A， LEAL DE SILVA W R， JONES S Z， et al. 3D printing using concrete extrusion： a roadmap for research［J］. Cement and Concrete Research， 2018， 112： 37-49.

[[34]]	CHANDRA M A， BEDI S S. Survey on SVM and their application in image classification［J］. International Journal of Information Technology， 2021， 13（5）： 1-11.

[[35]]	HAN B G， ZHANG L Q， OU J P. Smart and multifunctional concrete toward sustainable infrastructures［M］. Singapore： Springer， 2017.

[[36]]	SALAZAR D A， VÉLEZ J I， SALAZAR J C. Comparison between SVM and logistic regression： which one is better to discriminate？［J］. Revista Colombiana de Estadística， 2012， 35（2）： 223-237.

[[37]]	HILL H， DIETRICH S， YEATER D， et al. Developing a catalog of socio-sexual behaviors of beluga whales （Delphinapterus leucas） in the care of humans［J］. Animal Behavior and Cognition， 2015， 2（2）： 105-123.

[[38]]	ZHONG C T， LI G， MENG Z. Beluga whale optimization： a novel nature-inspired metaheuristic algorithm［J］. Knowledge-Based Systems， 2022， 251： 109215.

[[39]]	CAO Y， MIAO Q G， LIU J C， et al. Advance and prospects of AdaBoost algorithm［J］. Acta Automatica Sinica， 2013， 39（6）： 745-758.

[[40]]	WANG X L， JIN Y C， SCHMITT S， et al. Recent advances in Bayesian optimization［J］. ACM Computing Surveys， 2023， 55（13s）： 1-36.

[[41]]	SNOEK J， LAROCHELLE H， ADAMS R P. Practical Bayesian optimization of machine learning algorithms［C］//Neural Information Processing Systems. Lake Tahoe Nevada， Dec. 3-6， 2012.

[[42]]	BERKENKAMP F， KRAUSE A， SCHOELLIG A P. Bayesian optimization with safety constraints： safe and automatic parameter tuning in robotics［J］. Machine Learning， 2023， 112（10）： 3713-3747.

[1]	Bailin LI,Yunfan MA,Yurui CHEN,Yuanlin LUO,Fanwu CHU,Wenlong FU. Research on transformer hotspot temperature prediction and warning technology based on digital twin[J]. Chinese Journal of Engineering Design, 2025, 32(3): 281-295.

[2]	Heng CUI,Zongfang MA,Lin SONG,Chao LIU,Yixuan HAN. Path planning algorithm for concrete 3D printing based on continuous vertex partitioning[J]. Chinese Journal of Engineering Design, 2024, 31(3): 271-279.

[3]	Meng LI,Zong-jun YIN. Research on comprehensive quality assessment method for parts[J]. Chinese Journal of Engineering Design, 2022, 29(4): 410-418.

Viewed

Full text

Abstract

Cited

Shared

Discussed