Skip to main content
Springer Nature Link
Log in

DRL-MeshGen: automated block-structured mesh generation framework via deep reinforcement learning and optimal conformal mapping

  • Original Article
  • Published:
Save article
View saved research
Engineering with Computers Aims and scope Submit manuscript

Abstract

Block-structured mesh generation offers significant advantages in numerical computation and simulation, yet conventional methods often rely on manual intervention, struggling to balance automation with high-quality output. To address this, an automated block-structured mesh generation framework that integrates deep reinforcement learning and optimal conformal mapping techniques is proposed in this paper. This framework utilizes the triangulation of the geometric model as input and operates in the following four steps. Firstly, surface triangular meshes are mapped to planar parametric domains using the Ricci flow algorithm. Secondly, isocontours are extracted based on density variation before and after conformal mapping to guide partitioning. Thirdly, a reinforcement learning decision framework is constructed to formulate mesh generation as a sequential decision-making process, where topological template selection and singularity placement are optimized through reward functions. Finally, surface-structured meshes are generated through mesh smoothing and inverse conformal mapping. Experimental results demonstrate that the proposed method outperforms existing methods in both generation efficiency and mesh quality, providing an intelligent new solution for automated mesh generation in CAD/CAE applications.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+
from €37.37 /Month
  • Starting from 10 chapters or articles per month
  • Access and download chapters and articles from more than 300k books and 2,500 journals
  • Cancel anytime
View plans

Buy Now

Price includes VAT (Netherlands)

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
The alternative text for this image may have been generated using AI.
Fig. 2
The alternative text for this image may have been generated using AI.
Fig. 3
The alternative text for this image may have been generated using AI.
Algorithm 1
The alternative text for this image may have been generated using AI.
Fig. 4
The alternative text for this image may have been generated using AI.
Fig. 5
The alternative text for this image may have been generated using AI.
Fig. 6
The alternative text for this image may have been generated using AI.
Algorithm 2
The alternative text for this image may have been generated using AI.
Fig. 7
The alternative text for this image may have been generated using AI.
Fig. 8
The alternative text for this image may have been generated using AI.
Fig. 9
The alternative text for this image may have been generated using AI.
Fig. 10
The alternative text for this image may have been generated using AI.
Fig. 11
The alternative text for this image may have been generated using AI.
Fig. 12
The alternative text for this image may have been generated using AI.
Fig. 13
The alternative text for this image may have been generated using AI.
Fig. 14
The alternative text for this image may have been generated using AI.

Similar content being viewed by others

Explore related subjects

Discover the latest articles, books and news in related subjects, suggested using machine learning.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Baker T (2005) Mesh generation: Art or science? Prog Aerosp Sci 41:29–63

    Article  Google Scholar 

  2. Hughes TJR, Cottrell JA, Bazilevs Y (2005) Isogeometric analysis: Cad, finite elements, nurbs, exact geometry and mesh refinement. Comput Methods Appl Mech Eng 194(39–41):4135–4195

    Article  MathSciNet  Google Scholar 

  3. Tong Y, Alliez P, Cohen-Steiner D, Desbrun M (2006) Designing quadrangulations with discrete harmonic forms. In: Proceedings of the Fourth Eurographics Symposium on Geometry Processing, Cagliari, Sardinia, Italy, June 26-28, 2006

  4. Lee TY, Yen SW, Yeh IC (2008) Texture mapping with hard constraints using warping scheme. IEEE Trans Visual Comput Graphics 14(2):382–395

    Article  Google Scholar 

  5. Lu F, Qi L, Jiang X, Liu G, Liu Y, Chen B, Pang Y, Hu X (2020) Nnw-gridstar: interactive structured mesh generation software for aircrafts. Adv Eng Softw 145:102803

    Article  Google Scholar 

  6. Chuen A, Hosseini SS, Jensen JC, Chan W (2023) Aerodynamic simulations for complex geometries using automatically generated structured overset meshes. AIAA AVIATION 2023 Forum

  7. Fogg JH (2015) Automatic generation of multiblock decompositions for meshing. IEEE Trans Visual Comput Graphics 101(13):965–991

    MathSciNet  Google Scholar 

  8. Georgiadis C, Reberol M, Remacle JF (2022) Indirect all-quadrilateral meshing based on bipartite topological labeling. Eng Comput 38(5):4731–4747

    Article  Google Scholar 

  9. Zhang S, Xu G, Wu H, Gu R, Qi L, Pang Y (2024) Medial hex-meshing: high-quality all-hexahedral mesh generation based on medial mesh. Eng Comput 40(4):2537–2557

    Article  Google Scholar 

  10. Lei N, Li Z, Xu Z, Li Y, Gu X (2023) What’s the situation with intelligent mesh generation: A survey and perspectives. IEEE Trans Visual Comput Graphics 30(8):4997–5017

    Article  Google Scholar 

  11. Chen X, Liu J, Pang Y, Chen J, Chi L, Gong C (2020) Developing a new mesh quality evaluation method based on convolutional neural network. Eng Appl Comput Fluid Mech 14(1):391–400

    Google Scholar 

  12. Wang X, Yue Q, Liu X (2024) Intelligent mesh generation for crack simulation using graph neural networks. Comput Struct 292:107234

    Article  Google Scholar 

  13. Aygun A, Maulik R, Karakus A (2023) Physics-informed neural networks for mesh deformation with exact boundary enforcement. Eng Appl Artif Intell 125:106660

    Article  Google Scholar 

  14. Chen X, Liu J, Yan J, Wang Z, Gong C (2022) An improved structured mesh generation method based on physics-informed neural networks. arXiv preprint arXiv:2210.09546

  15. Li T, Shi Y, Sun X, Wang J, Yin B (2021) Pcgan: Prediction-compensation generative adversarial network for meshes. IEEE Trans Circuits Syst Video Technol 32(7):4667–4679

    Article  Google Scholar 

  16. Kowalski N, Ledoux F, Frey P (2015) Automatic domain partitioning for quadrilateral meshing with line constraints. Eng Comput 31:405–421

    Article  Google Scholar 

  17. Viertel R, Osting B (2019) An approach to quad meshing based on harmonic cross-valued maps and the ginzburg-landau theory. SIAM J Sci Comput 41(1):452–479

    Article  MathSciNet  Google Scholar 

  18. Qi L, Qiu J, Xu G, Liu Y, Xu J, Gu R, Lu F, Pang Y (2024) Rff-meshing: a parallel and anisotropic quad-dominant mesh generation framework based on riemann frame field. Eng Comput 1–23

  19. Lei N, Zheng X, Luo Z, Luo F, Gu X (2020) Quadrilateral mesh generation ii: Meromorphic quartic differentials and abel-jacobi condition. Comput Methods Appl Mech Eng 366:112980

    Article  MathSciNet  Google Scholar 

  20. Zheng X, Zhu Y, Chen W, Lei N, Luo Z, Gu X (2021) Quadrilateral mesh generation iii: Optimizing singularity configuration based on abel-jacobi theory. Comput Methods Appl Mech Eng 387:114146

    Article  MathSciNet  Google Scholar 

  21. Gu X, Yau S-T (2003) Global conformal surface parameterization. In: Proceedings of the 2003 Eurographics/ACM SIGGRAPH Symposium on Geometry Processing, pp. 127–137

  22. Lévy B, Petitjean S, Ray N, Maillot J (2023) Least squares conformal maps for automatic texture atlas generation. In: Seminal Graphics Papers: Pushing the Boundaries, Volume 2, pp. 193–202

  23. Jin M, Kim J, Luo F, Gu X (2008) Discrete surface ricci flow. IEEE Trans Visual Comput Graph 14(5):1030–1043

    Article  Google Scholar 

  24. Mullen P, Tong Y, Alliez P, Desbrun M (2008) Spectral conformal parameterization. In: Computer Graphics Forum, vol. 27, pp. 1487–1494. Wiley Online Library

  25. Chern A, Pinkall U, Schröder P (2015) Close-to-conformal deformations of volumes. ACM Trans Graph (TOG) 34(4):1–13

    Article  Google Scholar 

  26. Xu S, Wang B, Liu J (2015) On the use of schwarz-christoffel conformal mappings to the grid generation for global ocean models. Geosci Model Dev 8(10):3471–3485

    Article  Google Scholar 

  27. Vinhais C, Kociński M, Materka A (2018) Centerline-radius polygonal-mesh modeling of bifurcated blood vessels in 3d images using conformal mapping. In: 2018 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA), pp. 180–185. IEEE

  28. Chen X, Liu J, Gong C, Pang Y, Chen B (2020) An airfoil mesh quality criterion using deep neural networks. In: 2020 12th International Conference on Advanced Computational Intelligence (ICACI), pp. 536–541. IEEE

  29. Chen X, Liu J, Zhang Q, Liu J, Wang Q, Deng L, Pang Y (2023) Developing a novel structured mesh generation method based on deep neural networks. Phys Fluids 35(9)

  30. Chen X, Li T, Wan Q, He X, Gong C, Pang Y, Liu J (2022) Mgnet: a novel differential mesh generation method based on unsupervised neural networks. Eng Comput 38(5):4409–4421

    Article  Google Scholar 

  31. Zhang H, Li H, Li N (2024) Meshlink: a surface structured mesh generation framework to facilitate automated data linkage. Adv Eng Softw 194:103661

    Article  Google Scholar 

  32. Siddiqui Y, Alliegro A, Artemov A, Tommasi T, Sirigatti D, Rosov V, Dai A, Nießner M (2024) Meshgpt: Generating triangle meshes with decoder-only transformers. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 19615–19625

  33. Yu Y, Wei X, Li A, Liu JG, He J, Zhang YJ (2022) Hexgen and hex2spline: polycube-based hexahedral mesh generation and spline modeling for isogeometric analysis applications in ls-dyna. Geometric challenges in isogeometric analysis. Springer, Cham, pp 333–363

    Chapter  Google Scholar 

  34. Yu Y, Fang Y, Tong H, Zhang YJ (2025) Dl-polycube: deep learning enhanced generalized polycube method for high-quality hexahedral mesh generation and volumetric spline construction. Eng Comput 1–24

  35. Lorsung C, Farimani AB (2022) Meshdqn: a deep reinforcement learning framework for improving meshes in computational fluid dynamics. arXiv preprint arXiv:2212.01428

  36. Pan J, Huang J, Cheng G, Zeng Y (2023) Reinforcement learning for automatic quadrilateral mesh generation: A soft actor-critic approach. Neural Netw 157:288–304

    Article  Google Scholar 

  37. Tong H, Qian K, Halilaj E, Zhang YJ (2023) Srl-assisted afm: generating planar unstructured quadrilateral meshes with supervised and reinforcement learning-assisted advancing front method. J Comput Sci 72:102109

    Article  Google Scholar 

  38. Lim K, Lee S, Lee K, Yee K (2024) Mesh generation for flow analysis by using deep reinforcement learning. In: AIAA SCITECH 2024 Forum, p. 0382

  39. Zhao R, Ye J, Wang Z, Liu G, Chen Y, Wang Y, Zhu J (2025) Deepmesh: Auto-regressive artist-mesh creation with reinforcement learning. arXiv preprint arXiv:2503.15265

  40. Zhang M, Zeng W, Guo R, Luo F, Gu XD (2015) Survey on discrete surface ricci flow. J Comput Sci Technol 30:598–613

    Article  MathSciNet  Google Scholar 

  41. Yin X, Jin M, Luo F, Gu XD (2008) Discrete curvature flows for surfaces and 3-manifolds. LIX fall colloquium on emerging trends in visual computing. Springer, pp 38–74

    Google Scholar 

  42. Chow B, Luo F (2003) Combinatorial ricci flows on surfaces. J Differ Geom 63(1):97–129

    Article  MathSciNet  Google Scholar 

  43. Floater MS., Hormann K (2005) Surface parameterization: a tutorial and survey. Adv Multiresolution Geom Model 157–186

  44. Hasselt HV, Guez A, Silver D (2016) Deep reinforcement learning with double q-learning. In: National Conference on Artificial Intelligence

  45. Qiu B, Pang Y, Liu Y, Zhang X (2024) Reinforcement learning-driven adaptive region segmentation for structured mesh generation, 641–653

  46. Geuzaine C, Remacle J-F (2009) Gmsh: A 3-d finite element mesh generator with built-in pre- and post-processing facilities. Int J Numer Meth Eng 79(11):1309–1331

    Article  MathSciNet  Google Scholar 

  47. Reberol M, Georgiadis C, Remacle J-F (2021) Quasi-structured quadrilateral meshing in gmsh – a robust pipeline for complex cad models. Computing Research Repository abs/2103.04652

  48. Bommes D, Zimmer H, Kobbelt L (2009) Mixed-integer quadrangulation. ACM Trans Graph (TOG) 28(3):1–10

    Article  Google Scholar 

  49. Zhang Y, Bajaj C, Xu G (2009) Surface smoothing and quality improvement of quadrilateral/hexahedral meshes with geometric flow. Commun Numer Methods Eng 25(1):1–18

    Article  MathSciNet  Google Scholar 

  50. Yu Y, Zhang YJ, Takizawa K, Tezduyar TE, Sasaki T (2020) Anatomically realistic lumen motion representation in patient-specific space-time isogeometric flow analysis of coronary arteries with time-dependent medical-image data. Comput Mech 65:395–404

    Article  MathSciNet  Google Scholar 

  51. Lu F, Pang Y, Jiang X, Sun J, Huang Y, Wang Z, Ju J (2018) Automatic generation of structured multiblock boundary layer mesh for aircrafts. Adv Eng Softw 115:297–313

    Article  Google Scholar 

Download references

Acknowledgements

We extend our gratitude to the anonymous reviewers for their meticulous evaluation and for offering insightful and comprehensive feedback. We also gratefully acknowledge the Grid Generation Team at the Computational Aerodynamics Institute of CARDC for their essential support during manuscript preparation.

Funding

This work is partially supported by the National Key R & D Program of China under Grant No. 2023YFB3309100, “Pioneer” and “Leading Goose” R & D Program of Zhejiang Province (No. 2025C01086), the National Natural Science Foundation of China under Grant No. U22A2033.

Author information

Authors and Affiliations

  1. School of Computer Science, Hangzhou Dianzi University, Hangzhou, 310018, Zhejiang, China

    Long Qi, Gang Xu, Jiamin Xu & Renshu Gu

  2. Institute of Computational Aerodynamics, China Aerodynamics Research and Development Center, Mianyang, 621000, Sichuan, China

    Long Qi, Qiang Wu, Rushuang Mu, Yang Liu, Jingying Qiu & Yufei Pang

Authors
  1. Long Qi
    View author publications

    Search author on:PubMed Google Scholar

  2. Qiang Wu
    View author publications

    Search author on:PubMed Google Scholar

  3. Gang Xu
    View author publications

    Search author on:PubMed Google Scholar

  4. Rushuang Mu
    View author publications

    Search author on:PubMed Google Scholar

  5. Yang Liu
    View author publications

    Search author on:PubMed Google Scholar

  6. Jingying Qiu
    View author publications

    Search author on:PubMed Google Scholar

  7. Jiamin Xu
    View author publications

    Search author on:PubMed Google Scholar

  8. Renshu Gu
    View author publications

    Search author on:PubMed Google Scholar

  9. Yufei Pang
    View author publications

    Search author on:PubMed Google Scholar

Contributions

L.Q. and Q.W. wrote the main manuscript text, coding, and prepared all the figures, R.M. developed and validated the code for conformal parameterization, J.Q., Y.L., J.X. and R.G. checked and modified the manuscript grammar, G.X. designed the manuscript framework and revised the overall content. Y.P. provided technical guidance for the manuscript and designed the experiments. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Gang Xu or Yufei Pang.

Ethics declarations

Conflict of interest

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Qi, L., Wu, Q., Xu, G. et al. DRL-MeshGen: automated block-structured mesh generation framework via deep reinforcement learning and optimal conformal mapping. Engineering with Computers 41, 4293–4315 (2025). https://doi.org/10.1007/s00366-025-02199-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • Issue date:

  • DOI: https://doi.org/10.1007/s00366-025-02199-9

Keywords

Profiles

  1. Long Qi View author profile