Reinforcement learning for block decomposition of planar CAD models

DiPrete, Benjamin C.; Garimella, Rao; Cardona, Cristina Garcia; Ray, Navamita

doi:10.1007/s00366-023-01940-6

Reinforcement learning for block decomposition of planar CAD models

Journal Article · Wed Feb 14 00:00:00 EST 2024 · Engineering with Computers

DOI:https://doi.org/10.1007/s00366-023-01940-6· OSTI ID:2305429

DiPrete, Benjamin C.; ; Cardona, Cristina Garcia; Ray, Navamita

Abstract

The problem of hexahedral mesh generation of general CAD models has vexed researchers for over 3 decades and analysts often spend more than 50% of the design-analysis cycle time decomposing complex models into simpler blocks meshable by existing techniques. The decomposed blocks are required for generating good quality meshes (tilings of quadrilaterals or hexahedra) suitable for numerical simulations of physical systems governed by conservation laws. We present a novel AI-assisted method for decomposing (segmenting) planar CAD (computer-aided design) models into well shaped rectangular blocks. Even though the simple examples presented here can also be meshed using many conventional methods, we believe this work is proof-of-principle of a AI-based decomposition method that can eventually be generalized to complex 2D and 3D CAD models. Our method uses reinforcement learning to train an agent to perform a series of optimal cuts on the CAD model that result in a good quality block decomposition. We show that the agent quickly learns an effective strategy for picking the location and direction of the cuts and maximizing its rewards. This paper is the first successful demonstration of an agent autonomously learning how to perform this block decomposition task effectively, thereby holding the promise of a viable method to automate this challenging process for more complex cases.

View Journal Article

Research Organization:: Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)

Sponsoring Organization:: USDOE; USDOE Laboratory Directed Research and Development (LDRD) Program

Grant/Contract Number:: 89233218CNA000001

OSTI ID:: 2305429

Alternate ID(s):: OSTI ID: 2472580

Report Number(s):: LA-UR--23-29571; PII: 1940

Journal Information:: Engineering with Computers, Journal Name: Engineering with Computers; ISSN 0177-0667

Publisher:: Springer Science + Business MediaCopyright Statement

Country of Publication:: United States

Language:: English

References (33)

Automatic scheme selection for toolkit hex meshing White, David R.; Tautges, Timothy J. International Journal for Numerical Methods in Engineering, Vol. 49, Issue 1-2 https://doi.org/10.1002/1097-0207(20000910/20)49:1/2<127::AID-NME926>3.0.CO;2-V	journal	January 2000
Construction of curvilinear co‐ordinate systems and applications to mesh generation Gordon, William J.; Hall, Charles A. International Journal for Numerical Methods in Engineering, Vol. 7, Issue 4 https://doi.org/10.1002/nme.1620070405	journal	January 1973
Automatic three-dimensional mesh generation by the finite octree technique Shephard, Mark S.; Georges, Marcel K. International Journal for Numerical Methods in Engineering, Vol. 32, Issue 4 https://doi.org/10.1002/nme.1620320406	journal	September 1991
Hexahedral mesh generation by medial surface subdivision: Part I. Solids with convex edges Price, M. A.; Armstrong, C. G.; Sabin, M. A. International Journal for Numerical Methods in Engineering, Vol. 38, Issue 19 https://doi.org/10.1002/nme.1620381910	journal	October 1995
Neural Networks Based Mesh Generation Method in 2-D Ahmet, Çinar; Ahmet, Arslan EurAsia-ICT 2002: Information and Communication Technology https://doi.org/10.1007/3-540-36087-5_46	book	January 2002
MeshingNet: A New Mesh Generation Method Based on Deep Learning Zhang, Zheyan; Wang, Yongxing; Jimack, Peter K. Computational Science – ICCS 2020 https://doi.org/10.1007/978-3-030-50420-5_14	book	January 2020
Modeling 3D Shapes by Reinforcement Learning Lin, Cheng; Fan, Tingxiang; Wang, Wenping Computer Vision – ECCV 2020 https://doi.org/10.1007/978-3-030-58607-2_32	book	January 2020
Distortion Energy for Deep Learning-Based Volumetric Finite Element Mesh Generation for Aortic Valves Pak, Daniel H.; Liu, Minliang; Kim, Theodore Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 https://doi.org/10.1007/978-3-030-87231-1_47	book	January 2021
CCSweep: automatic decomposition of multi-sweep volumes White, David R.; Saigal, Sunil; Owen, Steven J. Engineering with Computers, Vol. 20, Issue 3 https://doi.org/10.1007/s00366-004-0290-6	journal	July 2004
Recognition of form features using convex decomposition Kim, Y. S. Computer-Aided Design, Vol. 24, Issue 9 https://doi.org/10.1016/0010-4485(92)90027-8	journal	September 1992
A general three-dimensional elliptic grid generation system on a composite block structure Thompson, Joe F. Computer Methods in Applied Mechanics and Engineering, Vol. 64, Issue 1-3 https://doi.org/10.1016/0045-7825(87)90047-8	journal	October 1987
Transfinite mappings and their application to grid generation Gordon, William J.; Thiel, Linda C. Applied Mathematics and Computation, Vol. 10-11 https://doi.org/10.1016/0096-3003(82)90191-6	journal	January 1982
Sheet operation based block decomposition of solid models for hex meshing Wang, Rui; Shen, Chun; Chen, Jinming Computer-Aided Design, Vol. 85 https://doi.org/10.1016/j.cad.2016.07.016	journal	April 2017
Fuzzy clustering based pseudo-swept volume decomposition for hexahedral meshing Wu, Haiyan; Gao, Shuming; Wang, Rui Computer-Aided Design, Vol. 96 https://doi.org/10.1016/j.cad.2017.10.001	journal	March 2018
Recurrent neural networks as optimal mesh refinement strategies Bohn, Jan; Feischl, Michael Computers & Mathematics with Applications, Vol. 97 https://doi.org/10.1016/j.camwa.2021.05.018	journal	September 2021
A mesh optimization method using machine learning technique and variational mesh adaptation Wu, Tingfan; Liu, Xuejun; An, Wei Chinese Journal of Aeronautics, Vol. 35, Issue 3 https://doi.org/10.1016/j.cja.2021.05.018	journal	March 2022
Evaluation of user-guided semi-automatic decomposition tool for hexahedral mesh generation Lu, Jean Hsiang-Chun; Quadros, William Roshan; Shimada, Kenji Journal of Computational Design and Engineering, Vol. 4, Issue 4 https://doi.org/10.1016/j.jcde.2017.05.001	journal	May 2017
An Art Gallery Approach to Submap Meshing Mukherjee, Nilanjan Procedia Engineering, Vol. 82 https://doi.org/10.1016/j.proeng.2014.10.393	journal	January 2014
A self-learning finite element extraction system based on reinforcement learning Pan, Jie; Huang, Jingwei; Wang, Yunli Artificial Intelligence for Engineering Design, Analysis and Manufacturing, Vol. 35, Issue 2 https://doi.org/10.1017/S089006042100007X	journal	April 2021
A Knowledge-Based Mesh Generation System for Forging Simulation Takata, Osamu; Nakanishi, Koukichi; Horinouchi, Nariaki Applied Intelligence, Vol. 11, Issue 2, p. 149-168 https://doi.org/10.1023/A:1008324413477	journal	January 1999
Human-level control through deep reinforcement learning Mnih, Volodymyr; Kavukcuoglu, Koray; Silver, David Nature, Vol. 518, Issue 7540 https://doi.org/10.1038/nature14236	journal	February 2015
On the use of Machine Learning to Defeature CAD Models for Simulation Danglade, Florence; Pernot, Jean-Philippe; Véron, Philippe Computer-Aided Design and Applications, Vol. 11, Issue 3 https://doi.org/10.1080/16864360.2013.863510	journal	December 2013
SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernels Fey, Matthias; Lenssen, Jan Eric; Weichert, Frank 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) https://doi.org/10.1109/CVPR.2018.00097	conference	June 2018
Reinforcement Cutting-Agent Learning for Video Object Segmentation Han, Junwei; Yang, Le; Zhang, Dingwen 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR.2018.00946	conference	June 2018
Dynamic Face Video Segmentation via Reinforcement Learning Wang, Yujiang; Dong, Mingzhi; Shen, Jie 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) https://doi.org/10.1109/CVPR42600.2020.00699	conference	June 2020
DeepCAD: A Deep Generative Network for Computer-Aided Design Models Wu, Rundi; Xiao, Chang; Zheng, Changxi 2021 IEEE/CVF International Conference on Computer Vision (ICCV) https://doi.org/10.1109/ICCV48922.2021.00670	conference	October 2021
A Reinforcement Learning Framework for Medical Image Segmentation Sahba, F.; Tizhoosh, H. R.; Salama, M. M. A. The 2006 IEEE International Joint Conference on Neural Network Proceedings https://doi.org/10.1109/IJCNN.2006.246725	conference	January 2006
Geometric Deep Learning: Going beyond Euclidean data Bronstein, Michael M.; Bruna, Joan; LeCun, Yann IEEE Signal Processing Magazine, Vol. 34, Issue 4 https://doi.org/10.1109/MSP.2017.2693418	journal	July 2017
Learning Direction Fields for Quad Mesh Generation Dielen, Alexander; Lim, Isaak; Lyon, Max Computer Graphics Forum, Vol. 40, Issue 5 https://doi.org/10.1111/cgf.14366	journal	August 2021
Feature Decomposition for Hexahedral Meshing Lu, Yong; Gadh, Rajit; Tautges, Timothy J. Volume 1: 25th Design Automation Conference https://doi.org/10.1115/DETC99/DAC-8618	conference	September 1999
Hex-Mesh Generation and Processing: A Survey Pietroni, Nico; Campen, Marcel; Sheffer, Alla ACM Transactions on Graphics, Vol. 42, Issue 2 https://doi.org/10.1145/3554920	journal	October 2022
A deep learning approach to the classification of 3D CAD models Qin, Fei-wei; Li, Lu-ye; Gao, Shu-ming Journal of Zhejiang University SCIENCE C, Vol. 15, Issue 2 https://doi.org/10.1631/jzus.C1300185	journal	February 2014
Unstructured and Semi-Structured Hexahedral Mesh Generation Methods Sarrate, J.; Ruiz-Gironés, E.; Roca, X. Computational Technology Reviews, Vol. 10 https://doi.org/10.4203/ctr.10.2	journal	September 2014

Similar Records

Volume Decomposition and Feature Recognition for Hexahedral Mesh Generation

Conference · Mon Sep 27 00:00:00 EDT 1999 · OSTI ID:14065

Feature based volume decomposition for automatic hexahedral mesh generation

Journal Article · Sun Feb 20 23:00:00 EST 2000 · ASME Journal of Manufacturing Science and Engineering · OSTI ID:751465

Related Subjects

97 MATHEMATICS AND COMPUTING
Block decomposition
Computer Science
Mathematics
Quadrilateral and Hexahedral Mesh generation
Reinforcement learning

Reinforcement learning for block decomposition of planar CAD models

Citation Formats

References (33)

Similar Records

Related Subjects