DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Reinforcement learning for block decomposition of planar CAD models

Journal Article · · Engineering with Computers

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.

Sponsoring Organization:
USDOE
OSTI ID:
2305429
Alternate ID(s):
OSTI ID: 2472580
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)

Sheet operation based block decomposition of solid models for hex meshing journal April 2017
Modeling 3D Shapes by Reinforcement Learning book January 2020
Evaluation of user-guided semi-automatic decomposition tool for hexahedral mesh generation journal May 2017
Hex-Mesh Generation and Processing: A Survey journal October 2022
Human-level control through deep reinforcement learning journal February 2015
A Knowledge-Based Mesh Generation System for Forging Simulation journal January 1999
Unstructured and Semi-Structured Hexahedral Mesh Generation Methods journal September 2014
DeepCAD: A Deep Generative Network for Computer-Aided Design Models conference October 2021
Recognition of form features using convex decomposition journal September 1992
Hexahedral mesh generation by medial surface subdivision: Part I. Solids with convex edges journal October 1995
Geometric Deep Learning: Going beyond Euclidean data journal July 2017
An Art Gallery Approach to Submap Meshing journal January 2014
A Reinforcement Learning Framework for Medical Image Segmentation conference January 2006
Fuzzy clustering based pseudo-swept volume decomposition for hexahedral meshing journal March 2018
A mesh optimization method using machine learning technique and variational mesh adaptation journal March 2022
SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernels conference June 2018
A self-learning finite element extraction system based on reinforcement learning journal April 2021
Transfinite mappings and their application to grid generation journal January 1982
Recurrent neural networks as optimal mesh refinement strategies journal September 2021
On the use of Machine Learning to Defeature CAD Models for Simulation journal December 2013
A deep learning approach to the classification of 3D CAD models journal February 2014
Automatic three-dimensional mesh generation by the finite octree technique journal September 1991
Reinforcement Cutting-Agent Learning for Video Object Segmentation conference June 2018
Neural Networks Based Mesh Generation Method in 2-D book January 2002
A general three-dimensional elliptic grid generation system on a composite block structure journal October 1987
Automatic scheme selection for toolkit hex meshing journal January 2000
Feature Decomposition for Hexahedral Meshing conference September 1999
Distortion Energy for Deep Learning-Based Volumetric Finite Element Mesh Generation for Aortic Valves book January 2021
Learning Direction Fields for Quad Mesh Generation journal August 2021
Construction of curvilinear co‐ordinate systems and applications to mesh generation journal January 1973
CCSweep: automatic decomposition of multi-sweep volumes journal July 2004
MeshingNet: A New Mesh Generation Method Based on Deep Learning book January 2020
Dynamic Face Video Segmentation via Reinforcement Learning conference June 2020