Search papers, labs, and topics across Lattice.
This paper introduces Dmsh, a fully automated reinforcement learning framework that integrates geometric decomposition and quadrilateral mesh generation, addressing the longstanding challenges in computational engineering. By employing a parametric Soft Actor-Critic architecture within a Markov Decision Process, Dmsh effectively coordinates three agents for topology simplification, geometric regularization, and mesh generation, allowing for efficient exploration of a hybrid discrete-continuous action space. The framework demonstrates superior performance across various benchmarks, achieving higher automation, robustness, and mesh quality compared to traditional methods, thus setting a new standard in learning-based mesh generation.
Dmsh achieves globally conforming all-quadrilateral meshes without the need for post hoc corrections, revolutionizing the automation of mesh generation in complex geometries.
Generating high-quality meshes for arbitrary geometries remains a fundamental bottleneck in computational engineering, often demanding heuristic tuning and semi-manual workflows. In this paper, we introduce Dmsh, a first fully automated reinforcement learning pipeline that unifies geometric decomposition and quadrilateral mesh generation within a single learning-based framework. Dmsh decomposes the problem through three coordinated agents handling topology simplification, geometric regularization, and mesh generation. The meshing process is formulated as a Markov Decision Process and solved using a parametric Soft Actor-Critic architecture with decoupled critics, enabling efficient exploration of a hybrid discrete-continuous action space. A curriculum learning strategy ensures scalability from simple domains to highly complex geometries, suppressing seed variance. By design, the recursive decomposition enables parallel meshing of subregions, yielding globally conforming all-quadrilateral meshes without post hoc correction. Across a wide range of benchmarks, Dmsh consistently outperforms existing methods in automation, robustness, and mesh quality, establishing a new paradigm for learning-based mesh generation.
