The article presents Neural Implicit Topology Optimization (NITO), a deep learning approach for topology optimization in engineering design. NITO offers a resolution-free and domain-agnostic solution, outperforming state-of-the-art models in structural efficiency and time. The novel method, Boundary Point Order-Invariant MLP (BPOM), is introduced to represent boundary conditions in a sparse and domain-agnostic way. NITO can train and generate solutions in various domains, eliminating the need for numerous domain-specific CNNs and their extensive datasets. The combination of versatility, efficiency, and performance highlights NITO’s potential to transform the landscape of engineering design optimization problems.
Publication date: 7 Feb 2024
Project Page: https://arxiv.org/abs/2402.05073v1
Paper: https://arxiv.org/pdf/2402.05073
