Researchers have introduced SpectONet, a novel physics-guided spectral deep operator network designed to solve Euler-Bernoulli beam vibration problems. This framework enhances the operator-learning capabilities of DeepONet by incorporating physics-informed constraints and a specialized Chebyshev-Gauss-Lobatto sensor placement strategy. Unlike traditional methods that use uniform sensor distribution, SpectONet employs nonuniform spectral locations, concentrating sensors near domain boundaries to better capture structural responses with fewer inputs. Numerical experiments on synthetic and real-world datasets show SpectONet significantly outperforms baseline models like Vanilla DeepONet, PI-DeepONet, PINN, and CNN-UNet, achieving prediction error improvements of at least 64% on synthetic problems and 37% on real-world data. AI
IMPACT This research offers a more accurate and computationally efficient framework for structural vibration analysis, potentially improving the design and monitoring of bridges and other structures.
RANK_REASON The cluster describes a new academic paper detailing a novel method for solving a specific type of engineering problem using AI. [lever_c_demoted from research: ic=1 ai=1.0]
- Arup Kumar Sahoo
- Chebyshev-Gauss-Lobatto
- CNN-UNet
- DeepONet
- Euler-Bernoulli beam
- physics-informed neural networks
- SpectONet
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
