Knowledge-guided generative surrogate modeling for high-dimensional design optimization under scarce data

Researchers developed RBF-Gen, a new AI framework that combines limited experimental data with domain expertise to create more accurate surrogate models for engineering optimization. The method uses radial basis functions and generator networks to address data scarcity challenges in mechanical design and manufacturing processes.

• →RBF-Gen framework combines scarce data with domain knowledge to improve surrogate modeling accuracy in engineering applications.
• →The method uses radial basis functions with more centers than training samples and leverages null space via generator networks.
• →Significantly outperforms standard RBF surrogates on structural optimization problems in data-scarce environments.
• →Successfully demonstrated superior predictive accuracy on real-world semiconductor manufacturing datasets.
• →Addresses the challenge of integrating subject matter expert knowledge with limited experimental data systematically.