Data-driven robust topology optimization using surrogate modeling, model reduction, and machine learning

doi:10.1016/j.engappai.2025.113136

Data-driven robust topology optimization using surrogate modeling, model reduction, and machine learning

2025-11-11

0

PRE

AI

DOI：10.1016/j.engappai.2025.113136

Original

Original request for help

Save

Abstract

En 中文

• A novel data-driven framework for robust topology optimization under load uncertainty. • Polynomial Chaos Expansion used for accurately computing the statistical moments. • Linearity-assumption- enhanced substructuring method for drastic model order reduction. • Physics-enhanced ANN enables real-time prediction of substructure shape functions. • Numerical examples and Monte Carlo verifies the proposed method performs well.

Data-driven robust topology optimization using surrogate modeling, model reduction, and machine learning

Abstract

Journal

Organization

Cited Papers