Yunhao Yang杨云浩

Continuous fiber paths and manufacturing constraints are incorporated into topology design. Finite element analysis, physics-inspired graph neural networks, and reinforcement learning connect digital structures to 3D-printed materials.

Zigzag silk fibroin ionic conductors connect geometric parameterization, performance prediction, and reinforcement learning with structural optimization for spatially resolved wearable motion detection.

This collaborative work incorporates network connectivity and physical information into graph neural networks to predict nonuniform deformation and stress–strain responses in disordered fibrous materials.

Tensile simulation and deep reinforcement learning are combined for biomimetic three-dimensional disordered fibrous networks, exploring structural changes that bring lightweight design and mechanical performance together.