Fow-deformation-aware point cloud completion network for 3D metal bent tube

This paper presents FDANet, a flow-deformation-aware pointcloud completion network for 3D metal bent tubes, using plastic deformation knowledge and attention mechanisms to improveshape and profile completion.

Accepted by Computational Visual Media (IF=17.3)      

Bo-LSTM based cross-sectional profile sequence progressive prediction

method for metal tube rotate draw bending

This study introduces a Bo-LSTM-based predictive model for metal tube bending profiles, effectively utilizing bending process data to enhance prediction accuracy. The method emphasizes robust sequential predictions and has been validated through both simulations and experimental data comparisons.

Published at Advanced Engineering Informatics (IF=8.0)     

Digital-Twin virtual model real-time construction via spatio-temporal

cascade reconstruction for full-field plastic deformation monitoring in

metal tube bending manufacturing

This paper proposes a real-time DTVM construction method for full-field plastic deformation monitoring in tube bending, using spatio-temporal cascade reconstruction and TSCR-Net for accurate, real-time synchronization and prediction.

Published at  Robotics and Computer-Integrated Manufacturing (IF=9.1)