Prediction of rupture instability during deep drawing using Gated Recurrent units based on conditional sequence generation

Key Points

• Prediction of rupture instability during deep drawing is enhanced using gated recurrent units, yielding better reliability.
• The model achieves 85% accuracy in predicting rupture events based on input sequences and manufacturing parameters.
• Analysis using gated recurrent units as a machine learning tool improves outcomes in manufacturing processes.
• This approach may enable better preparation for rupture events, although further validation in industrial settings is necessary.