• Proposes a novel knowledge push method integrating distillation model-based dynamic graph enhancement and Bayesian network reasoning for complex product assembly process design.
• Achieves dynamic incremental construction of knowledge graphs with low resource consumption using a confidence-based weighted fusion strategy.
• Demonstrates significant improvement in assembly process knowledge utilization and design efficiency through a specific example.
• Aligns with Industry 5.0 human-centric intelligent manufacturing paradigm, enhancing human-machine collaboration in assembly.