Key Points from Professor Jia Chuanbao’s Presentation (Shandong University)
–The Intersection of Advanced Technology and Craftsmanship (1)
Dated 15-Nov, 2024
1. Visual Sensing and Deep Learning:
– Uses visual sensors to capture images of the welding pool.
– Applies deep learning algorithms to identify and predict the welding penetration state.
2. CNN-LSTM Network:
– Combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for welding process prediction.
– Achieves a prediction error of less than 0.5 and an accuracy of 99% for full penetration welding.
3. Rotating Arc Narrow Gap GTAW Technology:
– This technology, which has independent intellectual property rights, is suitable for welding thick plates in high-end industries like nuclear power and marine engineering.
– It can handle gap widths of less than 5mm and solves issues with incomplete fusion on side walls by rotating the arc actively.
4. Control of Heat and Force Distribution:
– The technology can control the distribution of heat and force during welding, improving quality.
– For multi-layer welding, it reduces the amount of filler material needed by 40%, increasing efficiency.
5. Full Position Welding Capability:
– This technology supports welding in any position, enhancing flexibility and usability.
These advancements help improve the quality, efficiency, and automation of welding processes, especially in demanding industrial sectors.