Fuse solid-liquid transition state control method and system based on image processing, and storage medium
By using a deep learning network model based on image processing to identify the solid-liquid transition state of the molten wire in real time and dynamically adjust the laser wire feeding parameters, the problem of unstable molten pool in laser wire feeding additive manufacturing is solved, thus achieving stability in the processing process and high quality of the formed parts.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-10
- Publication Date
- 2026-03-13
AI Technical Summary
In existing laser-fed additive manufacturing technologies, it is difficult to accurately identify and control the solid-liquid transition state of the molten wire, leading to process instability and problems such as wire vibration, decreased dimensional accuracy of the formed wire, and process interruption.
An image processing-based method is adopted, which uses an image acquisition device to acquire images of the molten pool and wire in real time. A pre-trained deep learning network model is used to identify the solid-liquid transition state of the molten wire, and the laser power and deposition height are dynamically adjusted according to the identification results to maintain the molten pool in a normal state.
It enables precise identification and control of the solid-liquid transition state of the molten wire, improves the stability of the processing, avoids the defects caused by the lag in state identification in traditional methods, and ensures the continuity and reliability of the forming process.
Smart Images
Figure CN121661441A_ABST