基于多机协同下的机器人自学习智能焊接系统及方法
By using a multi-machine collaborative self-learning system, the welding path is optimized through visual recognition and reinforcement learning, which solves the problems of adaptability and efficiency of traditional welding systems in complex curves and environments, and achieves efficient and precise welding operations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING HEXIN AUTOMATION CO LTD
- Filing Date
- 2025-06-06
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional teach-in programming welding is difficult to accurately describe and control welding tasks with complex spatial curves or free surfaces, and it cannot adapt to workpiece position deviations and welding parameter adjustments in real time, affecting welding continuity and efficiency.
A multi-machine collaborative robot self-learning system is adopted, which uses a large-scale visual line scan camera and a weld seam positioning and tracking device to collaboratively determine the welding scene. Combined with an intelligent control cabinet, dynamic path optimization and collision avoidance decisions are made. Welding data is processed through a reinforcement learning environment to divide task priorities and working ranges. Real-time monitoring and a composite reward function guide the robot to learn optimization strategies.
It enables efficient and precise welding operations, improves welding quality and safety, and solves the problems of adaptability and efficiency of traditional welding systems in complex environments.
Smart Images
Figure CN120395881B_ABST