一种焊接机器人的焊前全局规划方法及系统
By combining sparse observation and neural implicit reconstruction techniques with Bayesian deep learning and Markov decision processes, autonomous global planning of welding robots is achieved, solving the problem of reliance on manual intervention in existing technologies and improving production efficiency and flexible automation capabilities.
Patent Information
- Application Number
- CN202610600503.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-30
- Publication Date
- 2026-07-17
- Estimated Expiration
- 2046-04-30
AI Technical Summary
Existing welding robots lack adaptability when dealing with non-standard components and cannot perform global planning autonomously, resulting in low efficiency and reliance on manual intervention, which cannot meet the needs of flexible automated production.
By combining sparsely observed RGB-D images with neural implicit 3D reconstruction, and through Bayesian deep learning and Markov decision process models, autonomous decision-making and global planning are achieved to generate welding paths.
It enables robots to autonomously perceive and plan, and can autonomously decide when to stop observing under sparse observation conditions, generating global welding planning tasks. This solves the problem of relying on human intervention in existing technologies and improves production efficiency and flexible automation capabilities.
Smart Images
Figure CN122100191B_ABST
Abstract
Citation Information
Patent Citations
Welding seam tracking method, computer storage medium and terminal equipment
CN117934550A
Method for updating scene representation model
CN119278116A