一种面向人形机器人灵巧手的自主操作训练系统及方法
An autonomous operation training system based on image acquisition and multi-source information fusion has solved the problem of missing joint angle and object pose information in complex environments for humanoid robot dexterous hands, realizing robust intelligent control of dexterous hands and improving their autonomous operation capabilities.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING RES INST OF PRECISE MECHATRONICS CONTROLS
- Filing Date
- 2024-11-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to achieve robust intelligent control of humanoid robot dexterous hands in complex, unstructured environments, especially for fine manipulation when joint angles and object pose information are lacking.
The system employs an image acquisition unit, a pose estimation network unit, a humanoid dexterous hand structure, a microcontroller unit, a feature fusion and extraction network unit, an autonomous strategy network unit, and a PD controller. By fusing multi-source information through a hierarchical attention recursive fusion module, autonomous operation training of the dexterous hand is achieved.
Robust intelligent control of dexterous hands in complex environments has been achieved, solving the problem of missing joint angle and object pose information and improving the autonomous operation capability of dexterous hands.
Smart Images
Figure CN119847010B_ABST