基于力控协调和深度强化学习的轴孔装配控制方法及系统
By dividing the robot assembly process into hole-finding and hole-insertion stages, and combining PD force controllers and deep reinforcement learning algorithms, the problems of insufficient assembly accuracy and adaptability in high-precision shaft and hole assembly are solved, and efficient and accurate shaft and hole assembly is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV
- Filing Date
- 2024-01-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing robot assembly technology suffers from problems such as low assembly accuracy, poor adaptability, and low level of intelligence in high-precision shaft and hole assembly. Furthermore, the uncertainty of the force and torque on the shaft during the assembly process can lead to assembly failure or damage to parts.
By employing a force-controlled coordination and deep reinforcement learning approach, the robot assembly process is divided into two stages: hole searching and hole insertion. By combining a PD force controller and a deep reinforcement learning algorithm, the robot's motion trajectory is adjusted through a priority experience playback mechanism and feedback from the PD force controller, thereby optimizing the assembly process.
It improves the success rate and accuracy of robot assembly in high-precision shaft and hole assembly, reduces the force/torque on shaft parts, and realizes autonomous learning and rapid assembly.
Smart Images
Figure CN118143928B_ABST