一种基于手-眼-力一体的接触网腕臂机器人装配方法

By employing a hand-eye-force integrated contact network carousel robot assembly method, and utilizing adaptive multi-exposure fusion technology and impedance control mode, efficient and precise carousel assembly is achieved. This overcomes the shortcomings of traditional visual recognition and manual operation, and improves the assembly success rate and system robustness.

CN121607896BActive Publication Date: 2026-07-17CHINA RAILWAY TENTH BUREAU GRP ELECTRIC ENG CO LTD +2

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA RAILWAY TENTH BUREAU GRP ELECTRIC ENG CO LTD
Filing Date
2025-11-27
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

The assembly of the wrist arm relies on manual operation, which is inefficient, labor-intensive, and difficult to guarantee consistent quality. Furthermore, traditional visual recognition algorithms have difficulty extracting features in strong reflective environments, resulting in large fluctuations in positioning accuracy. The robot also lacks compliance and adaptability during the assembly process.

Method used

A hand-eye-force integrated contact network wrist robot assembly method is adopted. Through system calibration and parameter initialization, visual coarse positioning is performed by combining adaptive multi-exposure fusion technology and deep learning framework, fine positioning is performed by combining depth information from 3D camera, and compliant grasping and assembly are achieved through impedance control mode and finite state machine.

Benefits of technology

It achieves efficient and precise wrist arm assembly, improves the assembly success rate and the engineering practicality of the system, solves the problems of reflection interference and uncertainty in visual recognition and assembly process, and the robot has the ability of "sharp eyes" and "skillful hands", increasing the assembly success rate from 62% to over 90%.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121607896B_ABST
    Figure CN121607896B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于手‑眼‑力一体的接触网腕臂机器人装配方法,包括基于自适应多曝光融合技术,将不同曝光的图像序列合成高质量图像,基于深度学习框架识别出图像中的工件轮廓,计算被识别零件在相机坐标系下的位姿;二维相机与三维相机配合进行精确扫描,通过最近点点云配准算法得到实际工件相对于理论模型的精确位姿偏差矩阵,计算修正后的运动目标位姿;机器人向运动目标位姿运动,当机器人与装配工件即将发生接触时,切换为阻抗控制模式进行腕臂安装,全称持续监控力矩数据,直至安装完成。本发明使机器人能智能应对腕臂装配过程中的不确定性,显著提升了装配成功率和装配效率。
Need to check novelty before this filing date? Find Prior Art