未标定视觉的双臂自适应控制  

Uncalibrated Vision Based Adaptive Control for Dual-Arm

作　　者：王勇[1,2] 刘治 

机构地区：[1]广东工业大学自动化学院,广东广州510006 [2]广东外语外贸大学教育技术中心,广东广州510420

出　　处：《计算机仿真》2016年第10期305-309,390,共6页Computer Simulation

摘　　要：在双机械臂协同工作的情况下,采用"手—眼"构型的视觉系统反馈目标对象的图像信息,通过未标定视觉伺服控制方法实现了图像空间和力空间上的跟踪控制任务。首先通过在线自适应相机参数估计方法估计相机参数,然后利用估计的相机参数将图像空间的误差映射到关节空间,最后通过位置/力混合控制方法利用图像空间和关节空间的滑模变量设计了驱动力控制器。在理论上控制器能够保证位置和力的控制误差趋于零,通过仿真验证了控制器的有效性。In the paper,an uncalibrated visual servo control scheme was used to implement the tracking control task in image space and force space under condition of cooperation of dual- arm,where the vision system with eye-to- hand configuration was adopted to feedback the image information of the target object. Firstly,the camera parameters were estimated by online adaptive estimating method,next the error in image space was mapped into joint space using the estimated camera parameters,subsequently the driving force controller was designed with the sliding mode variables defined in image space and joint space using position / force hybrid control scheme. The controller can guarantee the control errors of position and force convergent to zero,and finally the effectiveness of the proposed controller was demonstrated in simulations.

关 键 词：协同控制 自适应控制 视觉伺服 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]