空间机器人抓捕未知目标的动力学参数辨识  

Dynamics Parameters Identification in Space Robot Capturing Unknown Target

作　　者：阳方平[1,2] 陈鹏[1,2] 王雪竹[1,2] 李洪谊[1] 王越超[1] 

机构地区：[1]中国科学院沈阳自动化研究所机器人学国家重点实验室,沈阳110016 [2]中国科学院研究生院,北京100049

出　　处：《载人航天》2013年第2期56-60,共5页Manned Spaceflight

摘　　要：空间机器人抓捕非合作未知目标会改变系统的动力学参数,并且降低系统路径规划和控制等的精度和准确性。因此在运动控制时,需要对目标物体进行参数辨识。现有在线辨识方法主要基于线动量及角动量守恒方程,将目标物体作为末端杆件或末端执行器的一部分。这些方法的辨识结果是末端杆件的动力学参数而非目标物体本身的参数,并且需要测量多种传感信息。与现有方法不同,在惯性坐标系中基于达朗贝尔原理和欧拉方程,提出了利用六维腕力传感器测得机器人末端力和力矩对未知目标的动力学参数进行辨识的新途径。为验证方法的有效性,采用ABB公司生产的IRB120机器人设计了验证试验,并用等效失重模型来消除机器人自身重力和目标物体重力的影响。试验结果表明该方法可以获得较高的辨识精度,适合实际工程应用。Space robot＇s capturing non-cooperative unknown target will change the system＇s dynamics parameters , reduce the precision and accuracy of the system path planning and control. Therefore, orbiting parameters identification of unknown target is necessary for motion control in space robot. Based on the conservation principle of linear and angular momentum, the existing-identification methods, which consider target as one part of the the end-point and end-effectors of the space robot, are dynamics parameters of the end-point and end-effectors but not the dynamics parameters identification of the unknown target. Besides, they need to measure a variety of sensor information to identify the object parameters. A new method is proposed based on D＇Alembert principle and Euler equation in an inertial frame of reference, which could identify dynamics parameters of unknown with measurement of force and moment exerted on the end-point of the robot through the six-dimensional wrist force sensor. To show the validity of the proposed scheme, an experiment is conducted by using an equivalent-weightlessness model to remove the effects of the gravity of space robot and object, with the IRB 120 robot produced by ABB Company. The resuhs show that the method can gain high identification precision and is suitable for practical engineering application.

关 键 词：参数辨识 抓捕目标 空间机器人 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TP242[自动化与计算机技术—控制科学与工程]