Learning object models for whole body manipulation
Learning object models for whole body manipulation
复制标题
学习全身操纵的对象模型
DOI：
--
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
S. Kagami
中科院分区：
文献类型：
--
作者：
Mike Stilman;K. Nishiwaki;S. Kagami
我们提出了一个成功的实现刚性抓取操作的大型物体沿着指定的轨迹移动的人形机器人。HRP-2操纵脚轮上的桌子，其负载范围可达其自身质量。机器人通过控制重心来补偿反射力来保持动态平衡。为了实现对大型未确定动力学对象的高性能，机器人学习了每个对象的摩擦模型，并将其应用于躯干轨迹生成。我们将这种方法与纯粹的反应策略进行了经验比较，并显示出预测能力和稳定性的显着增加。
We present a successful implementation of rigid grasp manipulation for large objects moved along specified trajectories by a humanoid robot. HRP-2 manipulates tables on casters with a range of loads up to its own mass. The robot maintains dynamic balance by controlling its center of gravity to compensate for reflected forces. To achieve high performance for large objects with unspecified dynamics the robot learns a friction model for each object and applies it to torso trajectory generation. We empirically compare this method to a purely reactive strategy and show a significant increase in predictive power and stability.