基于最大力螺旋多指手抓取规划

Planning of Grasping with Multifingered Hands Based on the Maximal External Wrench

机器人多指手抓取物体时,为了保证抓取稳定性,必须对手指与被抓物体之间的接触位置进行合理的布局,寻求最佳的抓取位置。基于此,提出基于最大力螺旋的多指手抓取规划方法,将多指手可以平衡的最大外力螺旋作为评价规划的性能指标,建立力封闭约束条件下抓取位置与最大外力螺旋优化模型,并运用神经网络进行仿真,将该方法与基于广义力椭球方法进行比较。前者弥补了广义力椭球方法的局限性,可以清楚地描述多指手抓取位置与可平衡外力螺旋之间的对应关系,定量地描述被抓物体的稳定性程度,使得抓取规划的判断更直观、更具有实用性。

When grasping with a multifingered hand, a reasonable arrangement should be suggested for the contact points of the fingers on an object in order to guarantee stability. A planning method of grasping based on the maximum external wrench is proposed, which takes the maximum external wrench as a criterion to evaluate the performances of the planning. After force closed constraints are given, an optimal model of the maximum external wrench is established and the simulation by using the neural network method is carried out. To compare with the numerical method of generalized force ellipsoid, the present model overcomes its limitation and can clearly describe the relationship between the grasping positions and the balanceable external force wrench. ,The model can quantitatively describe the grasping stability so as to judge a grasping planning more directly and practically.