基于柔性腕手的最优抓取规划研究

Research on optimal grasping planning based on flexible wrist-hand

在食品自动化加工行业中,由于抓取对象具有软、脆、易损,尺寸、形状差异较大等特性,机器人多指手的抓取规划和稳定性问题尚未得到完善解决。为此,设计了一款柔性腕手,并提出了相应的最优抓取规划算法,以降低其抓取规划和控制的难度。首先,利用软体手指和仿人手掌的变掌机构研发了柔性腕手:基于变掌机构的构型变换,柔性腕手手指的工作空间以及抓取点位置发生变化,从而可抓取尺寸、形状差异较大的物体;基于软体手指的柔顺性,柔性腕手可抓取软、脆、易损的物体。然后,通过实际抓取分析,分别确定了柔性腕手利用四指、三指、二指抓取物体时的稳定性评价指标,提出了基于柔性腕手的最优抓取规划算法,并通过抓取算例验证了该算法的可行性。最后,制作了柔性腕手样机并搭建了相应的实验平台,选取8种不同的物体,演示了柔性腕手的构型变换,通过实验分析再次验证了基于柔性腕手的最优抓取规划算法的可行性。结果表明基于柔性腕手的最优抓取规划算法不仅能提高机器人多指手的适应性和抓取稳定性,而且能降低其抓取规划和控制的难度。研究结果为机器人多指手的应用提供了参考。

In the automatic food processing industry,due to the characteristics of soft,brittle,vulnerable and large differences in size and shape of grasped objects,the grasping planning and stability problems of robotic multi-finger hand have not been solved.Aiming at the aboved problems,a flexible wrist-hand was designed,and an optimal grasping planning algorithm was proposed to reduce the difficulty of grasping planning and control.Firstly,the flexible wrist-hand was developed by using the soft fingers and the alterable palm mechanism of humanoid hand.Based on the configuration transformation of the alterable palm mechanism,the working space and the grasping point position of fingers were able to change,so the flexible wrist-hand could grasp the objects with large differences in size and shape.Based on the flexibility of soft fingers,the flexible wrist-hand could grasp soft,brittle and vulnerable objects.Then,the stability evaluation indexes of the flexible wrist-hand when grasping objects with four,three and two fingers were determined through the actual grasping analysis,and the optimal grasping planning algorithm based on the flexible wrist-hand was proposed.The feasibility of this algorithm was verified by grasping instance calculation.Lastly,a flexible wrist-hand prototype was made and an experimental platform was built,and eight objects was selected to demonstrate the configuration transformation of flexible wrist-hand and the feasibility of optimal grasping planning algorithm based on the flexible wristhand was verified again by the experimental analysis.The results indicated that the optimal grasping planning algorithm based on the flexible wrist-hand could not only improve the adaptability and grasping stability of the robotic multi-finger hand,but also reduce the difficulty of grasping planning and control.The research results provide reference for the application of robotic multi-finger hand.