机器人灵巧手设计与自主操作关键技术研究

For urgent needs of the industrial robot and service robot for dexterous end effector, and according to the function mismatch between autonomous planning control and dexterous mechanism, in other words, the existing planning control methods can not manipulate the dexterous hand with high dimensional DOF, this project with autonomy, intelligence, dexterity, compliance to the target, introduces the human-machine hybrid intelligence, studies the design and autonomous control method of dexterous robot hand. This project explores motion/impedance collaborative law and mathematical characterization of the manipulation behavior of human hand, introduces local mechanical intelligence and multi-modal perception in the design of dexterous hand and motion/impedance coordination model of in the control system, and uses cross coupling control to achieve soft synergy in planning layer and control layer. The human-machine intelligent integration based on operation coordination model, self-adaptive mechanism and autonomous planning control, the function mismatch between autonomous planning control and dexterous mechanism is solved, and the dexterous hand's autonomous manipulation control is realized, and typical autonomous operation experiments are carried out. The research of this project will enhance the autonomous or semi-autonomous manipulation ability of dexterous hand under unknown environment and object, and provide theoretical and technological support for autonomous operation of dexterous robot hand.

面向制造业机器人和服务机器人等对灵巧末端执行器的迫切需求，针对自主规划控制与灵巧运动机构之间功能失配、现有规划控制方法无法自主操控灵巧手高维运动的问题，本项目以自主、智能、灵巧、顺应为目标，在自主操作中引入人机混合智能，研究机器人灵巧手的设计和自主控制方法。探索人手操作行为的运动/阻抗协同规律及数学表征，在灵巧手设计中引入局部机械智能和多模态感知、在控制中引入人手运动/阻抗协同模型，基于交叉耦合控制在规划层和控制层实现“软协同”。基于人手操作协同模型、机械顺应机构和自主规划控制的人机智能融合，解决自主规划控制与灵巧运动机构的功能失配问题，实现灵巧手的自主操作控制，并且进行典型自主操作实验。本项目的研究将提升多自由度灵巧手在未知环境和对象情况下的自主或半自主作业能力，为机器人灵巧手自主操作的应用提供理论和技术支持。

面向制造业机器人和服务机器人等对灵巧末端执行器的迫切需求，针对自主规划控制与灵巧运动机构之间功能失配、现有规划控制方法无法自主操控灵巧手高维运动的问题，本项目以自主、智能、灵巧、顺应为目标，在自主操作中引入人机混合智能，研究机器人灵巧手的设计和自主控制方法。本项目建立了人手操作特性综合测试系统及解析方法，研制了具有机械顺应特性和多模态感知功能的机器人灵巧手样机，提出了面向任务的多指手动态力规划方法，提出了基于人手运动/阻抗协同模型和交叉耦合策略的多指手同步协调控制方法，在规划层和控制层实现“软协同”，进行了灵巧手操作实验。本项目基于人手操作模型、机械顺应机构和自主规划控制的人机智能融合，解决自主规划控制与灵巧运动机构的功能失配问题，实现灵巧手的自主操作控制。本项目的研究成果将提升多自由度灵巧手在未知环境和对象情况下的自主或半自主作业能力，为机器人灵巧手自主操作的应用提供理论和技术支持。