基于仿人行为的快速连续反应-高避障灵巧臂轨迹规划与控制

项目名称： 基于仿人行为的快速连续反应-高避障灵巧臂轨迹规划与控制

项目编号： No.61273340

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 任子武

作者单位： 苏州大学

项目金额： 80万元

中文摘要： 七自由度灵巧臂轨迹规划与控制技术因其在仿人机器人领域中重要地位一直是一研究热点，它也是仿人机器人走向实际应用要解决关键问题之一。目前灵巧臂轨迹规划及控制方法已有一些成果，但在快速连续反应作业中还存在实时性低、运动控制适应性较弱等问题。本项目借鉴人体手臂对高速物体快速连续反应作业的行为机制，针对作业高实时性、避障及对载体反作用优化等问题，将灵巧臂对高速物体作业轨迹规划问题转化为轨迹参数化优选问题，研究一种仿人自然构形姿态相似性最优的灵巧臂轨迹参数化方法，和对载体反作用优化的灵巧臂仿人作业轨迹参数化方法，通过学习构建"知识"数据库并进化实现灵巧臂快速连续反应作业。同时借鉴人体上肢运动控制机理及其适应特性，研究一种前馈和神经网络自适应反馈控制相结合的柔性关节位置控制方法，实现灵巧臂高性能轨迹跟踪控制。本项目不仅为其它机器人相关领域研究提供一种新思路，而且对推动仿人机器人实际应用具有重要现实意义。

中文关键词： 冗余仿人臂；仿人行为机制；轨迹优化；柔顺位置控制；“知识”数据库

英文摘要： Trajectory planning and control of dexterous arm with seven degrees of freedom(7-DOF) is a hot research topic because of its important role in the field of humanoid robot. It is also one of the key problems to be solved during the process of humanoid robot towards its practical application. Currently there have been some results about the trajectory planning and control algorithms of the dexterous arm, but the shortcomings of low real-time and weak control adaptability in rapidly continuous response operations still exist. This project imitates the human behavior mechanisms that the human arm operates with rapid succession response on high-speed objects. Aiming to the problem of high real-time operation, obstacle avoidance, and the reaction optimization etc,the trajectory planning of dexterous arm for high-speed object is transformed into the trajectory parameterization optimization problem.Then a kind of trajectory parameterization method of dexterous arm based on similarity optimal of imitation human natural configuration & posture, and a kind of trajectory parameterization method of dexterous arm's humanoid motion based on the reaction optimization on the carrier are researched in this project. Through learning and evolving continuously knowledge database of dexterous arm is built to realize rapid succession

英文关键词： redundant manipulator；humanoid behavior mechanism；trajectory optimization；flexible position control；knowledge database