基于人体动作姿态识别的机器人仿人运动

发布时间：2017-12-26 21:07

本文关键词：基于人体动作姿态识别的机器人仿人运动　出处：《机械工程学报》2016年21期 　论文类型：期刊论文

【摘要】：以关节式机器人为对象,进行机器人仿人运动研究。从人体动作姿态识别、人-机动作映射、机器人运动控制等方面,详细阐述机器人仿人运动算法。提出人体动作姿态识别方法,利用Kinect传感器捕获人体运作的关节点位置信息,在建立人体基准坐标系的基础上,为了得到描述肩、肘运动的动作信息,计算人体手臂动作的关节角度,实现人体动作姿态的识别。在分析人体肩、肘等关节和机器人机构差异性的基础上,建立人体手臂与四自由度机械手臂的人-机动作映射规则。针对机器人自由度较少,无法完全复现人体运动的情形,分析、比较不同控制策略的优缺点和适用性,寻求适合机器人操作的复现控制策略。关节式机器人接收运动控制指令,执行相应的关节运动,从而实现机器人仿人运动。相关试验验证了人体动作姿态识别和机器人仿人运动控制算法的有效性。研究成果对于提高机器人控制和操作的简单易用性、提高人机交互能力具有借鉴意义,对于扩展机器人应用领域具有实践意义。
[Abstract]:The research of humanoid motion is carried out with the joint robot as the object. The humanoid motion algorithm of robot is described in detail from the aspects of human movement attitude recognition, human machine motion mapping, robot motion control and so on. The human gesture recognition method, joint position information capture operation of the human body by using Kinect sensor, based on the establishment of the human reference coordinate system, in order to get a description of the shoulder and elbow movement information, joint angle calculation of human arm movements, recognition of human gestures. On the basis of analyzing the differences of human shoulder and elbow joint and robot mechanism, the rule of human - machine action mapping of human arm and four degree of freedom mechanical arm is established. In view of the fact that the degree of freedom of robot is less and can not completely resume human motion, we analyze and compare the advantages and disadvantages and applicability of different control strategies, so as to find a suitable control strategy for robot operation. The joint robot receives the motion control instructions and performs the corresponding joint motion, thus realizing the humanoid motion of the robot. The correlation test verifies the effectiveness of the human motion attitude recognition and the robot motion control algorithm. The research results are of significance for improving robot's control and operation's simplicity and ease of use, improving human-machine interaction ability, and for extending the application area of robots.
【作者单位】：上海大学机电工程与自动化学院;
【基金】：国家自然科学基金资助项目(51305248,61203351)
【分类号】：TP242
【正文快照】： 0前言*随着机器人应用从工业领域向医疗、服务、娱乐、教育等行业不断地扩展和深入,对机器人的运动控制也提出了新的要求[1]。传统的工业机器人对运动控制的效率和精度有严格要求,往往需要专业人员进行复杂的编程和校准,才能满足最终的使用要求。然而,对家庭服务机器人等其他

【相似文献】