中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
工业机器人关节柔性分析与减振控制

文献类型:学位论文

作者强艳辉
学位类别工学博士
答辩日期2012-12-06
授予单位中国科学院大学
授予地点中国科学院自动化研究所
导师侯增广 ; 景奉水
关键词柔性关节工业机器人 关节空间轨迹规划 输入整形 动态面法 Chebyshev神经网络 Flexible Joint Industrial Robot Trajectory Planning in Joint Space Input Shaping Technique Dynamical Surface Control Chebyshev Neural Network
其他题名Analysis of Joint Flexiblity and Vibration Reduction for an Industrial Robot
学位专业控制理论与控制工程
中文摘要工业机器人在现代工业中发挥着越来越重要的作用,为提高机器人性能,学者们对机器人建模与控制进行了深入研究,大多数研究以刚性模型为基础。然而,工业机器人关节减速机构都存在一定的柔性,特别是高速及大负载作业时,柔性表现尤为明显。关节柔性使得机器人固有频率降低,在外力和惯性力作用下,极易产生振动。振动的存在不仅影响轨迹跟踪精度和定位精度,也破坏了系统运行的稳定性和可靠性。因此,刚性模型控制器在机器人高精度应用中受到了很大限制。如果要进一步提高工业机器人的动态性能和工作品质,必须十分重视关节柔性给机器人动力学带来的影响,对有害振动加以控制。这是提高工业机器人控制精度的重要方向。 本文考虑了关节传动机构存在的柔性,从振动控制方面入手,对工业机器人进行减振控制研究。首先介绍了本课题的研究背景及意义,阐述了关节柔性产生的原因、柔性关节机器人研究现状。针对关节柔性导致的振动问题,综述了国内外学者在柔性关节机器人振动控制方面的研究成果,分析了振动控制策略应用中面临的一些问题,以此为基础,开展了进一步研究。主要内容如下: 第一,以六关节工业机器人为原型,考虑传动机构的柔性因素,基于拉格朗日能量法建立了刚性连杆柔性关节机器人动力学模型。以该模型为基础,验证了柔性关节带来的振动效应,并分析了振动特性。动力学模型是后续研究的基础。 第二,提出了一种新型关节空间轨迹规划算法,获得了更加平滑的运动轨迹。该算法具有以下特点:(1)可以实现关节空间加加速度连续;(2)可以处理机器人位于奇异位姿时的速度;(3)既可以进行在线规划,也可以完成离线规划。为更加有效利用该轨迹蕴含的平滑信息,针对柔性关节机器人采用四阶前馈控制器,结合提出的轨迹规划算法实施振动控制,仿真验证了该方法具有良好的振动控制效果。 第三,提出了一种新的两脉冲输入整形器参数学习策略。通过振动幅值测量,利用寻优方法(外推内插法和黄金分割法)离线调整整形器参数,运用最优参数整形器实施振动控制。提出的整形器参数学习策略无需振动相位的测量,不仅适用于线性系统,也适用于柔性关节机器人这种存在非线性及时变振动的系统。将所提离线学习整形器应用于柔性关节机器人的残余振动控制,分析了振动幅值和电机力矩之间的关系,利用力矩来测量振动幅值,并给出了整形器参数初始值计算方法。通过数值仿真验证了该方法的有效性。 第四,提出了基于电机状态反馈的柔性关节机器人动态面轨迹跟踪控制器,利用连杆位置和速度估计信息,使用动态面法完成控制器设计。这种基于观测器的动态面控制器相比于传统反演控制器,有以下优点:一方面,动态面法的使用降低了控制器设计难度,同时使控制器具有可调参数,方便了性能调试;另一方面,动态面法避免了高阶信号测量,连杆状态观测器避免了连杆状态的测量,因此,只需电机状态反馈即可完成控制器设计。所设计的控制器可以克服连杆动力学的不确定性,具有一定的鲁棒性。此外,由于控制器基于电机状态反馈,无需安装额外传感器,可直接应用于现有工业机器人,具有工程应...
英文摘要With the rapid development of the modern manufacture, industrial robot makes a more important role in modern industry. A lot of researchers have devoted their efforts to modeling and control of robot based on the rigid model. Nevertheless, industrial robots generally have flexible elements in the transmission systems which introduce flexibility into the joints. Especially for a high-speed operating robot with large load, joint flexibility is remarkable. Joint flexibility will decrease the natural frequencies of robot and vibrations are easily induced by the external force and inertia force. These vibrations may limit the trajectory accuracy and positioning precision, destroy the system stability and reliability. Therefore, the control algorithm designed assuming perfect rigidity at joints has limits in high precision robot control. In order to improve dynamic performance and quality of industrial robot, joint flexibility should be taken into account in both modeling and control and vibrations should be reduced. In this thesis, joint flexibility is considered and vibration control methods for industrial robot are studied. First of all, the research background and research significance of this paper are introduced. Then reasons of the joint flexibility and the research history of the flexible joint robot are summarized. A survey of work concentrated on the vibration control of flexible joint robot is presented and problems faced in vibration control for industrial robot are also analyzed. Based on the review and analysis of vibration control method for flexible joint robot, the research work is conducted. The main work and contributions of this thesis are: Firstly, the dynamical model of the rigid link flexible joint (RLFJ) robot with the structural of industrial robot is built based on the Lagrange energy method. Moreover, the vibrations properties of the model are addressed and vibration control methods are studied based on this model. Secondly, a novel trajectory planning algorithm is proposed which can insure the continuity of the positions, velocities, accelerations and jerks. Besides, the algorithm can be used when robot is at singular configurations with the inexistence of velocity. Finally, the algorithm can be applied to on-line tasks as well as off-line tasks. A forth-order feedforward controller is used in the control system to reduce the vibration of the robot in some extent, which can use the smoother trajectory information from the propo...
语种中文
其他标识符200918014628007
源URL[http://ir.ia.ac.cn/handle/173211/6493]  
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
强艳辉. 工业机器人关节柔性分析与减振控制[D]. 中国科学院自动化研究所. 中国科学院大学. 2012.

入库方式: OAI收割

来源:自动化研究所

浏览0
下载0
收藏0
其他版本

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。