SIA OpenIR  > 机器人学研究室
基于电机电流的机器人运动控制与力控制技术研究
Alternative TitleResearch of Robot Motion Control and Force Control Based on Motor Current
宋吉来1,2
Department机器人学研究室
Thesis Advisor曲道奎 ; 徐方
ClassificationTP242
Keyword新一代机器人 动力学建模 伺服控制 奇异摄动控制 阻抗控制
Call NumberTP242/S 86/2016
Pages105页
Degree Discipline模式识别与智能系统
Degree Name博士
2016-06-01
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract本论文针对当前机器人发展中的难点和热点问题,研究了不使用力传感器的基于电机电流的机器人运动控制和力控制方法,并以工业生产中一种典型的洁净机器人为研究对象进行系统实验。洁净机器人是半导体加工行业中一种在洁净环境下工作的专用机器人,主要实现晶圆在真空腔中的搬运和传输,在搬运晶圆的过程中,如果与周围环境或人发生碰撞将可能损坏晶圆或机器人自身,甚至导致整个生产线的停止,严重影响生产线效率。因此,洁净机器人的安全保护功能对整个系统的可靠性、安全性尤为重要。本课题以科技部“极大规模集成电路制造装备及成套工艺”科技重大专项“硅片集成传输系统研发和示范应用”项目为依托,对洁净机器人的运动控制、安全保护和力控制方法等进行深入研究。本文通过底层的建模和分析,将伺服控制、机器人学、控制理论三方面技术融合,提出一种新的机器人运动控制和力控制方法,实现机器人高精度运动控制以及阻抗控制、柔顺控制等力控制功能。主要贡献有:(1) 针对SCARA结构的洁净机器人,研究了解耦运动学和动力学建模方法,应用关节角度和连杆长度的耦合关系对动力学方程进行简化,降低了动力学模型的复杂度。并应用最小二乘法对系统参数辨识,提高了模型的精度。(2) 针对交流永磁同步电机,建立关节伺服系统模型,并设计伺服控制器,建立电机转矩模型,通过电机电流观测机器人关节转矩,为机器人关节转矩模型和关节转矩观测器的建立打下基础。(3) 建立基于电流的关节转矩观测器和反应转矩观测器。通过分析状态观测器、卡尔曼滤波等不同的力估计方法,设计适用于干扰具有多变性、不确定性和随机性情况下的基于扰动观测器的力估计方法。根据电机系统和机器人系统模型,分别设计关节转矩观测器和反应转矩观测器,观测器的输出用于关节转矩闭环控制、阻抗控制、碰撞保护等。(4) 针对多关节机器人系统模型具有高阶、非线性、强耦合等特点,且模型的精度难以保证,控制结构复杂。提出了一种基于奇异摄动的控制方法,将系统按照时间尺度的不同分为快慢两个子系统,进而按照奇异摄动理论分别给出保证两个子系统稳定的控制规律。通过快慢两个子系统的分解,实现降阶作用,简化系统的控制器设计。(5)设计了笛卡尔空间基于力的阻抗控制器,并在关节空间实现笛卡尔空间经典阻抗控制问题。采用无源性理论对关节空间基于力的机器人阻抗控制策略进行理论分析,引入一种新的关节转矩反馈实现对电机转子惯量的整定,从物理上对关节转矩反馈进行解释。通过关节转矩反馈,电机动力学与关节转矩反馈所组成的系统是一个无源子系统,同时刚性机器人的动力学也是一个无源子系统,二者构成两个内部反馈连接的两个无源系统,根据无源性定理,两个子系统内部反馈连接所组成的闭环系统也是无源的。通过建立Lyapunov函数证明了系统的稳定性,并且结合LaSalle不变原理得到系统的渐进稳定性。
Other AbstractThis paper focuses on the current difficult and hot issue in the development of robots, and proposes an approach of robot motion and force control without force sensor but use motor current. Experiments are carried out on a kind of typical industrial cleaning room robot. Cleaning room robot is a specialized robot that work under the clean environment in semiconductor processing industry, who’s main task is to handling and transport the wafer in a vacuum chamber. In the process of handling wafers, if the collision of robot and environment or people around occured, it may damage the wafer or the robot itself, even affecting the efficiency of the whole production line. So the safety protection function of cleaning room robot is particularly important to the security and reliability of the whole system. This PhD thesis is based on the national science and technology major projects "silicon integrated transmission system development and demonstration application", and further research is carried out on the movement control, safe and protection and force control approach for cleaning room robot. Through detailed modeling and analysis of the system, servo control, robotics and control theory are integrated in this paper. These three theories put forward a new method for robot motion control and force control, which is useful to realize high precision motion control, and force control like impedance control, compliance control, etc. Main contributions of this paper are as follows: (1) Decoupled kinematics and dynamics modeling method is studied based on the structure of the SCARA type cleaning room robot. Dynamic equation of the robot can be simplified by applying joint angles and links length relationship, which reduces the complexity of the dynamic model. And the least square method is used for parameter identification of system, which improving the accuracy of the model. (2) The joint servo system model is established for permanent magnet synchronous motor. At the same time, by designing of the servo controller, motor torque model is established, which is the basic for the observation of robot joint torque by motor current. Model the joint servo system is the foundation for further work of joint torque model and joint torque observer. (3) Joint torque observer and reaction torque observer are established based on motor current. By analyzing the different force estimation methods, such as state observer, Kalman filter design and disturbance observer, a method based on disturbance observer is designed for force estimation, which is suitable for the variability, uncertainty and randomness of the system disturbance. Joint torque observer and reaction torque observer are respectively designed, the output of the observer is used to joint torque closed loop control, robotic active compliance control, collision protection, etc. (4) For multi-joint robot system model has the characteristics of high order, nonlinear, strong coupling, the precision of the model is difficult to guarantee, and control structure is complex, this paper proposes a control method based on singular perturbation, which divides the system into two subsystems according to the different time scales, and then according to the singular perturbation theory, respectively control laws guaranteed the stability of the two subsystems is given. By decomposing the two subsystem, the order reduction is realized, which simplify the controller design of system. (5) Cartesian impedance controller is designed based on force, and it is implemented in joint space. Passivity theory is introduced for the theoretical analysis of joint space impedance control strategy, and a new joint torque feedback to achieve the motor rotor inertia tuning is introduced, which physically explain the joint torque feedback. Through joint torque feedback, the motor dynamics and joint torque feedback system is composed of a passive subsystem, at the same time dynamics of rigid robot is also a passive subsystem, and the two passive sbusystem are internal feedback connected. According to the passive theorem, the closed loop system internal feedback connected of two passive subsystems is also passive. By establishing Lyapunov function the stability of the system is proved, and by combining with LaSalle invariance principle asymptotic stability of the system is got.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/19673
Collection机器人学研究室
Affiliation1.中国科学院沈阳自动化研究所;
2.中国科学院大学
Recommended Citation
GB/T 7714
宋吉来. 基于电机电流的机器人运动控制与力控制技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2016.
Files in This Item:
File Name/Size DocType Version Access License
基于电机电流的机器人运动控制与力控制技术(3724KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[宋吉来]'s Articles
Baidu academic
Similar articles in Baidu academic
[宋吉来]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[宋吉来]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.