2自由度机器人转动关节LuGre摩擦模型参数辨识

发布时间：2018-01-20 06:04

本文关键词： 2自由度机器人转动关节 LuGre摩擦模型关节转矩控制粒子群算法参数辨识　出处：《华侨大学》2013年硕士论文　论文类型：学位论文

【摘要】：机器人关节作为机器人系统的基础部件，其性能的好坏将直接影响机器人的性能。由于关节中都存在着摩擦，而由摩擦引起的定位与跟踪误差、极限环振荡等现象已成为阻碍机器人性能提高的主要障碍，因此合理地解决关节中存在的非线性摩擦的问题已成为机器人学研究的难点问题之一。对这一问题的分析准确与否将会对机器人的控制效果产生直接的影响。由于摩擦现象非常复杂，目前还很难从其原理上获得准确唯一的数学模型，在已知的模型中，LuGre摩擦模型能较好地解释摩擦力的静态和动态特性，但是LuGre摩擦模型参数的辨识却非常困难。不失一般性，本文研究2自由度机器人转动关节LuGre摩擦模型参数辨识问题，如下：首先，在2自由度机器人动力学方程的基础上，结合LuGre摩擦模型，建立2自由度机器人转动关节的LuGre摩擦模型。其次，通过对2自由度机器人转动关节的LuGre摩擦模型进行理论分析，，获得静力参数和动力参数辨识的最小化目标函数，同时指出摩擦参数的辨识必须在开展关节转矩控制实验的基础上方能实现；接着在详细分析粒子群算法结构、参数设置及算法改进等关键技术的基础上，用MatLab编写基于粒子群算法的LuGre摩擦模型静力参数和动力参数辨识程序。再次，构建基于交流伺服电机以实现关节转矩控制为目标的2自由度机器人实验样机，并对实验样机进行参数设置和调试，编写基于关节转矩控制的伺服电机运动数据读取程序。最后，进行基于实验样机的参数辨识实验研究，分别对转动关节进行独立控制的参数辨识实验和同时控制的参数辨识实验。实验结果比较理想，其表明本文的实验辨识方法是可行的，但辨识精度有待进一步提高。
[Abstract]:As the basic component of robot system, the performance of robot joints will directly affect the performance of the robot. Because of the friction in the joints, and the friction caused by the positioning and tracking errors. Limit cycle oscillation has become the main obstacle to the improvement of robot performance. Therefore, reasonably solving the problem of nonlinear friction in joints has become one of the difficult problems in robotics research. Whether the analysis of this problem is accurate or not will have a direct impact on the control effect of the robot. Friction is very complex. At present, it is difficult to obtain the exact and unique mathematical model from its principle. Among the known models, LuGre friction model can explain the static and dynamic characteristics of friction. But the parameter identification of LuGre friction model is very difficult. In this paper, the parameter identification of the LuGre friction model of a two-degree-of-freedom robot rotation joint is studied, which is as follows: Firstly, based on the dynamic equation of two-degree-of-freedom robot and LuGre friction model, the LuGre friction model of two-degree-of-freedom robot rotation joint is established. Secondly, the minimum objective function of static and dynamic parameters identification is obtained by theoretical analysis of the LuGre friction model of the rotating joint of a two-degree-of-freedom robot. At the same time, it is pointed out that the identification of friction parameters must be realized on the basis of joint torque control experiment. Then on the basis of detailed analysis of particle swarm optimization structure, parameter setting and algorithm improvement and other key technologies. The identification program of static and dynamic parameters of LuGre friction model based on particle swarm optimization (PSO) is programmed by MatLab. Thirdly, a two-degree-of-freedom robot experimental prototype based on AC servo motor is constructed to achieve joint torque control, and the parameters of the experimental prototype are set up and debugged. Write servo motor motion data reading program based on joint torque control. Finally, the parameter identification experiment based on the experimental prototype is carried out. The parameters identification experiment of independent control and simultaneous control of rotating joint is carried out, respectively. The experimental results are satisfactory. It shows that the experimental identification method in this paper is feasible, but the identification accuracy needs to be further improved.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP242;TH117.1

【相似文献】