基于改进傅里叶级数的机器人动力学参数辨识  

Dynamic Identification for Robot Manipulators Based on Modified Fourier Series

作　　者：吴文祥[1] 朱世强[1] 靳兴来[1] 

机构地区：[1]浙江大学流体动力与机电系统国家重点实验室,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2013年第2期231-237,共7页Journal of Zhejiang University：Engineering Science

基　　金：浙江省自然科学基金资助项目(Y1100693);浙江省重点科技创新团队资助项目(2009R50014);国家重点实验室资助项目

摘　　要：针对机器人动力学参数辨识的问题,提出一种基于改进傅里叶级数的辨识方法.采用静态连续摩擦模型描述机器人关节摩擦特性,以确保基于该模型的摩擦补偿是物理可实现的.推导得到机器人动力学模型的线性形式.设计严格满足速度、加速度边界条件的改进傅里叶级数作为激励轨迹,并根据条件数准则优化激励轨迹系数.考虑测量噪声的影响,采用最大似然法作为参数估计的方法.实验结果表明,采用所提方法的辨识结果能够准确重构机器人关节力矩值.且相比于传统的基于傅里叶级数的辨识方法,该方法提高参数辨识的精度.To deal with the problem of dynamic parameter identification of robot manipulators, a systematic identification procedure was proposed based on modified Fourier series. First, a static continuous friction model was involved to model joint friction for realizable friction compensation in controller design. Then, the robot dynamic model was expressed linearly with respect to dynamic parameter vector. Second, periodic modified Fourier series was designed as exciting trajectory, which satisfied velocity and acceleration boundary conditions. To minimize the sensitivity to measurement noise, the coefficients of modified Fourier series were optimized according to the condition number criterion using genetic algorithm. In addition, considering the measurement noise effects, maximum likelihood estimation method was adopted to obtain accurate parameter estimates. The proposed identification procedure had been implemented on the first three axes of 6-DOF robot manipulator. Experimental results show that the needed joint torques can be reconstructed accurately, and identification using modified Fourier series can achieve better accuracy than that using finite Fourier series.

关 键 词：机器人 动力学 辨识 摩擦力 激励轨迹 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置]