Servo System Identification Based on Curve Fitting to Phase-Plane Trajectory
Servo System Identification Based on Curve Fitting to Phase-Plane Trajectory
复制标题
DOI：
10.1115/1.4045431
复制
发表时间：
2020-03
影响因子：
1.7
通讯作者：
Xiao-Xing Shi;Yong Han;Jianhua Wu;Z. Xiong
中科院分区：
文献类型：
--
作者：
Xiao-Xing Shi;Yong Han;Jianhua Wu;Z. Xiong
伺服系统参数辨识方法很多。然而，目前广泛应用的离线辨识方法仍存在一些问题，如基于描述函数的继电反馈法存在难以消除的逼近误差，最小二乘法需要加速度信息等。为了用较少的伺服系统信息准确、高效地辨识系统参数，提出了一种借助一种优化方法对相平面轨迹进行曲线拟合的伺服系统参数辨识新方法。推导了单自由度系统的相平面轨迹表达式，详细描述了将伺服系统辨识问题转化为曲线拟合优化问题的过程，并介绍了初始参数设置的指导原则。仿真和实验验证了该方法的有效性。最后，设计了基于辨识参数的前馈控制器，进一步验证了辨识的准确性。
Many methods have been proposed to identify servo system parameters. However, problems still remain in widely applied offline identification methods, for example, the describing-function-based relay feedback method has the ineradicable approximation error, and acceleration information is indispensable for the least-squares method. In order to identify systems accurately and efficiently with less servo system information, this article proposes a novel method to identify servo system parameters through curve fitting to the phase-plane trajectory under the help of one optimization method. Specifically, the phase-plane trajectory expression of the single-degree-of-freedom system is derived; the process on how to convert the servo system identification problem to a curve-fitting optimization problem is described in detail; and the guidelines of the initial parameter setting are introduced. Simulations and experiments are carried out to verify the efficiency of the proposed method. Finally, a feed-forward control based on the identified parameters is designed to further validate the identification accuracy.