基于非参数模型辨识的机床伺服系统OILC跟踪研究  

Machine Tool Servo System Tracking Based on Nonparametric Model Identification and Optimal Iterative Learning Control

作　　者：杨光[1] 寇爽 路晓云[1] 李峰 YANG Guang;KOU Shuang;LU Xiaoyun;LI Feng(Department of Mechanical Manufacturing,Xinxiang Vocational and Technical College,Xinxiang Henan 453000,China;Department of Electrical Engineering,Kaifeng Technician College,Kaifeng Henan 475004,China;School of Mechanical Engineering,Henan Polytechnic University,Jiaozuo Henan 454000,China)

机构地区：[1]新乡职业技术学院机械制造系,河南新乡453000 [2]开封技师学院电气工程系,河南开封475004 [3]河南理工大学机械工程学院,河南焦作454000

出　　处：《机械设计与研究》2024年第6期319-323,共5页Machine Design And Research

基　　金：河南省高等学校重点科研项目计划资助项目(18A460002)。

摘　　要：以机床伺服系统具有时变特征为基础,将最优迭代学习控制(OILC)和非参数模型(NMI)识别深度融合,提出了一种NMI-OILC算法。对机床伺服系统持续性识别,有效填补OILC难以应对系统时变缺陷。利用该算法对伺服系统运动过程进行控制,具备较强的跟踪性能。仿真结果表明:在参数缓慢变化情况下,NMI-OILC的性能会暂时下降,经过数次迭代后渐渐趋向于收敛,跟踪误差同样符合要求。即使是对于参数变化的情况,NMI-OILC仍有较强的跟踪性能。实验结果表明:在系统参数改变后,NMI-OILC算法能够让目标函数渐渐趋向于收敛,跟踪误差符合控制要求,能够高效应对系统的时变特征,大大提高系统的跟踪能力。该研究可以拓展到其它的机械传动的参数识别领域,具有很好的应用价值。Based on the time-varying characteristics of machine tool servo systems,a NMI-OILC algorithm is proposed,which deeply integrates optimal iterative learning control(OILC)and non-parametric model(NMI)recognition.Machine servo systems are continuously identified to effectively address time-varying defects that OILC is unable to cope with.The algorithm is used to control the servo system motion process and has strong tracking performance.The simulation results show that the performance of NMI-OILC is temporarily degraded under slow parameter changes,but gradually converges after several iterations,and the tracking error is also satisfactory.The NMI-OILC provides excellent tracking performance even when the parameters change.Experimental results show that the NMI-OILC algorithm can gradually converge the objective function after the system parameters change,and the tracking error meets the control requirements.It can effectively cope with the time-varying characteristics of the system,and significantly improve the tracking capability of the system.This research can be extended to other fields of parameter identification of mechanical transmission and has good application value.

关 键 词：迭代学习 伺服系统 参数辨识 最优控制 

分 类 号：TH137[机械工程—机械制造及自动化]