电工技术学报  2019, Vol. 34 Issue (5): 917-923    DOI: 10.19595/j.cnki.1000-6753.tces.L80703
电机与电器 |
基于在线附加Q学习的伺服电机速度最优跟踪控制方法
邹晓敏1,肖曦1,何琪2,ShkodyrevVyacheslav3
1. 清华大学电机工程与应用电子技术系 北京 100084;
2. 陕西航空电气有限责任公司 西安 710077 3. 圣彼得堡彼得大帝理工大学 圣彼得堡 195251
Optimal Tracking Control of Servo Motor Speed Based on Online Supplementary Q-Learning
Zou Xiaomin1, Xiao Xi1, He Qi2, Shkodyrev Vyacheslav3
1. Department of Electrical Engineering Tsinghua University Beijing 100084 China;
2. AVIC Shaanxi Aero Electric Co. Ltd Xi’an 710077 China;;
3. Peter the Great St. Petersburg Polytechnic University St. Petersburg 195251 Russia
全文: PDF (1084 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 该文将在线Q学习方法与附加控制思想相结合,讨论了其在伺服系统中电机速度最优跟踪控制问题上的应用。首先在线性二次型跟踪器问题的框架下对待求解问题进行了定义;然后给出了在线附加Q学习迭代式地进行策略评价、策略改善的具体算法。仿真测试中,首先为电机速度跟踪问题设计了传统的PI控制器,然后将基于该文思路所设计的附加控制器与其并联,组成新的速度控制器。仿真结果表明,附加控制器显著改善了电机速度跟踪的动态响应特性,并且具备在被控系统参数发生改变时自动调优的自适应能力。非线性系统在特定条件下可进行局部线性化时,也可用该方法来得到更优的控制性能。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
邹晓敏
肖曦
何琪
ShkodyrevVyacheslav
关键词 在线Q学习最优跟踪控制附加控制电机控制速度控制    
Abstract:This paper combined online Q-learning with supplementary control and discussed its application to the optimal tracking control problem of servo motor speed. Firstly, the problem to be solved was defined in the framework of linear quadratic tracking. Then, the iterative algorithm of policy evaluation and policy improvement for online supplementary Q-learning was given. In the simulation test, for the motor speed tracking problem in servo system, the traditional PI controller was firstly designed, and then the supplementary controller proposed in this paper was connected to it in parallel to form a new speed controller. The simulation results showed that the additional controller significantly improves the dynamic response characteristics of motor speed tracking, and has the adaptive ability to automatically adjust when parameters of the controlled system changes. When the nonlinear system can be locally linearized under certain conditions, the proposed method can also be applied to obtain better control performance.
Key wordsOnline Q-learning    optimal tracking control    supplementary control    electric machine control    speed control   
收稿日期: 2017-11-30      出版日期: 2019-03-21
PACS: TM301.2  
基金资助:国家自然科学基金(51577095)和清华大学自主科研计划支持项目资助
通讯作者: 肖曦, 男,1973 年生,教授,博士生导师,研究方向为高性能伺服电机控制、机器人驱动控制技术、海浪发电技术、电力储能与微电网技术等。E-mail:xiao_xi@tsinghua.edu.cn   
作者简介: 邹晓敏,男,1993年生,硕士研究生,研究方向为半实物仿真系统、自适应动态规划、电机控制等。E-mail:zouxm16@mails.tsinghua.edu.cn
引用本文:   
邹晓敏,肖曦,何琪,ShkodyrevVyacheslav. 基于在线附加Q学习的伺服电机速度最优跟踪控制方法[J]. 电工技术学报, 2019, 34(5): 917-923. Zou Xiaomin, Xiao Xi, He Qi, Shkodyrev Vyacheslav. Optimal Tracking Control of Servo Motor Speed Based on Online Supplementary Q-Learning. Transactions of China Electrotechnical Society, 2019, 34(5): 917-923.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.L80703          https://dgjsxb.ces-transaction.com/CN/Y2019/V34/I5/917