伺服系统瞬态优化的模糊自适应深度强化学习方法

魏晓晗; 张庆; 蒋婷婷; 梁霖

doi:10.7652/xjtuxb202108009

您当前的位置：

首页 >

文章列表页 >

伺服系统瞬态优化的模糊自适应深度强化学习方法

更新时间：2025-07-09

- 伺服系统瞬态优化的模糊自适应深度强化学习方法
- Fuzzy Adaptive Deep Reinforcement Learning Method for Optimizing Transient Performance of Servo System
- 西安交通大学学报 2021年55卷第8期页码：68-77
- 作者机构：
  
  西安交通大学机械工程学院,西安,710049
- 作者简介：
- 基金信息：
  
  : 2020-12-02。作者简介: 魏晓晗(1995—),男,博士生;张庆(通信作者),男,副教授。基金项目: 国家自然科学基金资助项目(51675405)
- DOI：10.7652/xjtuxb202108009
  中图分类号： TP181
- 网络首发：2021-08-10，
  
  纸质出版：2021
- 稿件说明：
移动端阅览
魏晓晗, 张庆, 蒋婷婷, 等. 伺服系统瞬态优化的模糊自适应深度强化学习方法[J]. 西安交通大学学报, 2021,55(8):68-77.

Fuzzy Adaptive Deep Reinforcement Learning Method for Optimizing Transient Performance of Servo System[J]. 2021, 55(8): 68-77.
魏晓晗, 张庆, 蒋婷婷, 等. 伺服系统瞬态优化的模糊自适应深度强化学习方法[J]. 西安交通大学学报, 2021,55(8):68-77. DOI： 10.7652/xjtuxb202108009.

Fuzzy Adaptive Deep Reinforcement Learning Method for Optimizing Transient Performance of Servo System[J]. 2021, 55(8): 68-77. DOI： 10.7652/xjtuxb202108009.

摘要

为了提高伺服系统瞬态性能

满足工业生产线的高精度加工要求

提出模糊自适应深度强化学习方法

用于永磁同步电机伺服系统的控制性能优化。根据伺服系统瞬态运行的响应快速性和稳定性要求

通过构造奖励函数与Actor-Critic网络

在伺服控制器中引入瞬态响应优化环节

面向实时运算需求设计学习率自适应模糊控制器

构建模糊自适应瞬态性能优化方法; 建立伺服系统Simulink仿真模型

根据实际伺服系统拟合仿真模型

通过瞬态性能仿真获得优化参数

并将优化参数应用于西门子840D数控系统。以仿真运算及系统实验对该优化方法进行验证

结果表明

所提出方法使伺服系统调节时间缩短10%以上

显著提高了系统的瞬态响应速度

且未引入明显超调

验证了方法的可行性。提出的方法具有较强的通用性

为伺服系统智能控制优化提出了新的途径。

Abstract

To improve the transient performance of servo systems and satisfy the high-precision processing requirements of industrial production lines

a fuzzy adaptive deep reinforcement learning algorithm is proposed to optimize the control performance of permanent magnet synchronous motor servo system. According to the rapidity and stability requirements of transient response

the reward function and Actor-Critic networks are constructed

the transient optimization link for servo controller is introduced

and the learning rate fuzzy controller for real-time computing requirements is designed. A servo system Simulink simulation model is established and fitted in with the actual servo system

the optimized parameters are obtained via transient performance optimization simulation and then applied to a Siemens 840D computerized numerical control system. The verification of simulation calculation by system experiment shows that the proposed method of transient performance optimization shortens the servo system adjustment time by more than 10%

significantly heightens the system transient response rate without introducing an obvious overshoot. This method has strong versatility to provide a new way for the intelligent control optimization of the servo system.

关键词

Keywords

references

鲁文其, 胡育文, 梁骄雁, 等. 永磁同步电机伺服系统抗扰动自适应控制 [J]. 中国电机工程学报, 2011, 31(3): 75-81.

LU Wenqi, HU Yuwen, LIANG Jiaoyan, et al. Anti-disturbance adaptive control for permanent magnet synchronous motor servo system [J]. Proceedings of the CSEE, 2011, 31(3): 75-81.

刘海财, 李光军, 李树胜. 基于改进型滑模观测器PMSM转子位置估计 [J]. 微特电机, 2017, 45(11): 23-25, 29.

LIU Haicai, LI Guangjun, LI Shusheng. Rotor position estimation method for permanent magnet synchronous motor based on the modified sliding mode observer [J]. Small Special Electrical Machines, 2017, 45(11): 23-25, 29.

邢向上, 姜新建. 飞轮储能系统电机及其控制器概述 [J]. 储能科学与技术, 2015, 4(2): 147-152.

XING Xiangshang, JIANG Xinjian. Introduction to motors and controllers of flywheel energy storage systems [J]. Energy Storage Science and Technology, 2015, 4(2): 147-152.

CALNETIX Technologies. Vycon direct connect kinetic energy storage systems [EB/OL]. [2020-11-20]. http: ∥www.Calnetix.com.

ZHANG Qing, TAN Luyao, XU Guanghua. Evaluating transient performance of servo mechanisms by analysing stator current of PMSM [J]. Mechanical Systems and Signal Processing, 2018, 101: 535-548.

胡强晖, 胡勤丰. 全局滑模控制在永磁同步电机位置伺服中的应用 [J]. 中国电机工程学报, 2011, 31(18): 61-66.

HU Qianghui, HU Qinfeng. Global sliding mode control for permanent magnet synchronous motor servo system [J]. Proceedings of the CSEE, 2011, 31(18): 61-66.

吴振顺, 姚建均, 岳东海. 模糊自整定PID控制器的设计及其应用 [J]. 哈尔滨工业大学学报, 2004, 36(11): 1578-1580.

WU Zhenshun, YAO Jianjun, YUE Donghai. A self-tuning fuzy PID controller and its application [J]. Journal of Harbin Institute of Technology, 2004, 36(11): 1578-1580.

蔚永强, 郭宏, 谢占明. 直驱式伺服系统的神经网络自适应滑模控制 [J]. 电工技术学报, 2009, 24(3): 74-79, 85.

WEI Yongqiang, GUO Hong, XIE Zhanming. Neural network adaptive sliding mode control for direct-drive servo system [J]. Transactions of China Electrotechnical Society, 2009, 24(3): 74-79, 85.

PANE Y P, NAGESHRAO S P, KOBER J, et al. Reinforcement learning based compensation methods for robot manipulators [J]. Engineering Applications of Artificial Intelligence, 2019, 78: 236-247.

陈奇石. 强化学习在仿人机器人行走稳定控制上的研究及实现 [D]. 广州: 华南理工大学, 2016: 41-56.

HAN T, NAGESHRAO S, FILEV D P, et al. An online evolving framework for advancing reinforcement-learning based automated vehicle control [J]. IFAC-PapersOnLine, 2020, 53: 8118-8123.

HORGAN D, QUAN J, BUDDEN D, et al. Distributed prioritized experience replay [EB/OL]. [2020-11-20]. https: ∥arxiv. org/abs/1803.00933.

张肖平, 于金飞, 崔英英, 等. 考虑铁损的永磁同步电动机有限时间模糊动态面控制 [J]. 机械制造与自动化, 2021, 50(1): 49-53, 56.

ZHANG Xiaoping, YU Jinfei, CUI Yingying, et al. Finite-time fuzzy dynamic surface control of PMSM with iron losses [J]. Machine Building Automation, 2021, 50(1): 49-53, 56.

陈玄, 李祥飞, 周杨. 基于新型滑模扰动观测器的永磁同步电机控制 [J]. 电机与控制应用, 2020, 47(3): 23-27.

CHEN Xuan, LI Xiangfei, ZHOU Yang. Permanent magnet synchronous motor control based on new sliding mode disturbance observer [J]. Electric Machines Control Application, 2020, 47(3): 23-27.

潘玉成, 林鹤之, 陈小利, 等. 基于模糊RBF神经网络的PID控制方法及应用 [J]. 机械制造与自动化, 2019, 48(3): 215-219.

PAN Yucheng, LIN Hezhi, CHEN Xiaoli, et al. PID control method based on fuzzy-RBF neural network and its applicaiton [J]. Machine Building Automation, 2019, 48(3): 215-219.

赵珊珊, 李晓庆, 纪志成. 基于模型参考模糊自适应控制的永磁同步电机控制器设计 [J]. 电机与控制应用, 2005, 32(8): 20-25.

ZHAO Shanshan, LI Xiaoqing, JI Zhicheng. A new design of speed controller of PMSM based on model reference fuzzy adaptive control method [J]. Electric Machines Control Application, 2005, 32(8): 20-25.

张腾, 张小栋, 张英杰, 等. 引入深度强化学习思想的脑-机协作精密操控方法 [J]. 西安交通大学学报, 2021, 55(2): 1-9.

ZHANG Teng, ZHANG Xiaodong, ZHANG Yingjie, et al. A precise control method for brain-computer cooperation with deep reinforcement learning [J]. Journal of Xi'an Jiaotong University, 2021, 55(2): 1-9.

赵星宇, 丁世飞. 深度强化学习研究综述 [J]. 计算机科学, 2018, 45(7): 1-6.

ZHAO Xingyu, DING Shifei. Research on deep reinforcement learning [J]. Computer Science, 2018, 45(7): 1-6.

夏金, 孙宏波, 孙立民. 基于强化学习的生产再决策问题 [J]. 计算机集成制造系统, 2019, 25(11): 2935-2942.

XIA Jin, SUN Hongbo, SUN Limin. Reinforcement learning for production reschedule [J]. Computer Integrated Manufacturing Systems, 2019, 25(11): 2935-2942.

DABNEY W, KURTH-NELSON Z, UCHIDA N, et al. A distributional code for value in dopamine-based reinforcement learning [J]. Nature, 2020, 577(7792): 671-675.

郑飞, 吴钦木. 基于模糊PI的永磁同步电机矢量控制 [J]. 计算机与现代化, 2021(1): 7-11.

ZHENG Fei, WU Qinmu. Vector control of permanent magnet synchronous motor based on fuzzy PI [J]. Computer and Modernization, 2021(1): 7-11.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向变形滑翔飞行器的智能变形决策方法

联合联邦学习与深度强化学习的并行服务功能链部署算法

基于深度强化学习的变形滑翔飞行器智能变形决策方法

微型电动车用永磁无刷电机转矩脉动控制研究

伺服系统Hammerstein非线性模型及参数辨识方法研究