基于自适应动态规划机械臂运动策略仿真研究

doi:10.16182/j.issn1004731x.joss.23-FZ0828

摘要/Abstract

摘要：

针对机械臂在复杂、恶劣的环境下难以实现高精度运动跟踪的问题，提出了基于自适应动态规划（adaptive dynamic programming，ADP）与滑模导纳控制相结合的策略方法。将未知环境建模为线性模型，基于准滑动模态推导出滑模导纳控制器，用以抵抗扰动干扰；提出ADP与滑模导纳控制器相结合的最优控制方法，并对价值函数中R矩阵的定义进行了优化改进，进一步提升了跟踪精度；运用基于ADP的神经网络来逼近最优价值函数的解，提高了收敛速度，从而快速获得哈密顿-雅可比-贝尔曼（Hamilton-Jacobi-Behrmanequation，HJB）方程的近似最优策略。实验结果表明：该方法有效降低了机械臂运动的控制成本以及保证了最优轨迹跟踪。

关键词: 导纳控制, 自适应动态规划, 环境交互, 最优控制, 神经网络

Abstract:

Aiming at the difficulty of manipulator to realize high-precision motion tracking in complex and harsh environment, a strategy method based on the combination of adaptive dynamic programming (ADP) and sliding mode admittance control is proposed. The unknown environment is modeled as a linear model and based on quasi, a sliding mode admittance controller is derived to resist disturbance interference. An optimal control method that combines ADP with sliding mode admittance controller is proposed, in which the definition of R-matrix in value function is optimized and improved to further improve the tracking accuracy. The neural network based on ADP is used to approximate the solution of optimal value functionwhich improves the rate of convergence and quickly obtains the approximate optimal strategy of Hamilton-Jacobi-Behrmann equation (HJB). The method is applied in the trajectory tracking control of a robotic arm. The experimental results show that the method effectively reduces the cost of controlling robotic arm and ensures the optimal trajectory tracking.

Key words: admittance control, adaptive dynamic programming, environment interaction, optimal control, neural network

中图分类号:

TP241.2

李明,徐群,王艳等 . 基于自适应动态规划机械臂运动策略仿真研究[J]. 系统仿真学报, 2023, 35(10): 2182-2192.

Li Ming,Xu Qun,Wang Yan,et al . Simulation and Research of Manipulator Motion Strategy Based on Adaptive Dynamic Programming[J]. Journal of System Simulation, 2023, 35(10): 2182-2192.

图/表 20

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

图13

图14

图15

图16

图17

图18

图19

图20

参考文献 21

1	夏长高, 杨鹏程, 韩江义, 等. 基于遗传算法优化的除草机械臂模糊PID控制研究[J]. 农机化研究, 2022, 44(12): 15-21.
	Xia Changgao, Yang Pengcheng, Han Jiangyi, et al. Research on Control of Weeding Manipulator Arm Based on Fuzzy PID Optimized by Genetic Algorithm[J]. Journal of Agricultural Mechanization Research, 2022, 44(12): 15-21.
2	丁荣乐, 侯旋, 孙文建. 考虑变有效载荷的机械臂自适应PID控制[J]. 机电工程, 2022, 39(3): 291-299.
	Ding Rongle, Hou Xuan, Sun Wenjian. Adaptive PID Controlling Method for Robotic Manipulators Considering Variable Payloads[J]. Journal of Mechanical & Electrical Engineering, 2022, 39(3): 291-299.
3	刘鹤松, 颜文旭, 倪宏宇, 等. 基于新型自适应律的巡检机械臂滑模控制[J]. 控制工程, 2023, 30(3): 434-442.
	Liu Hesong, Yan Wenxu, Ni Hongyu, et al. Sliding Mode Control of Inspection Manipulator Based on New Adaptive Law[J]. Control Engineering of China, 2023, 30(3): 434-442.
4	许一鸣. 基于自适应动态规划的柔性机械臂组合控制[D]. 徐州: 中国矿业大学, 2019.
	Xu Yiming. Composite Control of Flexible Manipulators Based on Adaptive Dynamic Programming[D]. Xuzhou: China University of Mining and Technology, 2019.
5	Gao Hejia, He Wei, Zhou Chen, et al. Neural Network Control of a Two-link Flexible Robotic Manipulator Using Assumed Mode Method[J]. IEEE Transactions on Industrial Informatics, 2019, 15(2): 755-765.
6	Tayebi A. Adaptive Iterative Learning Control for Robot Manipulators[J]. Automatica, 2004, 40(7): 1195-1203.
7	李焕, 王奉文, 徐世杰, 等. 基于阻抗控制的机械臂末端工具的柔顺控制[J]. 空间控制技术与应用, 2019, 45(1): 20-26.
	Li Huan, Wang Fengwen, Xu Shijie, et al. Compliance Control of End-effector of Space Manipulator Based on Impedance Control[J]. Aerospace Control and Application, 2019, 45(1): 20-26.
8	杨国卿. 不确定参数条件下机械臂的滑模预测控制研究[D]. 济南: 山东大学, 2022.
	Yang Guoqing. Research on Sliding Mode Predictive Control of Robot Manipulator with Uncertain Parameters[D]. Ji'nan: Shandong University, 2022.
9	裴红蕾. 机械臂关节空间轨迹的神经网络滑模跟踪控制[J]. 制造技术与机床, 2020(5): 43-48.
	Pei Honglei. Manipulator Joint Space Trajectory Tracking Control Based on Neutral Network-sliding Mode Controller[J]. Manufacturing Technology & Machine Tool, 2020(5): 43-48.
10	刘洋洋, 院老虎, 腾英元, 等. 基于RBF神经网络补偿的自适应滑模机械手控制[J]. 组合机床与自动化加工技术, 2023(6): 119-122, 127.
	Liu Yangyang, Yuan Laohu, Teng Yingyuan, et al. Adaptive Sliding Mode Manipulator Control Based on RBF Neural Network Compensation[J]. Modular Machine Tool & Automatic Manufacturing Technique, 2023(6): 119-122, 127.
11	Peng Guangzhu, Yang Chenguang, He Wei, et al. Force Sensorless Admittance Control With Neural Learning for Robots With Actuator Saturation[J]. IEEE Transactions on Industrial Electronics, 2020, 67(4): 3138-3148.
12	韩秀梅. 基于自适应动态规划的非线性系统最优控制方法研究[D]. 大连: 大连理工大学, 2022.
	Han Xiumei. Adaptive-dynamic-programming-based Optimal Control Method for Nonlinear Systems[D]. Dalian: Dalian University of Technology, 2022.
13	王泰华, 马彬彬, 李亚飞. 基于神经网络观测器的机械臂力/位置控制[J]. 制造业自动化, 2023, 45(2): 184-188.
	Wang Taihua, Ma Binbin, Li Yafei. Mechanical Arm Force/Position Control Based on Neural Network Observer[J]. Manufacturing Automation, 2023, 45(2): 184-188.
14	张文辉, 沈金淼, 游张平, 等. 空间机械臂基于速度观测器的神经网络控制[J]. 空间控制技术与应用, 2022, 48(6): 12-21.
	Zhang Wenhui, Shen Jinmiao, You Zhangping, et al. Neural Network Control of Space Manipulator Based on Velocity Observer[J]. Aerospace Control and Application, 2022, 48(6): 12-21.
15	唐智浩. 空间机械臂动力学预测与控制方法研究[D]. 哈尔滨: 哈尔滨工业大学, 2021.
	Tang Zhihao. Research on Dynamic Prediction and Control Method of Space Manipulator[D]. Harbin: Harbin Institute of Technology, 2021.
16	Liu Derong, Yang Xiong, Li Hongliang. Adaptive Optimal Control for a Class of Continuous-time Affine Nonlinear Systems With Unknown Internal Dynamics[J]. Neural Computing and Applications, 2013, 23(7): 1843-1850.
17	Zhan Hong, Huang Dianye, Yang Chenguang. Adaptive Dynamic Programming Enhanced Admittance Control for Robots With Environment Interaction and Actuator Saturation[J]. International Journal of Intelligent Robotics and Applications, 2021, 5(1): 89-100.
18	Vamvoudakis K G, Lewis F L. Online Actor-critic Algorithm to Solve the Continuous-time Infinite Horizon Optimal Control Problem[J]. Automatica, 2010, 46(5): 878-888.
19	Zhang Beibei, Yang Xiao, Zhao Dongya, et al. Sliding Mode Control for Nonlinear Manipulator Systems[J]. IFAC-PapersOnLine, 2017, 50(1): 5127-5132.
20	Yoo B K, Ham W C. Adaptive Control of Robot Manipulator Using Fuzzy Compensator[J]. IEEE Transactions on Fuzzy Systems, 2000, 8(2): 186-199.
21	Rahmani M, Komijani H, Rahman M H. New Sliding Mode Control of 2-DOF Robot Manipulator Based on Extended Grey Wolf Optimizer[J]. International Journal of Control Automation and Systems, 2020, 18(6): 1572-1580.

[1]	张凯, 卢海鹏, 韩莹, 张龄允, 丁昱杰. 融合变分模态分解的时空卷积短时车速预测[J]. 系统仿真学报, 2023, 35(8): 1651-1660.
[2]	季玉琦, 谢欢, 史少彧, 和萍, 金楠, 王惠丽. 基于EWM-AHP-BP神经网络的地区电网电压无功组合评价[J]. 系统仿真学报, 2023, 35(4): 843-852.
[3]	蔡兴泉, 李治均, 奚梦瑶, 孙海燕. 基于神经网络的手绘服饰图纹上色及风格迁移[J]. 系统仿真学报, 2023, 35(3): 604-615.
[4]	杨丰玮, 陈鹏, 郗凯, 蒲华林, 刘雪垠. 融合空间信息的运动想象脑电在线分类方法[J]. 系统仿真学报, 2023, 35(2): 254-267.
[5]	王涛, 吴迎年, 杨睿, 孙乐音. 基于IMU与sEMG混合信号的实时手势分类算法研究[J]. 系统仿真学报, 2023, 35(2): 359-371.
[6]	李静, 张涛涛, 金凯, 袁胜智, 查子龙. 基于时变RBF神经网络的时变非线性系统控制器设计[J]. 系统仿真学报, 2023, 35(10): 2223-2236.
[7]	张会林, 金玉洁, 杨海马. ANFIS优化磁链滑模观测器的PMSM无传感器控制[J]. 系统仿真学报, 2022, 34(8): 1682-1690.
[8]	刘兴华, 耿晨, 谢胜寒, 田佳强, 曹晖. 考虑光伏发电不确定性的日前火电-光伏经济调度[J]. 系统仿真学报, 2022, 34(8): 1874-1884.
[9]	张晓青, 肖万芳, 郭英杰, 刘博文, 韩学森, 马经纬, 高高, 黄赫, 夏时洪. 融合LSTM和MoE的倒闸操作识别[J]. 系统仿真学报, 2022, 34(8): 1899-1907.
[10]	张森, 张孟炎, 邵敬平, 普杰信. 基于随机策略搜索的多机三维路径规划方法[J]. 系统仿真学报, 2022, 34(6): 1286-1295.
[11]	张立峰, 王会忍. 基于卷积神经网络及有限元仿真的电容层析成像图像重建[J]. 系统仿真学报, 2022, 34(4): 712-718.
[12]	康旭, 张晓峰. 基于生成对抗神经网络的雷达遥感数据增广方法[J]. 系统仿真学报, 2022, 34(4): 920-927.
[13]	周思锦, 陈棣成, 涂耿, 姜大志. 基于个性化和记忆机制的多模态情感计算模型[J]. 系统仿真学报, 2022, 34(4): 745-758.
[14]	王丽梅, 么红岩, 张康. 直接驱动H型平台鲁棒最优同步控制仿真研究[J]. 系统仿真学报, 2022, 34(3): 527-535.
[15]	孙晓安, 栾小丽, 刘飞. 基于智能优化灰色模型的电子固废预测[J]. 系统仿真学报, 2022, 34(3): 536-542.