基于耦合算法的类人机器人模仿学习控制方法

中国机械工程

基于耦合算法的类人机器人模仿学习控制方法

李文华1;杨子凝1;王来贵2

1.辽宁工程技术大学机械工程学院，阜新，123000
2.辽宁工程技术大学力学与工程学院，阜新，123000

出版日期:2017-07-25 发布日期:2017-07-26
基金资助:
国家自然科学基金资助项目(51474121)；
辽宁省教育厅资助项目(L2015214)
National Natural Science Foundation of China （No. 51474121）

Imitation Learning Control Method of Humanoid Robots Based on Coupling Algorithm

LI Wenhua1;YANG Zining1;WANG Laigui2

1.School of Mechanical Engineering,Liaoning Technical University,Fuxin,Liaoning,123000
2.School of Mechanics and Engineering,Liaoning Technical University,Fuxin,Liaoning,123000

Online:2017-07-25 Published:2017-07-26
Supported by:
National Natural Science Foundation of China （No. 51474121）

摘要/Abstract

摘要： 为提高类人机器人模仿学习的准确性及效率，建立了一种改进的粒子群算法优化超限学习机的模仿学习模型。采用非线性动态系统对示教时的相关数据进行建模；以动态自适应策略改进粒子群算法的惯性权重，并利用改进后的粒子群算法对超限学习机的网络参数进行寻优；利用该耦合学习模型对模仿学习动态系统的参数进行学习，并重现了模仿学习动作。实验结果表明，该耦合算法应用在类人机器人模仿学习方面具有很好的拟合精度、自适应性及泛化能力，重现模仿学习动作时的平均误差为0.0172。

关键词: 模仿学习, 超限学习机, 改进粒子群优化, 非线性动态系统, 耦合模型

Abstract: To improve accuracy and efficiency in learning from demonstrations by humanoid robots, an imitation learning model was established based on improved PSO to optimize extreme learning machine, to learn human motions on the robots herein. A set of motions which was performed by a human demonstrator were collected to model as a nonlinear autonomous dynamical system. PSO was improved with the dynamic adaptive inertia weight. Then the improved PSO was merged with ELM to optimize network parameters. Using a mathematical model of improved PSO-ELM to learn the parameters of the dynamic system and reproduce human motions. The experimental results prove the method has a better fitting precision, adaptability and generalization ability on imitation learning of humanoid robots. The average relative errors are as 0.0172 of human motion reproductions.

Key words: imitation learning, extreme learning machine(ELM), improved particle swarm optimization(MPSO), nonlinear dynamic system, coupling model

中图分类号:

TP242.6

李文华1;杨子凝1;王来贵2. 基于耦合算法的类人机器人模仿学习控制方法[J]. 中国机械工程.

LI Wenhua1;YANG Zining1;WANG Laigui2. Imitation Learning Control Method of Humanoid Robots Based on Coupling Algorithm[J]. China Mechanical Engineering.

参考文献

[1]徐莉，刘振方，王建丞，等.仿人机器人综述[J].黑龙江科学，2013，4(7)：57-61.
XU Li，LIU Zhenfang，WANG Jiancheng，et al.Review on Humanoid Robot[J]. Heilongjing Science，2013，4(7)：57-61.
[2]刘涛，于建均，阮晓钢.基于概率轨迹匹配的机器人模仿学习方法[J]. 计算机测量与控制，2015，23(11)：3713-3720.
LIU Tao，YU Jianjun，RUAN Xiaogang. Robot Imitation Learning Method Based on Trajectory Probability Matching[J]. Computer Measurement and Control，2015，23(11)：3713-3720.
[3]于建均，门玉森，阮晓钢，等.模仿学习在机器人仿生机制研究中的应用[J]. 北京工业大学学报，2016，42(2)：210-216.
YU Jianjun，MEN Yusen，RUAN Xiaogang，et al. Application of Imitation Learning in the Research of Bionic Mechanism of Robots[J]. Journal of Beijing University of Technology，2016，42(2)：210-216.
[4]KHANSARI-ZADEH S M，BILLARD A. Learning Stable Nonlinear Dynamical Systems with Gaussian Mixture Models [J]. IEEE Transactions and Robotics，2011，27(5)：943-957.
[5]JIANG Ping，LI Ziyu，CHEN Yangquan.Iterative Learning Neural Network Control for Robot Learning from Demonstration[J]. Control Theory and Applications，2004，21(3)：447-452.
[6]NEUMANN K，LEMME A，STEIL J J. Neural Learning of Stable Dynamical Systems Based on Data-driven Lyapunov Candidates[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems. Tokyo，2013：1216-1222.
[7]SICILIANO B，KHATIB O. Handbook of Robotics[M]. New York：Springer，2008.
[8]白彦伟.基于微分几何移动机械臂非线性控制的研究[D].哈尔滨：哈尔滨工业大学，2007.
BAI Yanwei. Nonlinear Control Research of Mobile Manipulators Based on Differential Geometry [D]. Harbin：Harbin Institute of Technology，2007.
[9]吴晓刚，王旭东，余腾伟，等. 基于粒子群优化的电磁离合器模糊控制研究[J]. 中国机械工程，2010，21(9)：1071-1077.
WU Xiaogang，WANG Xudong，YU Tengwei，et al. Research on Fuzzy Control of Electromagnetic Clutch Based on Particle Swarm Optimization[J]. China Mechanical Engineering，2010，21(9)：1071-1077.
[10]王辉，朱龙彪，朱天成，等.基于粒子群遗传算法的泊车系统路径规划研究[J].工程设计学报，2016，23(2)：195-200.
WANG Hui，ZHU Longbiao，ZHU Tiancheng，et al. Research on Path Planning of Parking System Based on PSO-Genetic Hybrid Algorithm[J]. Chinese Journal of Engineering Design，2016，23(2)：195-200.
[11]HUANG Guangbin，ZHOU Hongming，DING Xiaojian. Extreme Learning Machine for Regression and Multiclass Classification[J]. IEEE Transactions Systems，Man，and Cybernetics，2012，42(2)：513-528.
[12]FENG Guorui，HUANG Guangbin，LIN Qingping，et al. Error Minimized Extreme Learning Machine with Growth of Hidden Nodes and Incremental Learning[J]. Transactions on Neural Networks，2009，20(8)：1352-1357.
[13]ZHANG Yannan，DING Shifei，XU Xinzheng，et al. An Algorithm Research for Prediction of Extreme Learning Machine Based on Rough Sets [J]. Journal of Computers, 2013，8(5)：1335-1342.
[14]何群，李磊，江国乾，等. 基于PCA和多变量极限学习机的轴承剩余寿命预测[J]. 中国机械工程，2014，25(7)：984-989.
HE Qun，LI Lei，JIANG Guoqian，et al. Residual Life Predictions for Bearings Based on PCA and MELM[J]. China Mechanical Engineering，2014，25(7)：984-989.