双足机器人步态规划研究
|
摘要
双足机器人因其灵活的步行能力和类人的行为而具有广泛的应用前景,其步态的稳定性已成为研究热点。步态规划是机器人实现稳定步行的基础和实现类人动作的关键。本文在在收集、整理和分析相关文献资料的基础上,以双足机器人为研究对象,借助RoboCup3D仿真平台,对双足机器人的步态规划进行研究。本文主要工作如下：
1.采用三次样条插值算法规划步态。根据关键时刻主要关节的位姿,基于三次样条插值算法,在满足步行环境,几何约束和物理结构的条件下,在径向和侧向平面中,分别规划出踝关节、髋关节在整个步行运动周期内的轨迹。再利用几何关系通过双平面的合成计算得到膝关节、脚掌等双足机器人下肢主要部位的运动轨迹。
2.利用粒子群算法对步态参数进行优化,得到稳定裕度大的平滑步态。首先分析了步态参数对步态稳定性的影响,然后根据双足机器人步行过程中单、双足支撑期的特点,分别构造了稳定裕度评估函数,并综合步行特征得到了步态优化的目标函数。最后采用粒子群算法优化步态参数,得到稳定裕度大的步态。
3.在MATLAB环境中和RoboCup3D仿真平台上进行实验。在MATLAB环境中规划得到期望的双足机器人各关节步行运动轨迹和稳定裕度大的ZMP轨迹,能够较好满足双足机器人稳定性原则。然后,在RoboCup3D仿真平台中,将已规划好的步态应用到仿真Nao机器人,通过统计仿真Nao机器人步行过程中的相关数据,验证三次样条插值规划方法的有效性和粒子群步态优化算法的优越性。
最后进行了总结,说明了本文的主要研究成果,同时对课题中需要改进之处提出了展望。
Biped robot has great application prospect because of its flexible walking ability and humanoid action, the stability of its gait has become a hot issue. Gait planning is the basics for the robot to walk stably and the key part to implement humanoid motion. In this dissertation, with RoboCup3D simulation platform, the research on walking motion planning of biped robot has been done based on the collection, collation and analysis of relevant literatures. The primary contents are as follows:
1. Using the cubic spline interpolation algorithm to plan the gait. In the sagittal and lateral plane, according to the main joint postures of critical moments, the walking patterns of ankle and hip joint have been planned in the walking cycle based on the cubic spline interpolation algorithm. And the planning gaits must meet all the conditions such as walking environment, geometric constrains and robot's structure. Finally, the trajectories of knee, foot and other parts of the biped robot are obtained by calculating with the geometric relationship between the two planes.
2. Using the particle swarm optimization algorithm to optimize parameters of gait and generate a smooth gait with large stability margin. Firstly, the dissertation analyzes the effection of gait parameters on gait stability. Secondly, this assement function of stability margin is constrcuted according to the characteristics of walking biped robot in single-support and double-support phase, and the target evaluation function used for evaluating the gait stability is obtained in the process of synthesizing respectively and completely on walking features. Finally, optimizing the gait parameters with particle swarm optimization algorithm and obtaining perfect gait with large stability margin which achieve the purpose of gait optimization.
3. Do experiments in the MATLAB environment and RoboCup3D simulation platform. Planning the desired walking motion trajectories of biped robot and the ZMP trajectory with large stability margin in the MATLAB, which can meet the principle of stability of biped robot. Then, the planned gait has been used on the biped robot Nao on the simulation platform in RoboCup3D. With statistical simulation data in the Nao robot walking process, it can verify the effectiveness of cubic spline interpolation algorithm and the advantages of PSO algorithm in planning biped gait.
At the end of this dissertation, the main research is summarized. It makes out the main research results, and also the problems which need to further research.
1.采用三次样条插值算法规划步态。根据关键时刻主要关节的位姿,基于三次样条插值算法,在满足步行环境,几何约束和物理结构的条件下,在径向和侧向平面中,分别规划出踝关节、髋关节在整个步行运动周期内的轨迹。再利用几何关系通过双平面的合成计算得到膝关节、脚掌等双足机器人下肢主要部位的运动轨迹。
2.利用粒子群算法对步态参数进行优化,得到稳定裕度大的平滑步态。首先分析了步态参数对步态稳定性的影响,然后根据双足机器人步行过程中单、双足支撑期的特点,分别构造了稳定裕度评估函数,并综合步行特征得到了步态优化的目标函数。最后采用粒子群算法优化步态参数,得到稳定裕度大的步态。
3.在MATLAB环境中和RoboCup3D仿真平台上进行实验。在MATLAB环境中规划得到期望的双足机器人各关节步行运动轨迹和稳定裕度大的ZMP轨迹,能够较好满足双足机器人稳定性原则。然后,在RoboCup3D仿真平台中,将已规划好的步态应用到仿真Nao机器人,通过统计仿真Nao机器人步行过程中的相关数据,验证三次样条插值规划方法的有效性和粒子群步态优化算法的优越性。
最后进行了总结,说明了本文的主要研究成果,同时对课题中需要改进之处提出了展望。
Biped robot has great application prospect because of its flexible walking ability and humanoid action, the stability of its gait has become a hot issue. Gait planning is the basics for the robot to walk stably and the key part to implement humanoid motion. In this dissertation, with RoboCup3D simulation platform, the research on walking motion planning of biped robot has been done based on the collection, collation and analysis of relevant literatures. The primary contents are as follows:
1. Using the cubic spline interpolation algorithm to plan the gait. In the sagittal and lateral plane, according to the main joint postures of critical moments, the walking patterns of ankle and hip joint have been planned in the walking cycle based on the cubic spline interpolation algorithm. And the planning gaits must meet all the conditions such as walking environment, geometric constrains and robot's structure. Finally, the trajectories of knee, foot and other parts of the biped robot are obtained by calculating with the geometric relationship between the two planes.
2. Using the particle swarm optimization algorithm to optimize parameters of gait and generate a smooth gait with large stability margin. Firstly, the dissertation analyzes the effection of gait parameters on gait stability. Secondly, this assement function of stability margin is constrcuted according to the characteristics of walking biped robot in single-support and double-support phase, and the target evaluation function used for evaluating the gait stability is obtained in the process of synthesizing respectively and completely on walking features. Finally, optimizing the gait parameters with particle swarm optimization algorithm and obtaining perfect gait with large stability margin which achieve the purpose of gait optimization.
3. Do experiments in the MATLAB environment and RoboCup3D simulation platform. Planning the desired walking motion trajectories of biped robot and the ZMP trajectory with large stability margin in the MATLAB, which can meet the principle of stability of biped robot. Then, the planned gait has been used on the biped robot Nao on the simulation platform in RoboCup3D. With statistical simulation data in the Nao robot walking process, it can verify the effectiveness of cubic spline interpolation algorithm and the advantages of PSO algorithm in planning biped gait.
At the end of this dissertation, the main research is summarized. It makes out the main research results, and also the problems which need to further research.
引文
[1]吴广玉,姜复兴.机器人工程导论[M].哈尔滨：哈尔滨工业大学出版社,1988.
[2]棍田秀司著,管贻生译.仿人机器人[M].北京：清华大学出版社,2007.Kajita Shuuji. Humanoid Robots [M]. Beijing:Tsinghua University Press,2007.
[3]殷际英,何广平.关节型机器人[M].北京：化学工业出版社,2003.
[4]柳洪义,宋伟刚.机器人技术基础[M].北京：冶金工业出版社,2002.
[5]SaeedB.Niku著,孙富春,朱纪洪,刘国栋,等译.机器人学导论一分析、系统及应用[M].北京：电子工业出版社,2004.
[6]傅京逊,[美]R.C冈萨雷斯,C.s.G李.机器入学[M].北京：科学出版社,1989.
[7]蔡自兴.机器人学[M].北京：清华大学出版社,2000.
[8]郑嫦娥,钱桦.仿人机器人国内外研究动态[J].机床与液压,2006,(3)：1-4.
[9]徐建峰.双足机器人动态步行研究及仿真[D].哈尔滨：哈尔滨工业大学,2001.
[10]马宏绪.两足步行机器人动态步行研究[D].长沙：国防科技大学,1995.
[11]马培荪,M伍科布拉托维奇著.步行机器人和动力型假肢[M].北京：科学出版社.1983.
[12]Hanafiah Yussof, Masahiro Ohka, Mitsuhiro Yamano, et al. Biped Locomotion Strategy in Humanoid Robot Navigation:A Case of Speed-up Walk [C]. The 2007 IEEE/ASME International Conference on Advanced Intelligent Mechatronics. 2007:1-7.
[13]汤卿.仿人机器人设计及步行控制方法[D].杭州：浙江大学,2009.
[14]胡凌云,孙增圻.双足机器人步态控制研究方法综述[J].计算机研究与发展2005,42(5)：728-733.
[15]Seungsu Kim, ChangHwan Kim, Bumjae You, et al. Stable Whole-body Motion Generation for Humanoid Robots to Imitate Human Motions [C]. The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems.2009: 2518-2524.
[16]Kazuo H, Masato H, Haikawa Y, et al. The Development of Honda Humanoid Robot [C]. Proceedings of the 1998 IEEE International Conference on Robotics and Automation.1998:1321-1326.
[17]俞志伟.双足机器人拟人步态规划与稳定性研究[D].哈尔滨：哈尔滨工程大学,2008.
[18]Chenglong Fu, Ken Chen. Gait Synthesis and Sensory Control of Stair Climbing for a Humanoid Robot [J].2008,5(55):2111-2120.
[19]Qiang Huang, Kazuhito Yokoi, Shuuji Kajita, et al. Planning Walking Patterns for a Biped Robot [J]. IEEE Transactions on Robotics and Automation.2001, 3(17):280-289.
[20]S. Kajita, F. Kanehiro, K. Kaneko, et al. Biped Walking Pattern Generation by a Simple Three-dimensional Inverted Pendulum Model [J]. Advanced Robotics, 2003,17(2):131-147.
[21]H. Hirukawa, S. Kajita, F. Kanehiro, et al. The Human-size Humanoid Robot That Can Walk, Lie Down and Get Up [J]. International Journal of Robotics Research,2005,24(9):755-769.
[22]李建,陈卫东,王丽军,等.不平整地面上双足机器人的步态控制[J].东南大学学报,2009,(39)：129-133.
[23]缭克华.Robocup3D足球机器人体系结构与基本技能的研究与实现[D].厦门：厦门大学,2008.
[23]M. Vukobratovic, B. Borovac, D. Surla, D. Stokic. Biped Locomotion:Dynamics, Stability, Control and Application.1975.
[25]A. Takanishi, M. Ishida, Y. Yamazaki, et al. The Realization of Dynamic Walking by Biped Robot WL-10RD [C]. IEEE International Conference on Autonomous Robot.1985:459-467.
[26]伊强,陈恳,刘莉等.考虑综合步行约束的仿人机器人参数化3D步态规划方法[J].机器人,2009,4(31)：342-350.
[27]Nguyen Thanh Phuong, Dae Won Kim, Hak Kyeong Kim, et al. An Optimal Control Method for Biped Robot with Stable Walking Gait [C].2008 8th IEEE-RAS International Conference on Humanoid Robots.2008:211-218.
[28]柯显信.仿人形机器人双足动态步行研究[D].上海：上海大学,2005.
[29]夏泽洋,陈恳,熊琼等.仿人机器人运动规划研究进展[J].高技术通讯,2007,10(17)：1092-1099.
[30]王剑.仿人机器人在线运动规划方法研究[D].长沙：国防科技大学,2008.
[31]Dusko M. Katic, Miomir K. Vukobratovic. Survey of Connectionist Control Algorithms for Humanoid Robots [J]. EUROCON,2005,22(24):310-313.
[32]柯显信,龚振邦,吴家麒.基于遗传算法的双足机器人上楼梯的步态规划[J].应用科学学报,2002,(04)：341-343.
[33]于彦朝.双足步行机器人步态规划研究[D].哈尔滨：哈尔滨工程大学,2007.
[34]魏航信,刘明治,赵丽琴.仿人机器人跑步运动的仿真[J].系统仿真学报,2007,19(2)：396-399.
[35]http://en.wikipedia.org/wiki/Zero_Moment_Point.
[36]王剑.仿人机器人在线运动规划方法研究[D].长沙：国防科技大学,2008.
[37]LEE B J, Stonier D, KIM Y D, et al. Modifiable Walking Pattern Generation Using Real-time ZMP Manipula- tion for Humanoid Robots [C]. Proceedings of the 2007 IEEE/RSJ International Conference on Intelligent Robots and Systems. 2007:4221-4226.
[38]禹奇才,张亚芳,刘峰.工程力学(理论力学部分)[M].广州：华南理工大学出版社,2005：226-227.
[39]杨敏.拟人机器人的步态规划仿真研究[D].南京：南京工业大学,2006.
[40]王建文.仿人机器人运动学和动力学分析[D].长沙：国防科技大学,2003.
[41]许小勇,钟太勇.三次样条插值函数的构造与Matlab实现[J].自动测量与控制,2006,25(11)：76-78.
[42]R.de Boer, J. Kok, and F.Groen. UvA Trilearn 2001 Team Description [J]. In Robocup-2001:Robot Soccer World Cup V. Springer Verlag, Berlin,2002.
[43]Hanafiah Yussof, Masahiro Ohka, Mitsuhiro Yamano, et al. Biped Locomotion Strategy in Humanoid Robot Navigation:A Case of Speed-up Walk. The 2007 IEEE/ASME International Conference on Advanced Intelli- gent Mechatronics. 2007,1-7.
[44]Seungsu Kim, ChangHwan Kim, Bumjae You. Stable Whole-body Motion Generation for Humanoid Robots to Imitate Human Motions [C]. The 2009 IEEE/RSJ International Conference on Intelligent Robots and Syste- ms.2009, 2518-2524.
[45]曾建潮,介婧,崔志华.微粒群算法[M].北京：科学出版社,2004.
[46]http://www.robocup.org/.
[47]胡崔娜.Robocup3D仿真比赛关键技术的研究[D].江西：江西理工大学,2007.
[48]Jelle R.Kok,Nikos Vlassis.UvA Trilearn 2005 team description.InProceedings CD RoboCup 2005,Springer-Verlag,Osaka,Japan,July 2005
[49]Paul Carpenter, Patrick Riley, Manuela Veloso, et al. Fcportugalteam description [R]. In RoboCup-02:Robot Soccer World Cup Ⅵ.Springer-Verlag,2002.
[50]Lufs Paulo Reis, Nuno Lau and Eugnio Costa liveir. Situation Based Strategicpositioning for Coordinating a Team of Homogeneous Agents[EB/ OL]. http://www.jeeta.pt/robocup/archive,htm,2002-5-18/2002-10-24.
[51]Philipp Vorst, Alexander Ferrein, Gerhard Lakemeyer. AllemaniACs3D Team Description [R]. In Proceedings CD RoboCup 2005,2005.
[52]钟碧良.机器人足球系统的研究与实现[D].广州：广东工业大学,2003.
[53]肖明尧.足球机器人仿真系统研究[D].吉林：吉林大学,2006.
[54]SimSpark User's Manual.http://www.robocup.org.
[55]李实,徐旭明,叶榛等.国际机器人足球仿真比赛及其相关技术[J].机器人,2000,2(5)：420-426.
[56]http://www.aldebaran-robotics.com.
[57]中国自动化学会机器人竞赛工作委员会,中国机器人大赛官方网站：http://www.rcccaa.org.
[58]赵业锦.受限视觉下RoboCup3D中机器人的定位[J].电子测试,2010,(4)：81-84.
[2]棍田秀司著,管贻生译.仿人机器人[M].北京：清华大学出版社,2007.Kajita Shuuji. Humanoid Robots [M]. Beijing:Tsinghua University Press,2007.
[3]殷际英,何广平.关节型机器人[M].北京：化学工业出版社,2003.
[4]柳洪义,宋伟刚.机器人技术基础[M].北京：冶金工业出版社,2002.
[5]SaeedB.Niku著,孙富春,朱纪洪,刘国栋,等译.机器人学导论一分析、系统及应用[M].北京：电子工业出版社,2004.
[6]傅京逊,[美]R.C冈萨雷斯,C.s.G李.机器入学[M].北京：科学出版社,1989.
[7]蔡自兴.机器人学[M].北京：清华大学出版社,2000.
[8]郑嫦娥,钱桦.仿人机器人国内外研究动态[J].机床与液压,2006,(3)：1-4.
[9]徐建峰.双足机器人动态步行研究及仿真[D].哈尔滨：哈尔滨工业大学,2001.
[10]马宏绪.两足步行机器人动态步行研究[D].长沙：国防科技大学,1995.
[11]马培荪,M伍科布拉托维奇著.步行机器人和动力型假肢[M].北京：科学出版社.1983.
[12]Hanafiah Yussof, Masahiro Ohka, Mitsuhiro Yamano, et al. Biped Locomotion Strategy in Humanoid Robot Navigation:A Case of Speed-up Walk [C]. The 2007 IEEE/ASME International Conference on Advanced Intelligent Mechatronics. 2007:1-7.
[13]汤卿.仿人机器人设计及步行控制方法[D].杭州：浙江大学,2009.
[14]胡凌云,孙增圻.双足机器人步态控制研究方法综述[J].计算机研究与发展2005,42(5)：728-733.
[15]Seungsu Kim, ChangHwan Kim, Bumjae You, et al. Stable Whole-body Motion Generation for Humanoid Robots to Imitate Human Motions [C]. The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems.2009: 2518-2524.
[16]Kazuo H, Masato H, Haikawa Y, et al. The Development of Honda Humanoid Robot [C]. Proceedings of the 1998 IEEE International Conference on Robotics and Automation.1998:1321-1326.
[17]俞志伟.双足机器人拟人步态规划与稳定性研究[D].哈尔滨：哈尔滨工程大学,2008.
[18]Chenglong Fu, Ken Chen. Gait Synthesis and Sensory Control of Stair Climbing for a Humanoid Robot [J].2008,5(55):2111-2120.
[19]Qiang Huang, Kazuhito Yokoi, Shuuji Kajita, et al. Planning Walking Patterns for a Biped Robot [J]. IEEE Transactions on Robotics and Automation.2001, 3(17):280-289.
[20]S. Kajita, F. Kanehiro, K. Kaneko, et al. Biped Walking Pattern Generation by a Simple Three-dimensional Inverted Pendulum Model [J]. Advanced Robotics, 2003,17(2):131-147.
[21]H. Hirukawa, S. Kajita, F. Kanehiro, et al. The Human-size Humanoid Robot That Can Walk, Lie Down and Get Up [J]. International Journal of Robotics Research,2005,24(9):755-769.
[22]李建,陈卫东,王丽军,等.不平整地面上双足机器人的步态控制[J].东南大学学报,2009,(39)：129-133.
[23]缭克华.Robocup3D足球机器人体系结构与基本技能的研究与实现[D].厦门：厦门大学,2008.
[23]M. Vukobratovic, B. Borovac, D. Surla, D. Stokic. Biped Locomotion:Dynamics, Stability, Control and Application.1975.
[25]A. Takanishi, M. Ishida, Y. Yamazaki, et al. The Realization of Dynamic Walking by Biped Robot WL-10RD [C]. IEEE International Conference on Autonomous Robot.1985:459-467.
[26]伊强,陈恳,刘莉等.考虑综合步行约束的仿人机器人参数化3D步态规划方法[J].机器人,2009,4(31)：342-350.
[27]Nguyen Thanh Phuong, Dae Won Kim, Hak Kyeong Kim, et al. An Optimal Control Method for Biped Robot with Stable Walking Gait [C].2008 8th IEEE-RAS International Conference on Humanoid Robots.2008:211-218.
[28]柯显信.仿人形机器人双足动态步行研究[D].上海：上海大学,2005.
[29]夏泽洋,陈恳,熊琼等.仿人机器人运动规划研究进展[J].高技术通讯,2007,10(17)：1092-1099.
[30]王剑.仿人机器人在线运动规划方法研究[D].长沙：国防科技大学,2008.
[31]Dusko M. Katic, Miomir K. Vukobratovic. Survey of Connectionist Control Algorithms for Humanoid Robots [J]. EUROCON,2005,22(24):310-313.
[32]柯显信,龚振邦,吴家麒.基于遗传算法的双足机器人上楼梯的步态规划[J].应用科学学报,2002,(04)：341-343.
[33]于彦朝.双足步行机器人步态规划研究[D].哈尔滨：哈尔滨工程大学,2007.
[34]魏航信,刘明治,赵丽琴.仿人机器人跑步运动的仿真[J].系统仿真学报,2007,19(2)：396-399.
[35]http://en.wikipedia.org/wiki/Zero_Moment_Point.
[36]王剑.仿人机器人在线运动规划方法研究[D].长沙：国防科技大学,2008.
[37]LEE B J, Stonier D, KIM Y D, et al. Modifiable Walking Pattern Generation Using Real-time ZMP Manipula- tion for Humanoid Robots [C]. Proceedings of the 2007 IEEE/RSJ International Conference on Intelligent Robots and Systems. 2007:4221-4226.
[38]禹奇才,张亚芳,刘峰.工程力学(理论力学部分)[M].广州：华南理工大学出版社,2005：226-227.
[39]杨敏.拟人机器人的步态规划仿真研究[D].南京：南京工业大学,2006.
[40]王建文.仿人机器人运动学和动力学分析[D].长沙：国防科技大学,2003.
[41]许小勇,钟太勇.三次样条插值函数的构造与Matlab实现[J].自动测量与控制,2006,25(11)：76-78.
[42]R.de Boer, J. Kok, and F.Groen. UvA Trilearn 2001 Team Description [J]. In Robocup-2001:Robot Soccer World Cup V. Springer Verlag, Berlin,2002.
[43]Hanafiah Yussof, Masahiro Ohka, Mitsuhiro Yamano, et al. Biped Locomotion Strategy in Humanoid Robot Navigation:A Case of Speed-up Walk. The 2007 IEEE/ASME International Conference on Advanced Intelli- gent Mechatronics. 2007,1-7.
[44]Seungsu Kim, ChangHwan Kim, Bumjae You. Stable Whole-body Motion Generation for Humanoid Robots to Imitate Human Motions [C]. The 2009 IEEE/RSJ International Conference on Intelligent Robots and Syste- ms.2009, 2518-2524.
[45]曾建潮,介婧,崔志华.微粒群算法[M].北京：科学出版社,2004.
[46]http://www.robocup.org/.
[47]胡崔娜.Robocup3D仿真比赛关键技术的研究[D].江西：江西理工大学,2007.
[48]Jelle R.Kok,Nikos Vlassis.UvA Trilearn 2005 team description.InProceedings CD RoboCup 2005,Springer-Verlag,Osaka,Japan,July 2005
[49]Paul Carpenter, Patrick Riley, Manuela Veloso, et al. Fcportugalteam description [R]. In RoboCup-02:Robot Soccer World Cup Ⅵ.Springer-Verlag,2002.
[50]Lufs Paulo Reis, Nuno Lau and Eugnio Costa liveir. Situation Based Strategicpositioning for Coordinating a Team of Homogeneous Agents[EB/ OL]. http://www.jeeta.pt/robocup/archive,htm,2002-5-18/2002-10-24.
[51]Philipp Vorst, Alexander Ferrein, Gerhard Lakemeyer. AllemaniACs3D Team Description [R]. In Proceedings CD RoboCup 2005,2005.
[52]钟碧良.机器人足球系统的研究与实现[D].广州：广东工业大学,2003.
[53]肖明尧.足球机器人仿真系统研究[D].吉林：吉林大学,2006.
[54]SimSpark User's Manual.http://www.robocup.org.
[55]李实,徐旭明,叶榛等.国际机器人足球仿真比赛及其相关技术[J].机器人,2000,2(5)：420-426.
[56]http://www.aldebaran-robotics.com.
[57]中国自动化学会机器人竞赛工作委员会,中国机器人大赛官方网站：http://www.rcccaa.org.
[58]赵业锦.受限视觉下RoboCup3D中机器人的定位[J].电子测试,2010,(4)：81-84.