Compared to single arm robot system, dual arm robot has the ability of performing human-like dexterity and cooperation. Dual arm cooperative operation has attracted more and more attention in industrial applications, ...Compared to single arm robot system, dual arm robot has the ability of performing human-like dexterity and cooperation. Dual arm cooperative operation has attracted more and more attention in industrial applications, such as in assembly of complex parts, manufacturing tasks and handling objects. A unified dynamic control method, which is divided into three modes, namely, independent mode, dependent mode, and half dependent mode, is proposed for a redundant dual arm robot with focus on the movement and force of the desired task being operated. Attention is devoted to develop a unified formulation of the above three modes. In addition, a closed form of inverse kinematic solution instead of numerical integration approach is proposed with the aim to guarantee position accuracy. Different from traditional dynamic controllers, where the independent redundancy resolution is obtained based on particular velocity or acceleration levels, here the two dynamic controllers are improved by combining a closed form of inverse kinematic solution with velocity and acceleration levels. Furthermore, the theoretical results of the proposed control method are validated by simulations and experiments.展开更多
社会网络中社团核心的发现是目前研究界和产业界关注的热点问题。现有算法把社团处理为特定约束下的图后,将社团核心发现规约为紧凑子图的提取,但对于动态约束下的多图效率很低。为此,提出基于图密度的动态约束社团核心挖掘方法——CCDC...社会网络中社团核心的发现是目前研究界和产业界关注的热点问题。现有算法把社团处理为特定约束下的图后,将社团核心发现规约为紧凑子图的提取,但对于动态约束下的多图效率很低。为此,提出基于图密度的动态约束社团核心挖掘方法——CCDCD(community core mining with dynamic constrains based on graphdensity)。主要工作包括:(1)分析约束条件变化下,关于社团的图密度变化规律;(2)提出约束变化下,社团图密度的近似求解算法DCUE(dynamic calculation based on updated edges);(3)通过实验表明,与现有方法相比,对较大规模的社团图,新方法能获得更好解,降低时间消耗80%以上;验证了动态约束能发现更多有兴趣度的知识。展开更多
基金This paper is supported by the project of National Natural Science Foundation of China （No. 51405469, No. 51275505）, and science and technology support plan key projects of Jiangsu province （No. BE2013003）.
文摘Compared to single arm robot system, dual arm robot has the ability of performing human-like dexterity and cooperation. Dual arm cooperative operation has attracted more and more attention in industrial applications, such as in assembly of complex parts, manufacturing tasks and handling objects. A unified dynamic control method, which is divided into three modes, namely, independent mode, dependent mode, and half dependent mode, is proposed for a redundant dual arm robot with focus on the movement and force of the desired task being operated. Attention is devoted to develop a unified formulation of the above three modes. In addition, a closed form of inverse kinematic solution instead of numerical integration approach is proposed with the aim to guarantee position accuracy. Different from traditional dynamic controllers, where the independent redundancy resolution is obtained based on particular velocity or acceleration levels, here the two dynamic controllers are improved by combining a closed form of inverse kinematic solution with velocity and acceleration levels. Furthermore, the theoretical results of the proposed control method are validated by simulations and experiments.
文摘社会网络中社团核心的发现是目前研究界和产业界关注的热点问题。现有算法把社团处理为特定约束下的图后,将社团核心发现规约为紧凑子图的提取,但对于动态约束下的多图效率很低。为此,提出基于图密度的动态约束社团核心挖掘方法——CCDCD(community core mining with dynamic constrains based on graphdensity)。主要工作包括:(1)分析约束条件变化下,关于社团的图密度变化规律;(2)提出约束变化下,社团图密度的近似求解算法DCUE(dynamic calculation based on updated edges);(3)通过实验表明,与现有方法相比,对较大规模的社团图,新方法能获得更好解,降低时间消耗80%以上;验证了动态约束能发现更多有兴趣度的知识。
