Redundant muscle coordination of a multi-DOF robot joint by online optimization

The Strand-Muscle Actuator (StMA) proposed by the authors has nonlinear elastic properties, and easily realizes joint angle/stiffness control, even for multi-DOF joints, by antagonistic actuator installation on the joint. The actuator is suitable for complex motions in spite of their simple mechanism, and extendable joint mechanism will be possible with them. To verify the effectiveness they have already been applied to some mechanisms such as legged robots, multi-finger hands, and 3-DOF joints. The authors have manufactured a StMA-based Robot Arm (StMA-RArm), a human-arm-like robot manipulator using StMAs. It is composed of the StMA-based Robotic Shoulder (StMA-RS), an elbow, a wrist, and a 4-flngered hand. The StMA-RS realizes 3-DOF motions with joint stiffness control, and having redundant muscles it is failure tolerant. For smooth and dexterous motions real-time cooperative muscle tension control is necessary. In this article the mechanism and cooperative control scheme of the StMA-RArm are presented. The Vibrant Particle Swarm Optimization is introduced to solve time-variant optimization problems. And it is then successfully applied to the online coordination of redundant muscles.

[1]  Azumi Ichikawa,et al.  Toward Springy Robot Walk Using Strand-Muscle Actuators , 2005 .

[2]  Zhiwei Luo,et al.  Sensory-motor control of a muscle redundant arm for reaching movements - convergence analysis and gravity compensation , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  N. A. Bernstein Dexterity and Its Development , 1996 .

[4]  Masakazu Suzuki,et al.  Evolutionary acquisition of complex behaviors through intelligent composite motion control , 2005, 2005 International Symposium on Computational Intelligence in Robotics and Automation.

[5]  Ming Zhang,et al.  A function description for the human upper limb pointing movements performance , 2001, 2001 Conference Proceedings of the 23rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[6]  Richard Quint van der Linde,et al.  Design, analysis, and control of a low power joint for walking robots, by phasic activation of McKibben muscles , 1999, IEEE Trans. Robotics Autom..

[7]  R. Dillmann,et al.  A method for learning complex and dexterous behaviors through knowledge array network , 2001, Proceedings 2001 IEEE/RSJ International Conference on Intelligent Robots and Systems. Expanding the Societal Role of Robotics in the the Next Millennium (Cat. No.01CH37180).