Analysis of geometrically constrained natural human movements and its application to robot control

本研究涉及的数学建模的人一样休息休息达到运动。两个模型，确定性最小努力模型和概率最小方差模型，正在开发和比较。结果发现，在运动后的时间的限制情况下，这些模型给出了等效的预测下，假设的植物稳定。该模型进行了验证和测试下的实验，使用可编程的触觉接口实现各种完整的约束。非完整约束也在本研究的框架内解决。在这里，在滚动约束的存在下，达到运动正在调查中。建立了球形滚动机器人的数学模型，建立了其动态可实现的条件。提出了两种运动规划算法。设计了一种球形滚动机器人，并进行了实验验证.

The research deals with mathematical modeling of human-like rest-to-rest reaching movements. Two models, the deterministic minimum effort model and the probabilistic minimum variance model, are under development and comparison. It was found that in the limiting case of post-movement time these models gives equivalent predictions under the assumption of the plant stability. The models were verified and tested under experiments using programmable haptic interface implementing a variety of holonomic constraints. Non-holonomic constraints were also addressed in the framework of this research. Here, reaching movements in the presence of rolling constraint were under investigation. A mathematical model of a spherical rolling robot was developed and the condition of dynamic realizability was established. Two motion planning algorithms have been proposed. A spherical rolling robot was designed and tested under experiment.