Controlling the walking speed in limit cycle walking
Controlling the walking speed in limit cycle walking
复制标题
DOI：
10.1177/0278364908095005
复制
发表时间：
2008-09-01
影响因子：
9.2
通讯作者：
Wisse, M.
中科院分区：
文献类型：
--
作者：
Hobbelen, D. G. E.;Wisse, M.
“极限环行走”是设计和控制两足步行机器人的一种相对较新的范式。它指出，实现稳定的周期性步态是可能的，而不是像大多数步行机器人传统上所做的那样，在每一个时刻局部稳定行走轨迹。众所周知的极限环步行器的例子是被动动态步行器，但最近也有许多驱动的极限环步行器。极限环步行者通常比其他现有的两足动物使用更少的能量，但到目前为止，它们还没有那么多功能。这篇文章关注的是多功能性的一个方面：行走速度。我们研究如何改变步行速度，哪种方式对能量有益，以及步行速度如何影响步行者处理干扰的能力(即抑制干扰)。研究使用了一个样机和一个仿真模型。这两名步行者的速度是通过改变三个参数来调整的：脚踝推出量、上半身俯仰和步长。这项研究得出了四个结论。(1)获得了在0.24~0.68m S(-1)之间的稳态速度(对于0.6m的腿长度)，失稳决定了下限，而驱动极限决定了上限。这一结果表明了极限环行走对多用途步行机械的适用性。(2)对于任何速度，通过身体前倾来驱动步态比使用脚踝推开要消耗更少的能量。(3)与传统步行机器人在速度和稳定性之间的明显折衷相反，在极限环步行中，我们发现增加步行速度会自动导致增强的干扰抑制，而不是如何实现的。(4)前馈驱动调节和步进速度反馈的结合表明，在执行负重和坡道行走等任务时，只需几步就可以改变步速并保持所需的速度。特别是，这第四个结论强调了极限环步行的概念对多功能两足步行机器的适用性。
"Limit Cycle Walking" is a relatively new paradigm for the design and control of two-legged walking robots. It states that achieving stable periodic gait is possible without locally stabilizing the walking trajectory at every instant in time, as is traditionally done in most walking robots. Well-known examples of Limit Cycle Walkers are the Passive Dynamic Walkers, but recently there are also many actuated Limit Cycle Walkers. Limit Cycle Walkers generally use less energy than other existing bipeds, but thus far they have not been as versatile. This paper focuses on one aspect of versatility: walking speed. We study how walking speed can be varied, which way is energetically beneficial and how walking speed affects a walker's ability to handle disturbances (that is, disturbance rejection). The study is performed using one prototype and one simulation model. The speed of these two walkers is adapted by changing three parameters: the amount of ankle push-off, upper body pitch and step length. The study has resulted in four conclusions. (1) Steady-state speeds between 0.24 and 0.68 m s(-1) (for a 0.6 m leg length) were obtained, with loss of stability determining the lower limit and actuation limits determining the upper limit. This result shows the applicability of Limit Cycle Walking for versatile walking machines. (2) For any speed, powering the gait by leaning the body forward costs less energy than using ankle push-off. (3) In contrast to the apparent tradeoff between speed and stability in traditional walking robots, in Limit Cycle Walking we find that increasing the walking speed, independent of how this is done, automatically results in an increasing disturbance rejection. (4) A combination of feedforward actuation adjustment and step-to-step feedback from walking speed shows that it is possible to change walking speed in only a few steps and maintain a desired speed when performing tasks such as carrying loads and walking on slopes. In particular, this fourth conclusion underlines the applicability of the concept of Limit Cycle Walking for versatile two-legged walking machines.