Suppr超能文献

模仿有蹄类动物的机器人脚可改善在松软湿滑地面上的运动能力。

Robotic feet modeled after ungulates improve locomotion on soft wet grounds.

机构信息

Centre for Biorobotics, School of IT, Department of Computer systems, Tallinn University of Technology (Taltech), Tallinn, Estonia.

出版信息

Bioinspir Biomim. 2024 Nov 4;19(6). doi: 10.1088/1748-3190/ad839c.

Abstract

Locomotion on soft yielding grounds is more complicated and energetically demanding than on hard ground. Wet soft ground (such as mud or snow) is a particularly difficult substance because it dissipates energy when stepping and resists extrusion of the foot. Sinkage in mud forces walkers to make higher steps, thus, to spend more energy. Yet wet yielding terrains are part of the habitat of numerous even-toed ungulates (large mammals with split hooves). We hypothesized that split hooves provide an advantage on wet grounds and investigated the behavior of moose legs on a test rig. We found that split hooves of a moose reduce suction force at extrusion but could not find conclusive evidence that the hoof reduces sinkage. We then continued by designing artificial feet equipped with split-hoof-inspired protuberances and testing them under different conditions. These bio-inspired feet demonstrate an anisotropic behavior enabling reduction of sinkage depth up to 46.3%, suction force by 47.6%, and energy cost of stepping on mud by up to 70.4%. Finally, we mounted these artificial feet on a Go1 quadruped robot moving in mud and observed 38.7% reduction of the mechanical cost of transport and 55.0% increase of speed. Those results help us understand the physics of mud locomotion of animals and design better robots moving on wet terrains. We did not find any disadvantages of the split-hooves-inspired design on hard ground, which suggests that redesigning the feet of quadruped robots improves their overall versatility and efficiency on natural terrains.

摘要

在柔软的地面上移动比在坚硬的地面上更复杂且能量需求更高。湿软的地面(如泥或雪)是一种特别困难的物质,因为它在踩踏时会消耗能量,并阻碍脚部的挤出。在泥中下沉迫使步行者迈更高的步伐,因此消耗更多的能量。然而,湿软的地形是许多偶蹄类动物(有分蹄的大型哺乳动物)的栖息地的一部分。我们假设分蹄在湿地上有优势,并在测试台上研究了驼鹿腿的行为。我们发现,驼鹿的分蹄在挤出时减少了吸力,但不能确定蹄子是否能减少下沉。然后,我们继续设计配备分蹄启发式突起的人工脚,并在不同条件下进行测试。这些仿生脚表现出各向异性的行为,可将下沉深度减少高达 46.3%,吸力减少高达 47.6%,在泥中踩踏的能量成本减少高达 70.4%。最后,我们将这些人工脚安装在 Go1 四足机器人上,在泥中移动,并观察到机械运输成本降低了 38.7%,速度提高了 55.0%。这些结果帮助我们了解动物在泥地中的运动物理学,并设计更好的在湿地上移动的机器人。我们没有发现分蹄启发式设计在坚硬地面上的任何缺点,这表明重新设计四足机器人的脚可以提高它们在自然地形上的整体通用性和效率。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验