Supernumerary robotic limbs are emerging to augment human function. Unlike exoskeletons, these robots provide additional kinematic structures to the user that enable novel human-robot interactions. To assist walking, a supernumerary leg should be compliant to impacts, minimize efforts on users, move quickly when swinging and exert large assistive forces on the ground. Here, we study the potential of a supernumerary leg powered by delocalized magnetorheological clutches (MR leg) to assist walking with three different gaits. Simulations show that the MR leg's low actuation inertia reduces the impact impulse by a factor 4 compared to geared motors and that delocalizing the clutches reduces by half the inertial forces transmitted to the user during swing. An impedance controller receives a reference trajectory based on each ankle's position to move the MR leg in synchrony with the gait cycle. Experiments show that the MR leg can comfortably contact the ground and swing at 3.9 m/s for a 1.4 m/s walk. The MR leg tracks the ankle within 5% of the gait cycle for the leader-follower gait, alternately tracks both ankles for the double gait and contacts the ground in between each step for the three-legged gait. A theoretical upper limit suggests that the average transmitted power in a gait cycle could be 84 W for the leader-follower gait, which is 4 times higher than autonomous ankle exoskeletons.

中文翻译：

---

由磁流变驱动器驱动的辅助人类运动的多余机器人腿

多余的机器人肢体正在出现以增强人类功能。与外骨骼不同，这些机器人为用户提供了额外的运动学结构，实现了新颖的人机交互。为了辅助行走，多余的腿应该能够承受冲击，尽量减少对使用者的影响，摆动时移动迅速，并在地面上施加大的辅助力。在这里，我们研究了由离域磁流变离合器（MR 腿）驱动的多余腿的潜力，以帮助以三种不同的步态行走。模拟表明，与齿轮电机相比，MR 腿的低致动惯性将冲击脉冲降低了 4 倍，并且离域离合器可将摆动期间传递给用户的惯性力减少一半。阻抗控制器接收基于每个脚踝的参考轨迹 s 位置以与步态周期同步移动 MR 腿。实验表明，MR 腿可以舒适地接触地面并以 3.9 m/s 的速度摆动 1.4 m/s 的步行。MR 腿在领跑者-跟随步态的步态周期的 5% 内跟踪脚踝，交替跟踪双脚踝以实现双步态，并在每一步之间接触地面以实现三足步态。理论上限表明，领导者-跟随者步态的步态周期中的平均传输功率可能为 84 W，比自主脚踝外骨骼高 4 倍。交替跟踪双脚踝以进行双步态，并在三足步态的每一步之间接触地面。理论上限表明，领导者-跟随者步态的步态周期中的平均传输功率可能为 84 W，比自主脚踝外骨骼高 4 倍。交替跟踪双脚踝以进行双步态，并在三足步态的每一步之间接触地面。理论上限表明，领导者-跟随者步态的步态周期中的平均传输功率可能为 84 W，是自主踝关节外骨骼的 4 倍。