Sensing and Control of a Multi-Joint Soft Wearable Robot for Upper-Limb Assistance and Rehabilitation
Sensing and Control of a Multi-Joint Soft Wearable Robot for Upper-Limb Assistance and Rehabilitation
复制标题
用于上肢辅助和康复的多关节软可穿戴机器人的传感和控制
DOI：
--
复制
发表时间：
2021
影响因子：
5.2
通讯作者：
C. Walsh
中科院分区：
文献类型：
--
作者：
Tommaso Proietti;Ciarán T. O’Neill;Cameron J. Hohimer;Kristin Nuckols;Megan E. Clarke;Yu Meng Zhou;David J. Lin;C. Walsh
在可穿戴机器人领域，人们对创造软可穿戴机器人越来越感兴趣，为身体残疾的人提供帮助和康复。与传统机器人相比，这些设备具有完全便携和轻便的潜力，这种灵活性可以增加使用时间，也可以在临床环境之外使用。在这封信中，我们提出了一种基于纺织品的多关节软可穿戴机器人来辅助上肢，特别是肩部抬高和肘部伸展。在开发便携式流体供应系统之前，我们利用非机载驱动系统进行动力和控制，磨损部件的重量不到半公斤。我们展示了这个机器人在断电时可以是机械透明的，不会限制用户执行与日常生活活动相关的动作。在躯干、上臂和前臂放置三个IMU，测量肩和肘部的运动学。我们发现，与光学运动捕捉系统相比，平均RMSE为$sim！5$度。我们实现了动态重力补偿(GC)和关节轨迹跟踪(JTT)控制器，可以根据IMU读数主动调节执行器压力。控制器的性能在一项针对8名健康个体的研究中进行了评估。使用GC控制器，受试者的肩部肌肉活动随着辅助程度的增加而减少，而对于JTT控制器，我们获得了较低的跟踪误差(平均$sim！6$RMSE)。未来的工作将评估机器人协助中风后康复活动的潜力。
In the field of wearable robotics, there has been increased interest in the creation of soft wearable robots to provide assistance and rehabilitation for those with physical impairments. Compared to traditional robots, these devices have the potential to be fully portable and lightweight, a flexibility that may allow for increased utilization time as well as enable use outside of a clinical environment. In this letter, we present a textile-based multi-joint soft wearable robot to assist the upper limb, in particular shoulder elevation and elbow extension. Before developing a portable fluidic supply system, we leverage an off-board actuation system for power and control, with the worn components weighting less than half kilogram. We showed that this robot can be mechanically transparent when powered off, not restricting users from performing movements associated with activities of daily living. Three IMUs were placed on the torso, upper arm and forearm to measure the shoulder and elbow kinematics. We found an average RMSE of $sim!5$ degrees when compared to an optical motion capture system. We implemented dynamic Gravity Compensation (GC) and Joint Trajectory Tracking (JTT) controllers that actively modulated actuator pressure in response to IMU readings. The controller performances were evaluated in a study with eight healthy individuals. Using the GC controller, subject shoulder muscle activity decreased with increasing magnitude of assistance and for the JTT controller, we obtained low tracking errors (mean $sim!6$ degrees RMSE). Future work will evaluate the potential of the robot to assist with activities in post-stroke rehabilitation.