基于SEA的上肢康复机器人按需辅助控制策略

Assist-as-Needed Control Strategy for Upper Limb Rehabilitation Robots Based on Serial Elastic Actuators (SEA)

摘要: 随着老龄化程度加深和因神经损伤等原因导致的上肢功能障碍者越来越多，对上肢康复机器人的研究也越来越深入，但目前针对上肢康复机器人临床使用中的安全性与稳定性的研究仍较少。本文为串联弹性驱动器(SEA)驱动的上肢康复机器人开发了一种“按需辅助”(AAN)控制方案。基于奇异摄动法，将SEA驱动的机器人模型分为慢时间尺度动力学和快时间尺度动力学，并分别设计导数控制项与按需辅助控制项，以实现误差跟踪和目标阻抗调制，实现满意的AAN特性和患者积极参与康复训练的目的。仿真结果表明，在速度域中建立基于模型的并发学习自适应阻抗控制器能够在保证跟踪精度的同时实现有效的按需辅助康复训练。

Abstract: With the deepening of aging and more and more people with upper limb dysfunction caused by nerve injury and other reasons, the research on upper limb rehabilitation robots is more and more in-depth, however, there are few studies on the safety and stability of the upper limb rehabilitation robot. In this paper, an assist-as-needed (AAN) control strategy is developed for an upper limb re-habilitation robot driven by series elastic actuators (SEA). Based on the singular perturbation method, the SEA-driven robot model is divided into slow-time-scale dynamics and fast-time-scale dynamics, to achieve error tracking and target impedance modulation, to achieve satisfactory AAN characteristics and patients actively participate in rehabilitation training purposes. The simulation results show that the model-based concurrent learning adaptive impedance controller in the veloc-ity domain can achieve effective AAN rehabilitation training while ensuring tracking accuracy.