External force/velocity control for an autonomous rehabilitation robot

8th TSME-International Conference on Mechanical Engineering

Saekow, Peerayuth^1 ; Neranon, Paramin^1 ; Smithmaitrie, Pruittikorn^1

Prince of Songkla University, Faculty of Engineering, Songkla, Hat Yai

90110, Thailand^1

关键词: Multi-axis force/torque sensors;    Omnidirectional mobile robot;    Proportional plus integrals;    Real time operating system;    Rehabilitation activities;    Rehabilitation robotics;    Root mean square errors;    Upper-limb rehabilitation;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/297/1/012051/pdf
DOI  :  10.1088/1757-899X/297/1/012051

来源: IOP

【 摘 要 】

Stroke is a primary cause of death and the leading cause of permanent disability in adults. There are many stroke survivors, who live with a variety of levels of disability and always need rehabilitation activities on daily basis. Several studies have reported that usage of rehabilitation robotic devices shows the better improvement outcomes in upper-limb stroke patients than the conventional therapy-nurses or therapists actively help patients with exercise-based rehabilitation. This research focuses on the development of an autonomous robotic trainer designed to guide a stroke patient through an upper-limb rehabilitation task. The robotic device was designed and developed to automate the reaching exercise as mentioned. The designed robotic system is made up of a four-wheel omni-directional mobile robot, an ATI Gamma multi-axis force/torque sensor used to measure contact force and a microcontroller real-time operating system. Proportional plus Integral control was adapted to control the overall performance and stability of the autonomous assistive robot. External force control was successfully implemented to establish the behavioral control strategy for the robot force and velocity control scheme. In summary, the experimental results indicated satisfactorily stable performance of the robot force and velocity control can be considered acceptable. The gain tuning for proportional integral (PI) velocity control algorithms was suitably estimated using the Ziegler-Nichols method in which the optimized proportional and integral gains are 0.45 and 0.11, respectively. Additionally, the PI external force control gains were experimentally tuned using the trial and error method based on a set of experiments which allow a human participant moves the robot along the constrained circular path whilst attempting to minimize the radial force. The performance was analyzed based on the root mean square error (E-RMS) of the radial forces, in which the lower the variation in radial forces, the better the performance of the system. The outstanding performance of the tests as specified by the E-RMS of the radial force was observed with proportional and integral gains of Kp= 0.7 and Ki= 0.75, respectively.