Premilitary Performance Evaluation of Model-mediated Telerehabilitation System with Task Stiffness Estimation

For the telerehabilitation robotic system, the stable and transparent remote therapist-patient interaction under the internet communication latency remains an open challenge. In this paper, we proposed a model-mediated telerehabilitation system to avoid the delayed force feedback on the therapist side. Moreover, a variable stiffness robot is employed in the proposed telerehabilitation system to improve the patient-robot interaction on the slave side. On the slave side, a performance-based variable stiffness interactive assistance control law was utilized and the task stiffness estimation method is designed for master-side interaction model updating. On the master side, the master subsystem was implemented by a haptic device with visual and force feedback. The overall telerehabilitation was realized by the TCP/IP internet communication method. From the premilitary experimental results, the therapist subject could conduct the therapy with the aid of visual and force feedback, and the patient subject could complete the training task with their active efforts.