An Adaptive Fractional-order Admittance Model for a Lower Limb Rehabilitation Robot

Huanfeng Peng,Jie Zhou,Rong Song

DOI: https://doi.org/10.1109/icarm54641.2022.9959589

2022-01-01

Abstract:As a compliance strategy of physical human-robot interaction, the admittance control has been widely used in rehabilitation robots. The output of the admittance model based on a mass-damping-stiffness system is relative to the reference position, and this second-order model with stiffness is well suited for the early and middle stages of rehabilitation. In order to encourage the active participation of patients in the later stage of rehabilitation, removing the stiffness term is a solution to promote active compliance. However, for the multi-joint cooperative motion of the lower limb, setting the stiffness to zero directly may lead the patients to unsafe behavior. In this paper, we propose a fractional-order admittance model for a lower limb rehabilitation robot, of which the differential order can be adaptively adjusted based on the human-robot interaction torques. The effectiveness of the fractional-order admittance model is verified by simulation and experiments, and the results illustrate that the proposed model achieves the transitional adjustment between admittance model with and without stiffness.