Dynamic Stiffness Estimation of Human Body Based on sEMG

Rehabilitation robots can effectively assist stroke patients in motor function rehabilitation. Robots need to have movement flexibility to ensure comfortable and active interaction among stroke patients. As a mechanical arm with good flexibility, the stiffness adjustment characteristics of the human body have received widespread attention. However, static estimation methods are mostly used to estimate human body impedance, which can lead to inaccurate human body impedance information. To achieve accurate man-machine impedance transfer, this paper studies muscle contraction and proposes a framework for human dynamic stiffness identification based on sEMG and static stiffness. In this paper, experiments were conducted to estimate the dynamic stiffness of human upper limbs. The conclusion proves that the dynamic stiffness calculated by this framework is consistent with the changes in arm stiffness during actual human motion. The framework proposed in this paper can be widely used in control systems that require human-machine impedance transfer.