Answers are grounded in this paper and cite their sources.
Occupational active cable‐driven shoulder exoskeletons have many advantages compared to the traditional ones, but strong nonlinearity and complexity in modeling prevent the accurate torque transmission through Bowden cable, which is not conducive to its wide application. Besides, there is an urgent need to develop a more intelligent strategy for human‐robot interaction. In this paper, to overcome the dependency on the Bowden cable model and achieve the assist‐as‐needed (AAN) function during the application of the shoulder exoskeleton, a force‐velocity evaluation‐based model‐free sliding mode control (FVE‐MFSMC) is proposed. The controller contains a force‐velocity evaluation‐based fuzzy admittance control (FVEFAC) strategy as the outer loop controller to determine the needed joint angle of the exoskeleton and a model‐free super‐twisting global terminal sliding mode control (MF‐STGTC) as the inner loop controller to realize accurate motion control. In the outer FVEFAC, an admittance‐based reference angle is generated to achieve the needed assistance level. Both the interaction force and the angular velocity of the arm are considered to be the evaluation elements of the subject's intention, and then fuzzy logic is applied to adjust the stiffness parameter based on the evaluated subject's intention. In the inner MF‐STGTC, the ultra‐local model is used to approximate the original complex system, which avoids the accurate modeling of the Bowden cable. Time‐delay estimation is adopted to compensate for the lumped uncertainties, and a global fast terminal sliding mode control algorithm with super‐twisting design is used to stabilize the system while minimizing the chattering problem in traditional terminal sliding mode control. Co‐simulation experiments through MATLAB/Simulink and SolidWorks are carried out to demonstrate the performance of the proposed MF‐STGTC and FVE‐MFSMC, where the motion control can achieve the reduction percentage in root mean square error (RMSE) of tracking errors from 34% to 89% under fixed and time‐varying Bowden cable disturbance.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: