Active disturbance rejection super-twisting integral sliding mode control of wrist rehabilitation robot driven by foldable air bag

Abstract

Two high-speed switching valves (HSVs) were employed as the core control elements for high-precision angle tracking control in a wrist rehabilitation robot driven by foldable airbags, characterized by low cost, simple structure, and fast response. However, the dead-zone nonlinearity of the HSVs, typical nonlinearities of the pneumatic system, model inaccuracies, and external load disturbances pose significant challenges for achieving precise angle tracking. To address these challenges, an active disturbance rejection super-twisting integral sliding mode control (ADR-STISMC) was established and its convergence was proven, integrating active disturbance rejection control (ADRC), an integral sliding surface-based super-twisting sliding mode controller, and a dead-zone inverse function. A sine wave and an optimal wrist motion curve were employed in trajectory tracking experiments and five participants were enrolled in the verification experiments. The experimental results demonstrated the effectiveness of the proposed control algorithm where RMSE and MAE of ADR-STISMC were only one-third of those of ADRC respectively. The control scheme for the wrist rehabilitation device can provide sufficient technical guidance for other similar rehabilitation or automated devices.