Edinburgh Napier University

  Intention tremor is an involuntary and rhythmic muscle contraction that occurs during purposeful limb movements. Functional electrical stimulation based repetitive control (RC) has proven to be an effective approach to reject periodic tremor disturbance with a constant frequency. However, the performance in suppressing tremor slightly decreases with variations in tremor frequency and inevitable musculoskeletal model uncertainties. Therefore, improving the robustness of the repetitive controller remains a challenge. This article presents a composite control strategy that combines active disturbance rejection control (ADRC) with high-order RC (HORC) to address the above issues. The stability of the proposed closed-loop ADRC based high-order RC system is analyzed. Simulation results show that the proposed control strategy can not only suppress tremor disturbances of varying frequency, but also handle the model parameter uncertainty. Furthermore, comparison to experimental outcomes of unimpaired subjects and tremor patients also demonstrate the effectiveness and robustness of the proposed method. The average tremor suppression performance achieved by the proposed ADRC-based HORC (ADRC–HORC) method is 87.76%, which represents an improvement of approximately 11% over the traditional RC method and 27% over the filter based method.