Tremor Suppression Strategy Using Internal‐Model‐Based Equivalent‐Input‐Disturbance Approach With Real‐Time Frequency Estimation

Abstract

Wearable exoskeletons have recently emerged as a promising noninvasive solution for mitigating pathological tremor, providing an alternative to pharmacological and surgical treatments. A major difficulty in tremor suppression lies in reducing involuntary oscillations while preserving the user's natural voluntary motion. To address this problem, this study proposes an internal‐model‐based equivalent‐input‐disturbance (IMEID) approach for suppressing periodic tremor. The method first employs a frequency estimation algorithm to estimate the fundamental tremor frequency. An internal model of the tremor is then embedded into the disturbance estimator to enhance steady‐state attenuation capability. Subsequently, a Lyapunov‐based stability condition and an optimization‐based observer‐gain synthesis method are developed to guarantee the uniform ultimate boundedness (UUB) of the cascaded closed‐loop system. Simulation studies verify the effectiveness of the proposed scheme and highlight its performance advantages over other approaches.