Wrist tremor suppression based on repetitive control with multi-muscle electrical stimulation

Abstract

Tremor is a rhythmic, alternating swing motion caused by involuntary repetition of muscle contraction and relaxation. Although it does not endanger life, it will make the work and daily life of patients difficult. Functional electrical stimulation (FES) has been shown as a promising technique for tremor suppression. Wrist motion is produced by a group of muscles in a collective and coordinate way. However, existing FES-based design methods mostly aim at one pair of muscles associated with the wrist motion, thus limiting the performance of tremor suppression. Furthermore, the possible high level of stimulation required for a single muscle pair can also accelerate muscle fatigue of the patients. To address these problems, this paper uses multiple muscles FES to suppress tremor by fully considering the properties of wrist motion. This paper develops a wrist musculoskeletal model with Hammerstein structure, identifies its parameters, and proposes repetitive controllers based on frequency modified inverse algorithm to suppress tremor. Experimental results are presented to demonstrate its advantages over single muscle stimulation based tremor suppression.