Study On Surface Emg Analysis Methods Under Rehabilitation Exercise

With the development of the neurorehabilitation, the analysis of bio-signal and biofeedback have been introduced to the rehabilitation exercise. Considering the features that the convenient, non-invasive surface electromyographic(s EMG) with small disturbance noise can well reflect the motion intention and muscle fatigue, the s EMG analysis has been widely used. And as the hot spot for the research of s EMG signal, the effective feature index of s EMG is necessary. A novel approaches were proposed to extract the feature for motion intention and muscle fatigue based on the ensemble empirical mode decomposition(EEMD), information entropy and intermuscular coherence. The main work of this paper is as follows:Firstly, the acquisition system of s EMG was designed, and the s EMG signal was acquired accordingly. In addition, the preprocessing method was proposed to remove the artifacts in s EMG.Secondly, based on the generation and characteristics of s EMG, the methodologies were described. Concurrent, the merit and defect of these methods were compared based on the test data from biceps and triceps under different angles.Thirdly, EEMD energy entropy was proposed to extract features and identify patterns for s EMG from biceps and triceps under different angles. In order to verify the effectivity of above method, the features extracted by EEMD energy entropy and other methods were input to the support vector machine(SVM), respectively. By the comparison of the recognition rate, the results reveal that the above method is of higher sensitivity and less error to obtain effective features of s EMG than other methods.Finally, the comprehensive assessment model was built to evaluate the inherent feature of muscle fatigue by analyzing the s EMG signal from multi-aspects. a novel method EEMD and Hilbert transform, called EEMD-HHT, was proposed to extract EMG-based average instantaneous frequency from each muscle. At the same time, the coherence is introduced to quantitatively describe the coupling relationship among muscles depending on calculating remarkable coherent area by collecting surface EMG signals. On this basis, the inherent relationship between average instantaneous frequency of EMG and the coupling among muscles was researched based on the analysis of test data. This provides a better approach to assess the muscle fatigue.