基于表面肌电的肘关节运动角度预测

表面肌电信号(sEMG)由于能够反映用户的动作意图而被广泛应用在上肢康复治疗系统中.针对目前上肢康复机器人的手臂关节运动控制不灵活的问题,提出一种基于sEMG的肘关节运动角度预测方法.为解决单一的时域特征提取方法存在的时间效率高而稳定性不足的问题,从时域和频域分别提取特征值,采用BP人工神经网络建立表面肌电信号与肘关节角度的映射模型,实现肘关节角度的预测.实验结果表明,该模型的预测结果与真实角度值有高度的一致性,有助于提高上肢康复机械臂的灵活性.

Prediction of elbow joint movement angle based onsurface electromyography

In the development of upper limb rehabilitation treatment system,surface electromyography (sEMG),with the capacity of reflecting users'' action intention,is widely used.Aiming at the problem that the arm joints of the upper limb rehabilitation robot are not flexible,this paper proposes a sEMG based elbow joint motion angle prediction method.To solve the problem that the stability of time domain feature extraction methodis insufficient,we use electromyography signal sampling analyzer and three-dimensional inertial position tracker acquire sEMG and the rotation information of the joint movement,extracting featuresfrom the time domain and frequency domain respectively,utilizing BP artificial neural network to establish the mapping relationship between sEMG and elbow angle.The experimental results show that the predictive results of this model are highly consistent with the true angle values,which is helpful to improve the flexibility of the upper limb rehabilitation manipulator.