利用健肢sEMG信号对康复机械腿进行映射控制
|
摘要
研究了利用健肢表面肌电信号(sEMG)信号控制康复机械腿的问题。首先采集健全人在沿直线自然行走状态下,下肢八块肌肉的表面肌电信号和髋、膝、踝三个关节在矢状面内的关节角度信号,并对信号进行预处理、特征提取和标准化等操作;然后,建立了一个BP神经网络模型进行训练,将训练得到的神经网络模型用于关节角度的预测;最后以处理后的肌电信号作为模型输入,模型输出值经过移动均值滤波得到最终的关节角度预测值。再将此预测值作为机械腿的控制信号,实现机械腿的实时控制。实验结果显示,下肢三关节实际测量角度和预测值平均均方差(RMSE)为5.8295,平均皮尔逊相关系数(γ)为0.9312,可以较好实现机械腿的实时控制。
The problem of using the limb surface electromyography(sEMG) signal to control the rehabilitation mechanical leg was studied.Firstly,the surface EMG signals of the eight muscles of the lower limbs and the joint angle signals of the hip,knee and ankle joints in the sagittal plane were collected under the natural walking along the straight line,and the signal is preprocessed,characterized and standardized.Wait for the operation.Then,a BP neural network model is established for training,and the trained neural network model is used for the prediction of joint angle.Finally,the processed EMG signal is used as the model input,and the model output value is filtered by moving average to obtain the final joint angle prediction value.Then use this predicted value as the control signal of the mechanical leg to realize real-time control of the mechanical leg.The experimental results show that the average mean square error(RMSE) of the measured angles and predicted values of the lower limbs is 5.8295,and the average Pearson correlation coefficient(γ) is 0.9312,which can achieve real-time control of the mechanical legs.
The problem of using the limb surface electromyography(sEMG) signal to control the rehabilitation mechanical leg was studied.Firstly,the surface EMG signals of the eight muscles of the lower limbs and the joint angle signals of the hip,knee and ankle joints in the sagittal plane were collected under the natural walking along the straight line,and the signal is preprocessed,characterized and standardized.Wait for the operation.Then,a BP neural network model is established for training,and the trained neural network model is used for the prediction of joint angle.Finally,the processed EMG signal is used as the model input,and the model output value is filtered by moving average to obtain the final joint angle prediction value.Then use this predicted value as the control signal of the mechanical leg to realize real-time control of the mechanical leg.The experimental results show that the average mean square error(RMSE) of the measured angles and predicted values of the lower limbs is 5.8295,and the average Pearson correlation coefficient(γ) is 0.9312,which can achieve real-time control of the mechanical legs.
引文
[1]黄敬,梅元武,童萼塘.脑卒中后脑的可朔性与康复[J].中国康复,2004(1):50-52
[2]丁其川,熊安斌,赵新刚,等.基于表面肌电的运动意图识别方法研究及应用综述[J].自动化学报,2016,42(1):13-25
[3]Chu J U,Moon I,Lee Y J,et al.A supervised feature-projection-based real-time EMG pattern recognition for multifunction myoelectric hand control[J].IEEE/ASME Transactions on Mechatronics, 2007,12(3):282-290
[4]姚鹏飞,盛鑫军,郭伟超,等.基于表面肌电信号与近红外光谱技术联合解码的仿人假手控制系统[J].中国康复医学杂志,2016,31(5):547-551
[5]王震,张震,姚松丽,等.一种基于肌电信号的踝关节动作预测方法的研究[J].高技术通讯,2010,20(11):1173-1177
[6]姚松丽,章亚男,张震,等.利用选择性肌电信号控制踝关节神经运动康复装置[J].上海大学学报(自然科学版),2009,15(3)
[7]章亚男,景银平,沈林勇,等.下肢表面肌电信号的降维和映射分析[J].传感技术学报,2018,31(7):70-77
[2]丁其川,熊安斌,赵新刚,等.基于表面肌电的运动意图识别方法研究及应用综述[J].自动化学报,2016,42(1):13-25
[3]Chu J U,Moon I,Lee Y J,et al.A supervised feature-projection-based real-time EMG pattern recognition for multifunction myoelectric hand control[J].IEEE/ASME Transactions on Mechatronics, 2007,12(3):282-290
[4]姚鹏飞,盛鑫军,郭伟超,等.基于表面肌电信号与近红外光谱技术联合解码的仿人假手控制系统[J].中国康复医学杂志,2016,31(5):547-551
[5]王震,张震,姚松丽,等.一种基于肌电信号的踝关节动作预测方法的研究[J].高技术通讯,2010,20(11):1173-1177
[6]姚松丽,章亚男,张震,等.利用选择性肌电信号控制踝关节神经运动康复装置[J].上海大学学报(自然科学版),2009,15(3)
[7]章亚男,景银平,沈林勇,等.下肢表面肌电信号的降维和映射分析[J].传感技术学报,2018,31(7):70-77