一种基于肌电信号的踝关节动作预测方法的研究  被引量:3

Research on a method for prediction of ankle movement based on electromyography signals

在线阅读下载全文

作  者:王震[1] 张震[1] 姚松丽[2] 章亚男[1] 钱晋武[1] 

机构地区:[1]上海大学机电工程与自动化学院,上海200072 [2]上海工程技术大学机械工程学院,上海201620

出  处:《高技术通讯》2010年第11期1173-1177,共5页Chinese High Technology Letters

基  金:863计划(2006AA04Z224);国家自然科学基金(50975165);上海市自然科学基金(01ZR1411500);上海市教委科研创新项目(10YZ17)资助

摘  要:进行了使用肌电信号预测关节动作的研究,提出了一种基于肌电信号的预测踝关节动作的方法。首先,选取与踝关节动作相关的5块肌肉(胫骨前肌、腓肠内肌、腓肠外肌、腓骨长肌和比目鱼肌)以及踝关节角作为研究对象,采集这5块肌肉的肌电信号和踝关节角信号,并进行特征提取和归一化处理。然后,建立了一个四层前向神经网络模型,使用误差逆向传播(BP)算法进行训练。最后,神经网络预测输出值经过六层小波去噪处理。实验中,9名志愿者的踝关节在矢状面内做有规律的背屈和跖屈动作,采集踝关节角和上述5块肌肉表面肌电信号,然后用上述方法预测踝关节动作,用相关系数评价预测的效果。实验结果显示,所提出的方法可以准确预测踝关节动作。A new method for prediction of ankle movement based on electromyography signals was developed. First, surface electromyography signals of five muscles associated with the ankle movement and ankle joint angle signals were collected, and their features were extracted and the outputs were normalized. Then, a feed forward neural network model was built. The neural network was composed of four layers and the back-propagation training algorithm was used. Finally, the prediction output signal from the neural network was transformed by the six-layer wavelet to eliminate noise. In the experiment, nine subjects were asked to do regular dorsiflexion and plantarflexion ankle movement in the sagittal plane. The ankle joint angle data and the surface electromyography signals of the five muscles were recorded. Ankle movement was predicted by the method. At last, the quality of the neural network' s prediction of ankle movement was evaluated by cor- relation coefficients. It is shown that the proposed method can accurately forecast the movement of ankle joint.

关 键 词:肌电(EMG)信号 神经网络 BP算法 踝关节 小波去噪 

分 类 号:R318.0[医药卫生—生物医学工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象