基于深度学习和心电信号的运动疲劳识别方法

A Method for Identifying Exercise Fatigue Based on Deep Learning and ECG Signals

适当的体育运动有利于身体健康,但是大多数人在运动过程中,盲目地进行高强度的体育锻炼,很容易造成身体的损伤甚至危及生命。因此,针对这一问题,本研究提出了一种基于深度学习和心电信号的运动疲劳识别方法。首先采集运动中人体的心电信号,并使用连续小波变换提取信号的特征。然后将提取的心电特征融合并转换为二维图像数据集,再使用具有注意力机制的VGG神经网络对二维图像数据集进行训练和识别。最后构建了基于VGG-Attention Mechanism的运动疲劳检测模型。结果表明,所提方法构建的运动疲劳检测模型具有较高的诊断精度,平均准确率为98.03%,对可穿戴设备的开发具有重要意义。

Appropriate physical exercise is beneficial to physical health,but most people blindly carry out high-intensity physical exercise during exercise,which can easily cause physical damage or even endanger life.Therefore,in response to this problem,an exercise fatigue recognition method based on deep learning and ECG signals was proposed in this study.Firstly,the ECG signal of the human body in motion was collected,and the characteristics of the signal were extracted by continuous wavelet transform.Then the extracted ECG features were fused and converted into a two-dimensional image dataset.The VGG neural network with attention mechanism was used to train and recognize the two-dimensional image dataset.Finally,a motion fatigue detection model based on VGG-Attention Mechanism was constructed.The results showed that the motion fatigue detection model constructed by the proposed method has high diagnostic accuracy,with an average accuracy rate of 98.03%,which is of great significance for the development of wearable devices.