基于1D-ECNN的短时心电信号疲劳分类

Classification of fatigue states based on short-term ECG signal and 1D-ECNN

提出一维双卷积神经网络(1D-ECNN),基于采集的心电信号检测操作员的疲劳状态。1D-ECNN包括4个卷积层、2个最大池化层、1个全连接层和1个softmax输出层。本研究仅使用较少的卷积核数量,这将减少模型参数的数量,降低模型的复杂程度,提高模型训练的速度,同时避免传统方法中复杂的特征提取过程或特征选择过程。将心电信号分成时间长度为1 s的样本,送入1D-ECNN,基于短时心电信号进行操作员疲劳状态分类。仿真结果表明,本文方法的平均分类准确率高达95.72%,能够实时准确地检测操作员的疲劳状态。此外,可以较好地消除个体差异性的影响。

One-dimensional double convolutional neural network(1D-ECNN)which includes 4 convolutional layers,2 maximum pooling layers,1 fully connected layer and 1 softmax output layer is proposed for detecting the fatigue state of the operator based on the acquired electrocardiogram(ECG)signals.Only a small number of convolutional cores are used in this study,which can reduce the number of model parameters,decrease the complexity of the model and increase the speed of model training,and meanwhile,it avoids the complicated feature extraction process or feature selection process in traditional methods.The acquired ECG signal is divided into samples with a time length of 1 s and then put into 1D-ECNN for classifying the fatigue state of the operator based on the short-term ECG signal.The simulation results show that the average classification accuracy of the proposed method is up to 95.72%,indicating that the proposed method can accurately detect the fatigue state of the operator in real time.In addition,it can better eliminate the effects of individual differences.