In view of the drawbacks of the traditional machine learning algorithm,an improved algorithm based on deep learning for diagnosing arrhythmia signal and individual recognition of ECG signal is proposed.By adding an improved CBAM module and combining one-dimensional convolutional neural network with bidirectional LSTM,a new model L-CNN is proposed to diagnose and recognize arrhythmia ECG signals.The experimental results show that the recognition accuracy is 1.23% higher than that of the existing research.Compared with the traditional recognition technology,the recognition technology based on ECG signal has broad application prospect.

Complex Fault Diagnosis of Convolutional Neural Network Based on Short-Time Fourier Transform [J]. Journal of Shenyang University of Chemical Technology, 2023, 37(1): 68-73.