Non-contact sleep breathing disorder event detection method and system based on cardiopulmonary activity information fusion

The invention discloses a non-contact sleep breathing disorder event detection method and system based on cardiopulmonary activity information fusion, and the method comprises the steps: carrying out the preprocessing of a radar echo signal, obtaining a breathing signal, estimating the higher harmonic frequency of a heart rate through an echo signal estimation algorithm, and carrying out the detection of a non-contact sleep breathing disorder event. A time-varying heart rate value is inverted in combination with the frequency interval of the normal heart rate; fusing the respiratory signal data and the heart rate data, discriminating and detecting sleep breathing disorder events by using a convolutional neural network, and finally determining sleep apnea ventilation indexes according to discriminating results. According to the method, the higher harmonic frequency of the heart rate is estimated on the basis of estimation algorithms of synchronous compression wavelet transform and an adaptive multi-trajectory