Research On Heart Rate Detection Algorithm Of Photoplethysmography Signal In Sports Noise Environment

Heart rate monitoring is of great significance in medical diagnosis.Photoplethysmography(PPG)can be used to detect the change of heart rate in real time.The severe challenge in PPG signal denoising is to reduce the motion artifacts,which can lead to large errors in the predicted heart rate.At the same time,the amount of PPG signal data collected by the monitoring equipment is large,which will not only occupy a lot of memory of the monitoring equipment,but also spend a lot of time in the process of data transmission and signal processing.In order to improve the accuracy of estimating the heart rate value,this thesis will conduct research on the estimation of the motion artifact and the heart rate value in the PPG signal.The main work is as follows::1.Aiming at the problem that the pulse signal will eliminate the peak of the heart rate spectrum in the process of reducing the motion artifact,this thesis proposes an effective method to reduce the motion artifact.This method uses the three-axis acceleration signal as the reference signal.First,the three-axis gyroscope signal is used to correct the three-axis acceleration angle value;then the corrected three-axis accelerometer is analyzed through singular spectrum;finally,it is used as the reference signal of the cascaded FTRLS algorithm to reduce motion artifacts.Finally,the experimental results show that the filtering algorithm can not only reduce the motion artifacts superimposed on the pulse wave,but also retain the information of the double wave.2.In view of the large amount of pulse signals collected by monitoring equipment,which will waste a lot of time in data transmission and subsequent processing,the method of compressed sensing is adopted,and a method of calculating heart rate in the time domain is proposed.First,the method uses the training dictionary K-Singular Value Decomposition(K-SVD)to sparse the pulse signal,and uses the Orthogonal Matching Pursuit(OMP)algorithm to restore the original signal.Then,the main wave template is constructed to match the peak value of the compressed and reconstructed pulse signal,and the maximum index position of the main wave is obtained through the maximum correlation degree,so as to calculate the heart rate value in the time domain.finally,the average absolute error,Bland-Altman chart,and Peel are used The Xun correlation coefficient evaluates the error between the calculated heart rate value and the true value.the experimental results show that the error of the heart rate value calculated by this method is smaller.