Research Of Human Heart Rate Monitoring System Based On Photoplethysmography

Recently,with the social development in a high-speed and life quality improved significantly,the importance of health is becoming more and more high,especially in under the trend of population aging,the wrist belt type heart rate monitoring equipment increasingly active in the public’s view,the current studies of this type of equipment has become the direction of the electronic intelligence that a big hit.As an important parameter of human body,heart rate can reflect the physiological indexes of various organ functions and circulatory system in real time.Accurate and real-time monitoring of heart rate can reflect human sickness,which is meaningful to people’s health.At present,the wristband heart rate monitoring equipment on the market generally has limitations,such as not easy to use,high power consumption,failed to solve the problem of low accuracy caused by movement pseudo-error.To solve these problems,this paper developed a heart rate monitoring system based on human Photoplethysmography(PPG),which was optimized from hardware filtering,software algorithm filtering,heart rate extraction algorithm and other aspects,so as to achieve the goal of improving the measurement accuracy and real-time detecting of heart rate in movement.This paper has done the following research:Aiming at the serious influence of Motion Artifact(MA)in PPG signal under the moving state,which makes it difficult to extract the heart rate value effectively from the collected signal,an algorithm based on the Normalized Least Mean Square(NLMS)error adaptive filtering combined with the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)was proposed to filter the motion pseudo-error in PPG signal.NLMS adaptive filtering is performed on preprocessed PPG signals to remove high frequency noise and baseline drift,and then the objects are broken down into several Intrinsic Mode functions(IMF)by empirical Mode decomposition algorithm of fully adaptive noise sets.Then,the threshold value was determined by calculating the Multiscale Permutation Entropy(MPE)of each mode,and the appropriate IMF component was selected to reconstruct PPG signal within the given range to remove noise such as motion interference and baseline drift,thus improving the accuracy of heart rate calculation of the system.In the design of heart rate monitoring system,the hardware and software parts were designed respectively.A hardware acquisition platform based on PPG signal was built,and a mobile APP was developed.The hardware design covers the main control chip and Bluetooth communication module,PPG signal acquisition module,I/V conversion module,signal filtering and amplification module and three axis accelerometer sensor circuit.Software design mainly includes PPG signal processing flow and heart rate extraction algorithm design.The main control chip STM32F103 is designed and improved to improve the switching mechanism of sleep mode and low power operation mode,which effectively reduces the system power consumption and improves the standby time.The local data management module is used to filter the collected PPG signal,and display the processed pulse wave and calculated heart rate value on the mobile APP.Aiming to verify reliability of the algorithm proposed in this paper ultimately,PPG signals from the wrist of 8 subjects in four states of sitting,normal walking(swinging arm),jogging and fast running were collected.The proposed method and the other two algorithms were used to filter PPG signals respectively,with signal-to-noise ratio and mean square error as evaluation indexes.Experimental results show that the proposed algorithm has higher SNR and smaller mean square error.In addition,the heart rate values in the four states were calculated by PPG signal after processing.Compared with the other two methods,the Average Absolute Error(AAE)of the proposed method was reduced by0.6%～1.4%.The Pearson Correlation Coefficient(PCC)between the real heart rate and the estimated heart rate increased by 0.3%,and the calculated heart rate value was more accurate,which proved that the human heart rate monitoring system developed in this paper has a higher accuracy of heart rate measurement,and meets the practical requirements.