Research On Heart Rate Extraction Method Based On Photoplethysmography Signal Under The Background Of Motion Noise

With the development of the times,the fast-developing economy has accelerated the pace of people's life.People's pursuit of life is no longer limited to the improvement of the level of clothing,food,housing and transportation.They are also paying more and more attention to their own health.More and more people are actively participating Sports,improve one's own physical fitness,relax body and mind is the most real portrayal.Heart rate is one of the most important physiological characteristics of the human body.Through heart rate monitoring,you can understand your physical condition.However,the heart rate will change with the change of exercise intensity.People can reasonably arrange their exercise intensity according to the change of heart rate,so as to avoid excessive exercise to cause damage to themselves.Compared with traditional heart rate detection methods,heart rate monitoring equipment designed based on PPG signals can make it easier for people to understand their own health status in real time.But heart rate monitoring devices based on PPG signals also have their own limitations.When in motion,the collected PPG signal is easily affected by motion interference noise,which reduces the accuracy of the finally obtained heart rate value.Therefore,this paper studies how to extract a more accurate heart rate value from the PPG signal collected by the existing motion interference noise.The work is as follows:(1)Usually,accurate and reliable heart rate extraction for PPG signals with motion interference noise is mainly divided into three stages: preprocessing,removal of motion interference noise,and heart rate estimation.Each stage requires careful analysis before implementation.Choosing the right filter is important for accurate heart rate extraction,and filter coefficients vary depending on the size of the filter.Based on the trade-off between accuracy and performance,this paper proposes to first use the SVD-based heart rate extraction method to remove the existing motion interference noise from the reference acceleration signal in the PPG signal,while retaining the main features of the PPG signal;Expectation-maximizing Kalman filter heart rate extraction method for parameter estimation.Validated by simulations and experiments on available datasets,the method used in this paper shows more desirable results than the fixed parameters.(2)For the error in the filtering state of each individual signal when using dual PPG sensors for signal acquisition,through the comparative analysis of the state vector fusion method and the measurement fusion method,it is proved that the measurement fusion method can reduce the acquisition of two sensors.The error brought by the process is better than that of the state vector fusion method,and then the improvement of the Kalman filter heart rate extraction method based on expectation maximization is realized.(3)Since the available real data is limited by many aspects,a tool for synthesizing PPG and acceleration signals is realized through the exploration of PPG signals and acceleration signals.The PPG signal and the acceleration signal are synthesized according to the preset different motion states and times,and the improved method of heart rate extraction based on the expectation-maximization Kalman filter proposed in this paper is compared with the method based on the linear model and the exponential model.It is verified that the method proposed in this paper can obtain accurate heart rate information.