Improved Research And Hardware Implementation Of Ambulatory Blood Pressure Algorithm Based On Pulse Wave

Nowadays,unhealthy lifestyle leads to the deterioration of people’s physical fitness.The prevalence of hypertension is increasing year by year,which may induce serious cardiovascular and cerebrovascular diseases,and pose a huge threat to the life and health of patients.The research and development of continuous non-invasive blood pressure monitoring methods and wearable devices based on pulse waves have good development prospects and practical application values.This paper proposes to use the cross-correlation algorithm to define the pulse wave transit time,which can effectively reduce the influence of noise and other objective conditions on the results and solve the problem of insufficient pulse waveform when collecting data.Based on the principle of measuring blood pressure based on pulse wave transit time,several common pulse wave characteristic parameters are extracted.According to the experimental results,a characteristic parameter was redefined,and a multiple linear regression model was established using the stepwise regression method to overcome the problem of weak correlation between diastolic blood pressure and conduction time.It is proposed to use Elman neural network to further improve the calculation model.Considering the actual application requirements,the hardware circuit design is optimized.The article mainly elaborates on basic theory,signal acquisition,algorithm improvement,hardware design,system testing,etc.Firstly,starting from the research background and significance of this topic,introduce the formation principle and potential connection of blood pressure and pulse wave,analyze the shortcomings of several common measurement methods,and propose the basic principles and advantages of blood pressure measurement based on pulse wave.The adjustment mechanism analyzes the factors that affect blood pressure.Secondly,build a data acquisition system based on the STM32 development board and Pulse Sensor sensor module,select appropriate measurement points and use photoplethysmography to obtain the pulse wave signal,and estimate the time delay of the two waveforms based on the cross-correlation function,and write the lower computer Data acquisition program.Develop a reasonable experimental program,select multiple volunteers and collect a large number of blood pressure and pulse sample data,establish a multiple linear regression equation,and use the stepwise regression method to screen parameters,and remove some parameters that have little influence on the results.The final calculation model is established based on the Elman neural network.And through random inspection,it is proved that the measurement accuracy of the model is in line with expectations.After comparison,it is found that the improved measurement error is smaller,which is basically within 5mm Hg specified by the AAMI standard.Finally,considering the miniaturization and power consumption of the device,the hardware circuit and the upper computer software based on MATLAB are redesigned.The function,schematic design and related components selection of signal acquisition module,microprocessor,power supply module,serial communication module and wireless communication module are introduced in detail.