Research On Intelligent Detection Method Of Human Blood Oxygen For Hip Measurement And Design Of Networked System

In today’s society,with the development of science and technology,people have higher and higher requirements for their own quality of life.With the increasing demand of human beings,the electronic equipment for monitoring human health has gained higher market potential,and people’s demand for it has also increased significantly.Blood oxygen saturation is one of the most important physiological health parameters.This index can reflect the oxygen content in human body.It is one of the important parameters to detect cardiopulmonary function,and can also be used as a reference index to prevent some diseases.This paper mainly studies the network detection system of blood oxygen extraction.Firstly,two pairs of sensors(red light(660nm)probe and infrared light(905nm))are respectively placed in two different parts of the human buttocks to obtain two pairs of PPG(photoplethysmography)signals,each pair of two channels(red light and infrared light).Firstly,the PPG signal is preprocessed,and a baseline drift noise removal algorithm based on morphological filtering is proposed to remove the noise of the original signal.Then,by estimating the correlation dimension of each pair of PPG signals,the correlation dimension of each pair of PPG signals can be obtained,and the correlation dimension can be used as the characteristic parameter of a PPG signal to evaluate the signal quality;The two PPG signals collected by the same wavelength photoelectric sensor at different parts of the same detector’s buttock are used for the optimal judgment.The extracted characteristic parameters are used to obtain the better two PPG signals through the relevant judgment criteria,and a pair of PPG signals after the high-quality judgment are used to calculate the subsequent blood oxygen.Secondly,the paper introduces the algorithm of extracting the peak and valley value of pulse wave.The traditional method has some defects more or less,or the peak and valley value extraction is not accurate or the peak and valley value extraction is incomplete.For the problem of extracting the peak and valley value feature point information,this paper improves the algorithm based on the extremum point,and the improved algorithm can adaptively search the peak and valley value information alternately,It lays a foundation for the prediction of blood oxygen saturation.Thirdly,the R value extracted from the processed PPG signal is used to train the model by adding the human attributes such as height,weight and BMI.The improved elm model is used to predict the blood oxygen saturation value.The cdessa-elm model based on chaos differential evolution bottle ascidian algorithm is proposed,which improves the pan Chinese ability of elm model and reduces the probability of over fitting Through the comparison with the other four models,it is found that the error rate of the same training set in this model is very low,and the effect in practical application is also better.At last,the realization method of the whole network system is described,and the hardware and software of the blood oxygen detection system and its network structure are introduced.Firstly,the realization method and background significance of the network are briefly introduced,and then the related hardware system is introduced,including the selection of blood oxygen sensor,the function of Wi Fi module and the acquisition and processing module;Finally,in the aspect of software,it mainly introduces the data communication,the realization of blood oxygen saturation algorithm and the design of software interface.