A Research Of Sleep Staging Algorithm Based On Support Vector Machine And The Implementation Of Application

It is a hot topic to correctly assess the quality of sleep and improve the quality of sleep.It is the main idea of sleep research to identify sleep quality and to evaluate the quality of sleep.All the time,EEG is the most frequently used signal to study the sleep stage,but the acquisition equipment of this signal is usually complex,it can disturb normal sleep,and is not suitable for daily family monitoring.Compared with the traditional EEG monitoring,the mattress with piezoelectric sensors can monitor sleep and hardly affect normal sleep.The piezoelectric sensing mattress is a BCG signal,including heart rate,including heart rate and other information.The purpose of this study is to establish a sleep stage algorithm based on the BCG signal,and apply it to sleep monitoring in the application of the realization of sleep monitoring.First,we preprocess the heartbeat interval sequence from the BCG signal,then use the mathematical statistics and autoregressive model to extract the characteristics of heart rate variability,and then use the principal component analysis(PC A)to reduce the feature dimension.Finally,the support vector machine is used to classify the classification results,and the classification results are obtained.The results of classification are compared with the results of expert phased experiments.The results show that the accuracy of the algorithm can reach 70.3%.At the same time,we compare the staging results of the algorithm with the staging results of the support vector machine model determined by grid search method and the sleep stage based on the hidden Markov model,and find that the algorithm has a higher rate of accuracy.In this paper,the mobile sleep monitoring application is designed and implemented.In the application,the sleep staging algorithm designed in this paper is integrated to staging the user's sleep data in real time,and the real time sleep information is displayed for the users through the mobile phone application,and the functions of sleep record display,sleep quality assessment and sleep knowledge display are also provided.In this paper,a support vector machine is used for the first time to carry out sleep stages of heart rate variability extracted from BCG signals.Compared with other research,this algorithm has improved the feature extraction and processing of HRV and the optimization of SVM parameters,and has achieved a better stage effect.The implementation of mobile terminal sleep monitoring system provides a convenient way for family sleep monitoring,and has practical application value.