Research On Sleep Information Based On Piezoelectric Ceramic Smart Mattress

With the rapid development of modern society,people are increasingly satisfied with the material,and the pace of life is becoming faster.People are facing greater life pressure,which leads to more common sleep problems.People hope that they can monitor their sleep status in the home environment.Intelligent sleep has become a research hotspot for a while.At present,the sleep monitoring system at home and abroad is mainly wearable,with poor sleep comfort at night and low accuracy.In this thesis,an unconstrained intelligent mattress sleep monitoring system based on piezoelectric ceramic sensor is proposed.The weak sleep signal is transformed into electrical signal by piezoelectric effect,and then the physiological characteristics in the signal are extracted by using the algorithm.At the same time,the data is extracted by wireless module According to upload to the user terminal,users can observe their sleep status on the mobile phone.The system is suitable for home use,but also has a high accuracy.According to the original data collected by the intelligent mattress,this thesis mainly makes the following research:(1)The optimal channel selection scheme is proposed.The optimal channel is selected from 18 channels for signal processing.For the optimal channel,the DC signal is removed by de-mean method,and then it is smoothed by five points and three times smooth filtering to eliminate the burr noise and peak phenomenon of the signal.Using singular spectrum analysis algorithm to extract physiological features,that is,to construct Hankel matrix for the preprocessed signal,and then carry out singular value decomposition to obtain the singular value distribution map.Put forward the principle of singular value screening,select the appropriate singular value to reconstruct the respiratory signal and heartbeat signal under the quiet sleep.Finally,the feasibility of the intelligent mattress system and the high accuracy of the singular spectrum analysis algorithm are verified by comparing the experimental results with the standard values.(2)This thesis introduces sleep disordered breathing.Aiming at the common problems of sleep disordered breathing,including snoring and apnea,it proposes to use permutation entropy algorithm to detect them,calculate permutation entropy value in different sleep states,and get permutation entropy value range in sleep disordered breathing states through a large number of experimental results,so as to achieve the purpose of detecting abnormal breathing.Because the periodicity of respiratory signal is broken when sleep breathing is abnormal,and the heartbeat signal and snoring signal are overlapped together,an improved integrated empirical mode decomposition algorithm with better signal decomposition effect is selected to process such signals.Finally,the respiratory signal and heartbeat signal are successfully extracted to ensure the integrity of the function of the intelligent mattress system.