Design And Research Of Non-contact Respiration Detection And Analysis System Based On Microwave

Respiration is a vital physiological parameter.Human life cannot be maintained without breathing movement.Detection of breathing can timely and effectively obtain human health information and prevent the occurrence of diseases.This subject uses microwave non-contact method to detect human respiratory signals,which has many advantages over traditional methods,such as penetrating barriers such as clothes and bedding,avoiding direct contact between electronic sensors and human body,and being able to detect breathing noninvasively over a long distance.In this paper,a front-end system of microwave breathing detection is developed to detect human breathing.Empirical mode decomposition and cohesive hierarchical clustering algorithm are proposed to identify abnormal breathing conditions during long breathing.Two methods of apnea recognition are proposed,and a sleep apnea monitoring system based on three microwave radar channels is built.The main research contents of this paper include:(1)The reliability of microwave non-contact breathing detection was verified.A front-end system for detecting human respiration based on microwave radar is developed,which includes amplifier module,power supply driver module,alarm module and near infrared detection module.The traditional commercial nasal airflow equipment YH-600B and microwave detection front-end are used to collect human respiratory signals at the same time.The reliability of microwave detection of respiratory signals is verified by comparative analysis.(2)In addition to normal breathing,there are various abnonnal breathing states in the process of human breathing.Each abnormal breathing state occurs frequently for a long time and has different pathological inducements.It is necessary to identify and judge the breathing state of human body during long breathing process in order to achieve early diagnosis of diseases.In view of this,an empirical mode decomposition(EMD)algorithm is proposed to identify abnormal breathing frequency and extract abnormal state time,and the results of EMD,EEMD and CEEMD are compared and analyzed.A cohesive hierarchical clustering algorithm is proposed to recognize abnormal breathing amplitude and extract abnormal state time.On the basis of the recognition of the two algorithms,Kusmore breathing,shallow fast breathing,deep slow breathing and deep slow breathing are completed.Shallow and slow breathing was identified and abnormal time was extracted.(3)In order to identify and judge sleep apnea and to solve the influence of changes in sleeping posture and posture on judging apnea.In this paper,three algorithms,short-time zero-crossing rate method,short-time integration method and K-means clustering method,are proposed to recognize body motion signals.Based on the difference of amplitude of breathing signals collected by microwave channels on both sides of sleeping human body,body position signals are recognized.On the basis of the above research,two kinds of apnea recognition algorithms are proposed in this paper:windowed energy spectrum method and WA-EMD energy spectrum metlhod.More analysis.Sleep apnea monitoring system based on three microwave channels was built.Sleep tests of 10 groups of normal people showed that the system could recognize and judge sleep apnea.The research work of this subject realizes the non-contact detection of human breathing based on microwave,identifies various abnormal breathing states in the process of human long-term breathing,and prevents the harm caused by long-term breathing abnormalities to human body.This research also realizes the judgment recognition of sleep apnea and resolves the influence of sleep posture change on the judgment of sleep apnea,which can effectively diagnose sleep apnea to reduce the harm to people.