Study On The Respiratory Signal Monitoring System Based On PVDF Piezoelectric Thin Film Sensor

Some characteristic parameters in respiratory signals such as respiration amplitudes,respiration periods and so on are the significant means to determine sleeping qualification and diagnose breathing-related diseases.Traditional invasive monitoring tools are disadvantageous of being too specialized,complex operation,too difficult to realize real-time monitoring and too much physiological and psychological burden.For out-of-hospital patients,instead of complicated clinical testing devices,they are in great request to obtain a home-style,convenient technique to monitor respiratory signals.In recent years,an extremely popular polymer film-Polyvinylidene Fluoride(PVDF)piezoelectric film has the advantages of high piezoelectric coefficients,light weight,fine flexibility and so on,thus it is highly expected to measure physiological signals.To avoid the disturbance derived from body casual movements existed in traditional PVDF monitoring belts and sensing mattresses,a headset-based sensing device based on PVDF piezoelectric film sensor that is extremely suitable to be worn when people are in sleep is developed,then its corresponding monitoring system to detect respiratory signals is investigated.The main researches are listed as follows:Firstly,as the sensing element in this paper,the mechanical and electrical properties of PVDF piezoelectric film material were researched.The headset-based sensing device based on PVDF piezoelectric film was designed,after which the bending deformation process of the PVDF film as a cantilever was analyzed.The whole size of the PVDF film embedded in this sensing device is finally determined as 20 mm×10 mm×30 ?m.Besides,simple and reliable processes to package the PVDF sensor were demonstrated in detail.Secondly,as for conditioning circuits,charge amplifier and filter circuits were emphasized,relevant core elements and key parameters including operation amplifier,feedback resistance,feedback capacitor and cutoff frequency were determined by performance comparison.The designed and manufactured circuits are proved to possess satisfactory ability to transform,amplify and filter signals both via the Multisim software and experimental results of detection respiratory signals.Thirdly,the respiratory signal monitoring system was developed,in which the periodical bending deformation caused by airflow during inspiration and expiration activities could be detected accurately.This method can achieve respiratory signals with fewer noises,more smoothness and higher visibility compared with other traditional detecting manners.Sorts of respiratory patterns such as respiratory arrest,apnea,coughing,inadequate breathing as well as deep breathing,shallow breathing and eupnea were simulated to validate the dynamic responding performance of the system.The experimental results from time and frequency domain indicate the system can respond to various respiratory patterns timely and effectively.Finally,in order to realize detection of peaks and troughs,Wavelet Transform is adopted to cancel noises.Wavelet coefficients at high orders were abandoned directly to reduce baseline wander.Besides,a wavelet improved thresholding function was adopted to cancel disturbance,because the algorithm can relieve the natural defects such as inadequate amplitudes in soft thresholding and discontinuous points in hard thresholding.Time interval of two adjacent peaks(or troughs)and the adaptive voltage threshold of peaks(or troughs)were set to determine the peaks and troughs.In this way,characteristic parameters such as average periods,average amplitudes and frequency were obtained successfully and automatically.This establishes a good foundation for the further study.