Study On The Method Of RQA To Recognize The Number Of The People In Bioradar

Bioradar is a kind of new concept radar, which synchronizes the radar and the technology of the bio-medical engineering. It can detect the vital signs (respiration and heartbeat) without contacting the subjects and it can penetrate through the non-golden medium of a certain thickness (brick wall and wasteland) to detect whether the beings exit. So it can be used in disaster succor and antiterrorism battle. In these occasions, we care not only whether the beings exit but also the number of beings.But at present, there is not an effective method to recognize the number of the people. Traditional signal identification methods usually pick-up the statistic character of the time region or the frequency region to identify the mode. Also someone picked-up the coefficients of the AR model to do it through building the time series parameter model of the vital signs. But these methods regard the vital signs as the calm signal or the calm signal in each subsection. And these methods can not achieve the satisfying effect. It has been proved in the former studies that the human respiration signal which we detected by the UWB bioradar is a kind of vital signs which are narrowband and non-calm.So these method mentioned above are not adaptive to analyse the signal detected by our UWB bioradar.At present,many methods(time domain cumulation,wavelet arithmetic et al)only do the simulated study in the background of white noise inland and the study of detecting the moving object and the many objects in many antennae have been done overseas. There is no correlative reports about the single antenna,many resting objects. But the single antenna,many resting objects is the basis to recognize the number of the people.RQA method is a kind of nonlinear analysis method. We can received the more information and conclusions using the RQA method to analyze the signal which can not received using other methods because the life system changed in many time scales and it is very hard to obtain the long-time stable time series, but the RQA method does not care the size and stability of the time series. We try to use the RQA method to analyse the data detected by the single antenna UWB bioradar in order to recognize the number of tbe people in this thesis.This thesis focuses on the following three aspects:1. Accomplish the system of data acquisition using the UWB bioradar in order to get the experiment data. And we got a lot of experiment data in different conditions.2. Using the RQA method to analyse the experiment data which we got. And we got an elementary conclusion that the PR can be used to recognize the number of the people. Through analysing the graph of PR distribution,we find that the PR is higher where someone exists than where nobody exists ,the PR is smaller as it depart this position and approximately to be a maximum. So we can estimate the number and the position of the people according to the number and the position of the maximum.3. Adopt the method of wavelet filtering combined with the RQA to improve the SNR. After the wavelet filtering, the graph of PR distribution is more perfect and the peak value is more outstanding in the position of the maximum. This is propitious to estimate the position more correctly.Innovations in this thesis are listed below:1. The study adopts the nonlinear method (RQA) to analyse the echo signal of the bioradar for the first time. And we get the elementary conclusion that it can be used to recognize the number of the people.2. Adopt the method of linear method combined with the nonlinear method (wavelet filtering combined with the RQA) to improve the SNR. And this is propitious to identify the position information correctly.