WO2023004688A1 - 一种基于多普勒雷达的非接触式呼吸监测方法 - Google Patents

一种基于多普勒雷达的非接触式呼吸监测方法 Download PDF

Info

Publication number
WO2023004688A1
WO2023004688A1 PCT/CN2021/109286 CN2021109286W WO2023004688A1 WO 2023004688 A1 WO2023004688 A1 WO 2023004688A1 CN 2021109286 W CN2021109286 W CN 2021109286W WO 2023004688 A1 WO2023004688 A1 WO 2023004688A1
Authority
WO
WIPO (PCT)
Prior art keywords
monitoring
frequency
respiratory
distance
chest
Prior art date
Application number
PCT/CN2021/109286
Other languages
English (en)
French (fr)
Inventor
韩毅
叶建平
王硕玉
Original Assignee
南京浙溧智能制造研究院有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京浙溧智能制造研究院有限公司 filed Critical 南京浙溧智能制造研究院有限公司
Priority to CN202180029015.9A priority Critical patent/CN115460980A/zh
Priority to PCT/CN2021/109286 priority patent/WO2023004688A1/zh
Publication of WO2023004688A1 publication Critical patent/WO2023004688A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
    • A61B5/0816—Measuring devices for examining respiratory frequency
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves 
    • A61B5/0507—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves  using microwaves or terahertz waves
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/1126—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb using a particular sensing technique
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/113—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb occurring during breathing
    • A61B5/1135—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb occurring during breathing by monitoring thoracic expansion
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/50—Systems of measurement based on relative movement of target
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications

Definitions

  • the invention belongs to the technical field of respiratory frequency monitoring, and in particular relates to a non-contact respiratory monitoring method based on Doppler radar.
  • non-contact physiological characteristic monitoring is in the ascendant, and various non-contact physiological characteristic monitoring methods are emerging. They are based on a variety of monitoring media, including infrared, acoustic and optical.
  • the non-contact vital sign monitoring technology based on the above media has more or less limitations.
  • infrared detectors will be affected by heat sources near the monitoring object;
  • acoustic detectors use air as the transmission medium, which is easily affected by air humidity and airflow, and cannot be used in materials that absorb sound waves such as cotton yarn;
  • optical detectors are susceptible to All kinds of light interference are difficult to play a role in the case of a lot of smoke, poor line of sight or the body of the monitoring object is blocked.
  • the non-contact vital signs monitoring technology based on Doppler radar is also widely known as bio-radar technology, which uses microwave as the detection medium, has strong anti-interference ability, strong penetrating power to non-metallic obstacles, and high integration , low power consumption and many other advantages.
  • bio-radar technology which uses microwave as the detection medium, has strong anti-interference ability, strong penetrating power to non-metallic obstacles, and high integration , low power consumption and many other advantages.
  • most of the existing Doppler radar-based physiological feature monitoring methods are relatively simple in algorithm, and often only directly extract the vital sign signal in the radar echo through the corresponding band-pass filter and other means, which makes the extracted signal Still contains a lot of noise and interference, resulting in a low-quality extracted signal.
  • the present invention proposes a new respiratory monitoring method based on Doppler radar. Through the innovative design of the radar receiving data processing algorithm, a better respiratory monitoring effect is achieved.
  • Preprocessing of the radar echo signal determination of the chest position of the monitored object, extraction of the chest movement information of the monitored object, extraction of the respiratory waveform of the monitored object, and estimation of the respiratory frequency.
  • the preprocessing of the radar echo signal mainly includes mixing the radar echo signal, and the output of the mixing is an intermediate frequency signal x out ,
  • the amplitude A of the intermediate frequency signal is constant, and the frequency and phase are proportional to the distance R between the reflector and the antenna, as shown in the following formula:
  • S is the rate of change of the frequency of the linear frequency-modulated continuous wave with time
  • c is the speed of light
  • ⁇ is the wavelength of the frequency-modulated continuous wave used, all of which are constants.
  • the FFT processing is performed on the intermediate frequency signal output by the frequency mixing, and the components of different frequencies in the frequency spectrum are obtained, which represent the radar echo signals reflected by objects at different distances.
  • the sequence of the data received by the nth antenna in the antenna array after the above preprocessing is x n , and the kth data in this sequence is recorded as x n,k , then at the current moment, the receiving
  • the reception vector at a certain distance received by the array can be expressed as:
  • N the number of receiving antennas in the receiving array.
  • the extraction of the movement information of the thorax of the monitoring object is based on the position parameters obtained by the above-mentioned algorithm for determining the position of the thorax of the monitoring object, specifically including the distance and direction angle of the thorax of the monitoring object relative to the antenna array to extract the movement information of a specific position.
  • the specific calculation process is: according to the distance parameter of the monitoring object relative to the antenna array and the characteristic that the frequency of the preprocessed intermediate frequency signal is proportional to the distance, select the corresponding intermediate frequency signal component from the preprocessed data center. From the phase information of the intermediate frequency signal according to the following formula Extract the motion information R at the distance of the thorax of the monitoring object,
  • the motion information at the angle of the chest of the monitoring object is further extracted from the motion information R received by the antenna array at the distance of the chest of the monitoring object.
  • the extraction of the respiratory waveform of the monitoring object adopts an improved complete ensemble empirical mode decomposition method, that is, an improved CEEMD algorithm.
  • u(t) and v(t) represent the upper envelope sequence and the lower envelope sequence of the current signal component respectively.
  • IMF component There are two criteria for judging the IMF component:
  • the endpoint effect in the EMD method is suppressed by using the envelope extremum continuation method.
  • the specific implementation method is: in the previous iterative process of decomposing a specific IMF component, continuation on the left side of the original signal sequence by 2 1 maximum value point and 2 minimum value points, and 2 maximum value points and 2 minimum value points are extended on the right;
  • ⁇ k-1 usually takes 0.1-0.3 times of the standard deviation of the (k-1)th residual component sequence.
  • the IMF component performs FFT processing on each IMF component to obtain its spectrum, and calculate the proportion of its energy in the respiratory frequency band (0.1Hz-0.5Hz) to the total energy. It is considered that when the proportion is higher than a certain threshold, the IMF component is the respiratory signal or part of the breathing signal, to be preserved. Finally, the sum of all the retained IMF components is considered as the respiratory waveform of the monitoring object.
  • Estimating the respiratory frequency is to obtain the above-mentioned extracted respiratory waveform through FFT, and obtain the corresponding frequency spectrum. Search to find the peak in the spectrum, and the respiratory rate can be estimated according to the following formula:
  • f is the breathing frequency obtained by this method, in bpm; f 0 is the sampling frequency, in Hz. k and N are the subscript and maximum subscript corresponding to the peak value in the spectrum, respectively.
  • the advantage of the present invention is that: the present invention proposes a non-contact respiratory monitoring algorithm based on Doppler radar.
  • the echo signals received by the radar antenna are preprocessed sequentially to extract the thoracic position, extract the thoracic motion information, extract Respiratory waveform information, respiratory frequency estimation and other steps, finally obtain the respiratory waveform and frequency parameters of the monitoring object.
  • the thorax of the monitored object in this method has higher positioning accuracy, which is the basis for accurate extraction of thoracic motion signals.
  • the improved CEEMD algorithm is used to separate the breathing and heartbeat signals. This algorithm has a good separation effect on body motion interference, background noise and vital sign signals. Through the improvement, it also effectively improves the problem of the endpoint effect in the original CEEMD algorithm. , while increasing the decomposition speed.
  • Figure 1 shows the effect of thoracic position determination.
  • Figure 2 shows the effect of breathing and heartbeat signals separated by the improved CEEMD algorithm.
  • Figure 3 is a comparison between the respiratory waveform extracted from the radar echo signal and the reference waveform.
  • the preprocessing of the radar echo signal mainly includes mixing the radar echo signal, and the output of the mixing is an intermediate frequency signal x out ,
  • the amplitude A of the intermediate frequency signal is constant, and the frequency and phase are proportional to the distance R between the reflector and the antenna, as shown in the following formula:
  • S is the rate of change of the frequency of the linear frequency-modulated continuous wave with time
  • c is the speed of light
  • ⁇ is the wavelength of the frequency-modulated continuous wave used, all of which are constants.
  • the FFT processing is performed on the intermediate frequency signal output by the frequency mixing, and the components of different frequencies in the frequency spectrum are obtained, which represent the radar echo signals reflected by objects at different distances.
  • the sequence of the data received by the nth antenna in the antenna array after the above preprocessing is x n , and the kth data in this sequence is recorded as x n,k , then at the current moment, the receiving
  • the reception vector at a certain distance received by the array can be expressed as:
  • N the number of receiving antennas in the receiving array.
  • the extraction of the movement information of the thorax of the monitoring object is based on the position parameters obtained by the above-mentioned algorithm for determining the position of the thorax of the monitoring object, specifically including the distance and direction angle of the thorax of the monitoring object relative to the antenna array to extract the movement information of a specific position.
  • the specific calculation process is: according to the distance parameter of the monitoring object relative to the antenna array and the characteristic that the frequency of the preprocessed intermediate frequency signal is proportional to the distance, select the corresponding intermediate frequency signal component from the preprocessed data center. From the phase information of the intermediate frequency signal according to the following formula Extract the motion information R at the distance of the thorax of the monitoring object,
  • the motion information at the angle of the chest of the monitoring object is further extracted from the motion information R received by the antenna array at the distance of the chest of the monitoring object.
  • the extraction of the respiratory waveform of the monitoring object adopts an improved complete ensemble empirical mode decomposition method, that is, an improved CEEMD algorithm.
  • u(t) and v(t) represent the upper envelope sequence and the lower envelope sequence of the current signal component respectively.
  • IMF component There are two criteria for judging the IMF component:
  • the endpoint effect in the EMD method is suppressed by using the envelope extremum continuation method.
  • the specific implementation method is: in the previous iterative process of decomposing a specific IMF component, continuation on the left side of the original signal sequence by 2 1 maximum value point and 2 minimum value points, and 2 maximum value points and 2 minimum value points are extended on the right;
  • ⁇ k-1 usually takes 0.1-0.3 times of the standard deviation of the (k-1)th residual component sequence.
  • the IMF component performs FFT processing on each IMF component to obtain its spectrum, and calculate the proportion of its energy in the respiratory frequency band (0.1Hz-0.5Hz) to the total energy. It is considered that when the proportion is higher than a certain threshold, the IMF component is the respiratory signal or part of the breathing signal, to be preserved. Finally, the sum of all the retained IMF components is considered as the respiratory waveform of the monitoring object.
  • Estimating the respiration frequency is to obtain the above-mentioned extracted respiration waveform through FFT to obtain a corresponding frequency spectrum. Search to find the peak in the spectrum, and the respiratory rate can be estimated according to the following formula:
  • f is the breathing frequency obtained by this method, in bpm; f 0 is the sampling frequency, in Hz. k and N are the subscript and maximum subscript corresponding to the peak value in the spectrum, respectively.
  • Figure 1 shows the effect of determining the position of the thorax.
  • the green line in the figure indicates the reference position of the thorax in the monitoring process obtained by the optical method, and the blue line indicates the position of the thorax estimated from the radar echo signal based on the above method. It can be seen from the figure that the above method can very accurately estimate the position of the chest of the monitored object.
  • Figure 2 shows the effect of respiration and heartbeat signals separated by the improved CEEMD algorithm. It can be seen from the figure that IMF1 and IMF2 are high-frequency noises that appear in the environment and the system, and IMF3 and IMF4 are the separated heartbeat signal components of the monitoring object. , IMF5 is the respiratory signal component of the monitoring object.
  • Figure 3 is a comparison between the respiratory waveform extracted from the radar echo signal and the reference waveform. It can be seen from the figure that the respiratory waveform extracted from the radar echo by the above algorithm is almost consistent with the reference respiratory waveform collected by the optical method of. It shows that the respiratory waveform extracted based on the above algorithm has high quality.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Veterinary Medicine (AREA)
  • Molecular Biology (AREA)
  • Pathology (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physiology (AREA)
  • Pulmonology (AREA)
  • Signal Processing (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Psychiatry (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Electromagnetism (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

一种基于多普勒雷达的非接触式呼吸监测方法，对雷达回波信号进行预处理，从雷达回波信号中提取监测对象胸廓所在的位置，提取监测对象胸廓所在位置的运动信息，从监测对象胸廓的运动信息中提取呼吸波形分量，基于呼吸波形的相应频谱估计得到监测对象的呼吸频率。这种监测对象胸廓的方法具有更高的定位精度，是能够准确提取胸廓运动信号的基础。通过改进之后的CEEMD算法进行呼吸和心跳信号的分离，对体动干扰、背景噪声和生命体征信号具有良好的分离效果，通过改进有效地改善了原始CEEMD算法中的端点效应问题，提高了分解速度。

Description

一种基于多普勒雷达的非接触式呼吸监测方法 技术领域
本发明属于呼吸频率监测技术领域，具体涉及一种基于多普勒雷达的非接触式呼吸监测方法。
背景技术
近年来关于非接触式生理特征监测的研究方兴未艾，各种非接触式生理特征监测方法不断涌现。它们基于的监测媒介多种多样，包括红外，声波和光学等。然而基于上述媒介的非接触式生命体征监测技术或多或少都具有局限性。比如红外探测器会受到监测对象附近的热源的影响；声波探测器以空气为传输介质，容易受到空气湿度和气流的影响，且在棉纱这类吸收声波的材料中无法使用；光学探测器容易受到各种光的干扰，在烟雾较多，视线不佳或者监测对象身体有遮挡的情况下都很难发挥作用。这些局限性阻碍了其进一步的发展。
基于多普勒雷达的非接触式生命体征监测技术，也被广泛地称为生物雷达技术，其采用微波作为探测媒介，具有抗干扰能力强，对非金属障碍物穿透力强，集成度高，功耗低等众多优点。但是现有基于多普勒雷达的生理特征监测方法在算法上大多比较简单，往往仅通过相应的带通滤波器等手段直接提取雷法回波中的生命体征信号，这使得提取到的信号中仍包含大量噪声和干扰，导致提取的信号质量不高。
发明内容
为了解决现有非接触式呼吸监测方法测量精度不高，使用条件受限，易受环境干扰等方面的问题，本发明提出一种基于多普勒雷达的新型呼吸监测方法。通过对雷达接收数据处理算法的创新设计，实现较好的呼吸监测效果。
为了实现上述目的，本发明采用如下的技术方案：
对雷达回波信号的预处理，对监测对象胸廓位置的确定，对监测对象胸廓运动信息的提取，对监测对象呼吸波形的提取以及对呼吸频率的估计。
S1、向监测对象所在区域发射周期性的线性调频连续波脉冲，并使用天线阵列采集相应的回波信号；
S2：所述对雷达回波信号的预处理主要包括对雷达回波信号进行混频，混频输出为一个中频信号x out，
Figure PCTCN2021109286-appb-000001
该中频信号的幅值A为常数，频率和相位均正比于反射面和天线的距离R，如下式所示：
Figure PCTCN2021109286-appb-000002
其中：S为线性调频连续波频率随时间的变化率，c为光速，λ为所用调频连续波的波长，均为常数。
最后对混频输出的中频信号进行FFT处理，得到频谱中不同频率的成分即代表不同距离处的物体反射的雷达回波信号。
S3：本方法中对监测对象胸廓位置的确定使用了一种基于MVDR的二维位置搜索算法：
S3-1、记某时刻，天线阵列中第n根天线接收到的数据经过上述预处理后的序列为x n，该序列中的第k个数据记为x n,k，则当前时刻，接收阵列所接收到的某一特定距离处的接收向量可以表示为：
x k＝[x 1,k x 2,k ... x N,k] H
其中N表示接收阵列中接收天线的根数。
S3-2、利用该接收向量x k计算协方差矩阵R xx，
R xx＝x kx k H；
S3-3、结合相应方向的导向矢量α(θ)，
Figure PCTCN2021109286-appb-000003
即可得到该距离和角度处的估计功率谱P k(θ)，
Figure PCTCN2021109286-appb-000004
S3-4、遍历监测范围内的所有的方向角θ并代入上式，即可得到监测范围内 当前距离处随波达方向角θ变化的功率谱P k(θ)；再改变上式中的k，按同样的方法得到监测范围内其他距离处随角度变化的功率谱。合并上述结果即可得到当前时刻监测范围的关于与天线间距离和相对天线阵列方向角的二维功率谱。
S3-5、找到该功率谱中的峰值位置，即认为是监测对象胸廓所在的位置。
S4、所述对监测对象胸廓运动信息的提取是根据上述对监测对象胸廓位置确定算法得到的位置参数，具体包括监测对象胸廓相对天线阵列的距离和方向角来提取特定位置的运动信息。其具体的计算过程为：根据监测对象相对天线阵列的距离参数以及上述预处理后中频信号的频率正比于该距离的特性，从预处理后的数据中心选择相应的中频信号成分。按照下式从中频信号的相位信息
Figure PCTCN2021109286-appb-000005
中提取监测对象胸廓所在距离处的运动信息R，
Figure PCTCN2021109286-appb-000006
根据监测对象胸廓相对天线阵列的方向角参数θ生成相应的权向量ω θ，
Figure PCTCN2021109286-appb-000007
通过该权向量，从天线阵列接收到监测对象胸廓所在距离处的运动信息R中进一步提取监测对象胸廓所在角度处的运动信息。
S5、所述对监测对象呼吸波形的提取采用了改进的完全集合经验模态分解方法，即改进的CEEMD算法。
S5-1、首先定义算子E j(.)为使用经典EMD算法求解第j个IMF分量。设x[n]表示原始信号序列，w i[n]表示分布为N(0,1)的高斯白噪声序列。
S5-2、为了提高计算速度，对IMF分量的判断条件进行了改进。
定义：
Figure PCTCN2021109286-appb-000008
其中u(t)和v(t)分别表示当前信号分量的上包络序列和下包络序列。IMF分量的判断条件共有2个：
Figure PCTCN2021109286-appb-000009
Figure PCTCN2021109286-appb-000010
其中#表示集合中元素的个数。通常取θ 1＝α＝0.05。
S5-3、采用包络极值延拓的方式对EMD方法中的端点效应进行抑制，具体的执行方法为：在分解某一特定IMF分量的历次迭代过程中，在原始信号序列左边延拓2个极大值点和2个极小值点，右边延拓2个极大值点和2个极小值点；
S5-4、对于提取到监测对象胸廓的运动信息序列，首先随机取M个不同的白噪声序列w i[n],i＝1,2,...,M，将它们的方差缩放到合适的尺度后分别叠加到原始信号序列x[n]上，通常ε取为原始信号序列标准差的0.1～0.3倍。按照下式进行第一次EMD分解并在原始信号序列减掉第一次的分解结果。
Figure PCTCN2021109286-appb-000011
r 1[n]＝x[n]-IMF 1[n]；
对于从第2个开始的IMF分量，均按照如下方式进行分解
Figure PCTCN2021109286-appb-000012
r k[n]＝r k-1[n]-IMF k[n]；
其中ε k-1通常取第(k-1)个剩余分量序列的标准差的0.1～0.3倍。
当剩余分量不再满足继续分解的条件时，停止分解，得到一个剩余的残差。
Figure PCTCN2021109286-appb-000013
对每个IMF分量进行FFT处理得到其频谱，统计其在呼吸频段(0.1Hz～0.5Hz)的能量占总能量的比例，认为当该比例高于某一阈值时，该IMF分量即为呼吸信号或呼吸信号的一部分，加以保留。最后将所有保留的IMF分量求和，即认为是监测对象的呼吸波形。
S6、对呼吸频率的估计是通过FFT获得上述提取到的呼吸波形，得到相应 的频谱。搜索找到频谱中的峰值，即可根据下式估计得到呼吸频率：
Figure PCTCN2021109286-appb-000014
其中f为通过此方法得到的呼吸频率，单位为bpm；f 0为采样频率，单位为Hz。k和N分别为频谱中峰值对应的下标和最大下标。
本发明的优点在于：本发明提出了一种基于多普勒雷达的非接触式呼吸监测算法，对于雷达天线接收到的回波信号，依次经过预处理，提取胸廓位置，提取胸廓运动信息，提取呼吸波形信息，呼吸频率估计等步骤，最终得到监测对象的呼吸波形和频率参数。相比传统的监测方法，本方法的监测对象的胸廓具有更高的定位精度，这是能够准确提取胸廓运动信号的基础。通过改进之后的CEEMD算法进行呼吸和心跳信号的分离，该算法对体动干扰，背景噪声和生命体征信号具有良好的分离效果，通过改进，也有效地改善了原始CEEMD算法中的端点效应的问题，同时提高了分解速度。
附图说明
图1为胸廓位置确定效果展示。
图2为通过改进的CEEMD算法分离得到呼吸和心跳信号效果展示。
图3为通过雷达回波信号提取到的呼吸波形与基准波形的对比。
具体实施方式
为使本发明实现的技术手段、创作特征、达成目的与功效易于明白了解，下面结合具体实施方式，进一步阐述本发明。
本实施例中提出了一种基于多普勒雷达的非接触式呼吸监测方法，具体如下：
S1、向监测对象所在区域发射周期性的线性调频连续波脉冲，并使用天线阵列采集相应的回波信号；
S2：所述对雷达回波信号的预处理主要包括对雷达回波信号进行混频，混频输出为一个中频信号x out，
Figure PCTCN2021109286-appb-000015
该中频信号的幅值A为常数，频率和相位均正比于反射面和天线的距离R，如下式所示：
Figure PCTCN2021109286-appb-000016
其中：S为线性调频连续波频率随时间的变化率，c为光速，λ为所用调频连续波的波长，均为常数。
最后对混频输出的中频信号进行FFT处理，得到频谱中不同频率的成分即代表不同距离处的物体反射的雷达回波信号。
S3：本方法中对监测对象胸廓位置的确定使用了一种基于MVDR的二维位置搜索算法：
S3-1、记某时刻，天线阵列中第n根天线接收到的数据经过上述预处理后的序列为x n，该序列中的第k个数据记为x n,k，则当前时刻，接收阵列所接收到的某一特定距离处的接收向量可以表示为：
x k＝[x 1,k x 2,k ... x N,k] H
其中N表示接收阵列中接收天线的根数。
S3-2、利用该接收向量x k计算协方差矩阵R xx，
R xx＝x kx k H；
S3-3、结合相应方向的导向矢量α(θ)，
Figure PCTCN2021109286-appb-000017
即可得到该距离和角度处的估计功率谱P k(θ)，
Figure PCTCN2021109286-appb-000018
S3-4、遍历监测范围内的所有的方向角θ并代入上式，即可得到监测范围内当前距离处随波达方向角θ变化的功率谱P k(θ)；再改变上式中的k，按同样的方法得到监测范围内其他距离处随角度变化的功率谱。合并上述结果即可得到当前时刻监测范围的关于与天线间距离和相对天线阵列方向角的二维功率谱。
S3-5、找到该功率谱中的峰值位置，即认为是监测对象胸廓所在的位置。
S4、所述对监测对象胸廓运动信息的提取是根据上述对监测对象胸廓位置确定算法得到的位置参数，具体包括监测对象胸廓相对天线阵列的距离和方向角来提取特定位置的运动信息。其具体的计算过程为：根据监测对象相对天线阵列的距离参数以及上述预处理后中频信号的频率正比于该距离的特性，从预处理后的数据中心选择相应的中频信号成分。按照下式从中频信号的相位信息
Figure PCTCN2021109286-appb-000019
中提取监测对象胸廓所在距离处的运动信息R，
Figure PCTCN2021109286-appb-000020
根据监测对象胸廓相对天线阵列的方向角参数θ生成相应的权向量ω θ，
Figure PCTCN2021109286-appb-000021
通过该权向量，从天线阵列接收到监测对象胸廓所在距离处的运动信息R中进一步提取监测对象胸廓所在角度处的运动信息。
S5、所述对监测对象呼吸波形的提取采用了改进的完全集合经验模态分解方法，即改进的CEEMD算法。
S5-1、首先定义算子E j(.)为使用经典EMD算法求解第j个IMF分量。设x[n]表示原始信号序列，w i[n]表示分布为N(0,1)的高斯白噪声序列。
S5-2、为了提高计算速度，对IMF分量的判断条件进行了改进。
定义：
Figure PCTCN2021109286-appb-000022
其中u(t)和v(t)分别表示当前信号分量的上包络序列和下包络序列。IMF分量的判断条件共有2个：
Figure PCTCN2021109286-appb-000023
Figure PCTCN2021109286-appb-000024
其中#表示集合中元素的个数。通常取θ 1＝α＝0.05。
S5-3、采用包络极值延拓的方式对EMD方法中的端点效应进行抑制，具体的执行方法为：在分解某一特定IMF分量的历次迭代过程中，在原始信号序列左边延拓2个极大值点和2个极小值点，右边延拓2个极大值点和2个极小值点；
S5-4、对于提取到监测对象胸廓的运动信息序列，首先随机取M个不同的白噪声序列w i[n],i＝1,2,...,M，将它们的方差缩放到合适的尺度后分别叠加到原始信号序列x[n]上，通常ε取为原始信号序列标准差的0.1～0.3倍。按照下式进行第一次EMD分解并在原始信号序列减掉第一次的分解结果。
Figure PCTCN2021109286-appb-000025
r 1[n]＝x[n]-IMF 1[n]；
对于从第2个开始的IMF分量，均按照如下方式进行分解
Figure PCTCN2021109286-appb-000026
r k[n]＝r k-1[n]-IMF k[n]；
其中ε k-1通常取第(k-1)个剩余分量序列的标准差的0.1～0.3倍。
当剩余分量不再满足继续分解的条件时，停止分解，得到一个剩余的残差。
Figure PCTCN2021109286-appb-000027
对每个IMF分量进行FFT处理得到其频谱，统计其在呼吸频段(0.1Hz～0.5Hz)的能量占总能量的比例，认为当该比例高于某一阈值时，该IMF分量即为呼吸信号或呼吸信号的一部分，加以保留。最后将所有保留的IMF分量求和，即认为是监测对象的呼吸波形。
S6、对呼吸频率的估计是通过FFT获得上述提取到的呼吸波形，得到相应的频谱。搜索找到频谱中的峰值，即可根据下式估计得到呼吸频率：
Figure PCTCN2021109286-appb-000028
其中f为通过此方法得到的呼吸频率，单位为bpm；f 0为采样频率，单位为 Hz。k和N分别为频谱中峰值对应的下标和最大下标。
采用上述实施例的方法对进行患者呼吸监测，得到如下结果：
图1所示为胸廓位置确定效果展示，图中绿色的线表示采用光学方法得到监测过程中胸廓的基准位置，蓝色的线表示基于上述方法从雷达回波信号中估计得到的胸廓的位置，从图中可以看出，上述方法能够十分准确地估算出监测对象胸廓的位置。
图2为通过改进的CEEMD算法分离得到呼吸和心跳信号效果展示，从图中可以看出IMF1和IMF2为环境与系统中出现的高频噪声，IMF3和IMF4为分离出的监测对象的心跳信号分量，IMF5为监测对象的呼吸信号分量。
图3为通过雷达回波信号提取到的呼吸波形与基准波形的对比，从图中可以看出通过上述算法从雷达回波中提取到的呼吸波形与通过光学方法采集的基准呼吸波形几乎是一致的。说明基于上述算法提取到的呼吸波形具有较高的质量。
由技术常识可知，本发明可以通过其它的不脱离其精神实质或必要特征的实施方案来实现。因此，上述公开的实施方案，就各方面而言，都只是举例说明，并不是仅有的。所有在本发明范围内或在等同于本发明的范围内的改变均被本发明包含。

Claims (7)

  1. 一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，包括如下步骤：
    S1、向监测对象所在区域发射周期性的线性调频连续波脉冲，并使用天线阵列采集相应的回波信号；
    S2、采用混频和FFT的方式对雷达回波信号进行预处理；
    S3、从雷达回波信号中提取监测对象胸廓所在的位置，具体包括相对于天线阵列的距离和方向角参数；
    S4、采用相位相关的方法，从雷达回波信号中提取监测对象胸廓所在位置的运动信息；
    S5、采用改进的CEEMD算法，从监测对象胸廓的运动信息中提取呼吸波形分量；
    S6、利用提取到监测对象的呼吸波形得到相应的频谱，基于此频谱估计得到监测对象的呼吸频率。
  2. 根据权利要求1所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于：所述步骤S1中的线性调频连续波脉冲调频范围为77Ghz～81GHz，单个脉冲持续周期为50us，每次采样时会连续发射3个上述脉冲，采样频率为20Hz，用于采集雷达回波信号的天线阵列包含4根等距的贴片天线。
  3. 根据权利要求1所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，所述步骤S2的预处理过程中，混频输出为一个中频信号，其幅值为一常数，频率和相位均正比于反射物体相对天线阵列的距离。
  4. 根据权利要求1所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，所述步骤S4中获取监测对象胸廓所在的位置的方法如下：
    S3-1、基于接收天线阵列接收到的数据，经过预处理后得到某一距离处的接收向量x k＝[x 1,k x 2,k ... x N,k] H，其中N表示接收阵列中接收天线的根数；
    S3-2、基于接收向量x k，根据R xx＝x kx k H计算得到当前距离处的协方差矩阵；
    S3-3、根据
    Figure PCTCN2021109286-appb-100001
    生成导向矢量α(θ)，基于协方差矩阵R xx生成当前距离不同波达方向的功率谱
    Figure PCTCN2021109286-appb-100002
    其中θ为监测对象胸廓相对天线阵列的方向角参数，H表示对导向矢量进行共轭转置，λ为使用多普勒雷达发出线性调频波的波长；
    S3-4、遍历所有的距离下标，得到监测范围内每一距离处随波达方向变化的功率谱，合并结果，即得到当前监测范围的二维功率谱；
    S3-5、搜索上述监测范围二维功率谱的峰值，找出其相对于天线阵列的距离和方向角参数，即认为是监测对象胸廓所在的位置。
  5. 根据权利要求4所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，所述步骤S4中监测对象胸廓运动信息提取的方法如下：
    根据监测对象相对天线阵列的距离参数以及上述预处理后中频信号的频率正比于该距离的特性，从预处理后的数据中心选择相应的中频信号成分，按照下式从中频信号的相位信息
    Figure PCTCN2021109286-appb-100003
    中提取监测对象胸廓所在距离处的运动信息R：
    Figure PCTCN2021109286-appb-100004
    根据监测对象胸廓相对天线阵列的方向角参数θ生成相应的权向量ω θ：
    Figure PCTCN2021109286-appb-100005
    通过该权向量ω θ，从天线阵列接收到监测对象胸廓所在距离处的运动信息R中进一步提取监测对象胸廓所在角度处的运动信息。
  6. 根据权利要求1所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，所述步骤S5中使用改进后的CEEMD算法从监测对象胸廓的运动信息中提取其呼吸波形，该算法的具体过程如下：
    S5-1、采用CEEMD的方式对监测对象胸廓的运动序列进行分解；
    S5-2、计算
    Figure PCTCN2021109286-appb-100006
    其中u(t)和v(t)分别表示当前信号分量的上包络序列和下包络序列，通过判断以下两个条件作为判断EMD分离出IMF分量的判断条件：
    (1)
    Figure PCTCN2021109286-appb-100007
    (2)
    Figure PCTCN2021109286-appb-100008
    其中，θ 1＝α＝0.05；
    S5-3、通过极值包络延拓的方法对经典EMD算法中的端点效应进行抑制，其具体步骤为在原EMD分解过程中，在原极大值序列和极小值序列两端各自延拓2个点；
    S5-4、对于采用改进CEEMD方法得到的各阶IMF分量，通过计算其在呼吸频段的能量占总能量的比例，判断其是否为呼吸波形或呼吸波形的分量。
  7. 根据权利要求1所述的一种基于多普勒雷达的非接触式呼吸监测方法，其特征在于，所述步骤S6中通过选取一定时长的呼吸波形序列，使用FFT得到该序列对应的频谱，搜索该频谱中的峰值，并由此估计得到监测对象的呼吸频率：
    Figure PCTCN2021109286-appb-100009
    其中，f为通过此方法得到的呼吸频率，单位为bpm；f 0为采样频率，单位为Hz，k和N分别为频谱中峰值对应的下标和最大下标。
PCT/CN2021/109286 2021-07-29 2021-07-29 一种基于多普勒雷达的非接触式呼吸监测方法 WO2023004688A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202180029015.9A CN115460980A (zh) 2021-07-29 2021-07-29 一种基于多普勒雷达的非接触式呼吸监测方法
PCT/CN2021/109286 WO2023004688A1 (zh) 2021-07-29 2021-07-29 一种基于多普勒雷达的非接触式呼吸监测方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/109286 WO2023004688A1 (zh) 2021-07-29 2021-07-29 一种基于多普勒雷达的非接触式呼吸监测方法

Publications (1)

Publication Number Publication Date
WO2023004688A1 true WO2023004688A1 (zh) 2023-02-02

Family

ID=84294747

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/109286 WO2023004688A1 (zh) 2021-07-29 2021-07-29 一种基于多普勒雷达的非接触式呼吸监测方法

Country Status (2)

Country Link
CN (1) CN115460980A (zh)
WO (1) WO2023004688A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116776083A (zh) * 2023-06-19 2023-09-19 浙江华东建设工程有限公司 一种用于多波束海底地形测量的信号采集降噪方法
CN116840805A (zh) * 2023-08-30 2023-10-03 长沙莫之比智能科技有限公司 基于mimo雷达与波束形成的人体生命体征检测方法
CN118303876A (zh) * 2024-06-11 2024-07-09 长春大学 一种多尺度特征融合呼吸运动预测方法及系统
CN118356176A (zh) * 2024-06-19 2024-07-19 大连海事大学 一种适用于非接触式疲劳检测的无线体征挖掘方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116148850B (zh) * 2023-04-23 2023-07-14 中南大学 一种远距离人体呼吸信号检测的方法、系统及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106646457A (zh) * 2016-11-02 2017-05-10 南京慧尔视智能科技有限公司 基于微波的人员行为检测方法与装置
EP3384843A1 (en) * 2017-04-05 2018-10-10 Nokia Technologies Oy Determining breathing attributes by a detection device
CN110118966A (zh) * 2019-05-28 2019-08-13 长沙莫之比智能科技有限公司 基于毫米波雷达的人员检测与计数系统
CN111373282A (zh) * 2017-12-07 2020-07-03 德州仪器公司 用于fmcw雷达系统的雷达处理链
CN112674740A (zh) * 2020-12-22 2021-04-20 北京工业大学 一种基于毫米波雷达的生命体征检测方法
CN112816981A (zh) * 2021-01-22 2021-05-18 深圳迈睿智能科技有限公司 提高人体存在探测可靠度的微波探测装置和探测方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106646457A (zh) * 2016-11-02 2017-05-10 南京慧尔视智能科技有限公司 基于微波的人员行为检测方法与装置
EP3384843A1 (en) * 2017-04-05 2018-10-10 Nokia Technologies Oy Determining breathing attributes by a detection device
CN111373282A (zh) * 2017-12-07 2020-07-03 德州仪器公司 用于fmcw雷达系统的雷达处理链
CN110118966A (zh) * 2019-05-28 2019-08-13 长沙莫之比智能科技有限公司 基于毫米波雷达的人员检测与计数系统
CN112674740A (zh) * 2020-12-22 2021-04-20 北京工业大学 一种基于毫米波雷达的生命体征检测方法
CN112816981A (zh) * 2021-01-22 2021-05-18 深圳迈睿智能科技有限公司 提高人体存在探测可靠度的微波探测装置和探测方法

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
GAO HUI: "Research on Key Technology of 77GHz FM CW Radar Scene Perception", CHINESE MASTER'S THESES FULL-TEXT DATABASE, 1 December 2019 (2019-12-01), pages 1 - 81, XP093028639, DOI: 10.27241/d.cnki.gnjgu.2019.002157 *
LIU YI: "Research on key technology of lung sound signal processing based on EMD technology", CHINESE DOCTORAL DISSERTATIONS FULL-TEXT DATABASE, 5 June 2019 (2019-06-05), pages 1 - 122, XP093028660, ISSN: 1013-7653, DOI: 10.27170/d.cnki.gjsuu.2019.000353 *
WANG YONG: "Design of non-contact sign monitoring module in mobile nursing system", CHINA CIO NEWS, 25 November 2017 (2017-11-25), pages 34 - 35, XP093028636 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116776083A (zh) * 2023-06-19 2023-09-19 浙江华东建设工程有限公司 一种用于多波束海底地形测量的信号采集降噪方法
CN116776083B (zh) * 2023-06-19 2024-07-16 浙江华东岩土勘察设计研究院有限公司 一种用于多波束海底地形测量的信号采集降噪方法
CN116840805A (zh) * 2023-08-30 2023-10-03 长沙莫之比智能科技有限公司 基于mimo雷达与波束形成的人体生命体征检测方法
CN116840805B (zh) * 2023-08-30 2023-11-10 长沙莫之比智能科技有限公司 基于mimo雷达与波束形成的人体生命体征检测方法
CN118303876A (zh) * 2024-06-11 2024-07-09 长春大学 一种多尺度特征融合呼吸运动预测方法及系统
CN118356176A (zh) * 2024-06-19 2024-07-19 大连海事大学 一种适用于非接触式疲劳检测的无线体征挖掘方法

Also Published As

Publication number Publication date
CN115460980A (zh) 2022-12-09

Similar Documents

Publication Publication Date Title
WO2023004688A1 (zh) 一种基于多普勒雷达的非接触式呼吸监测方法
CN111352102B (zh) 一种基于调频连续波雷达的多目标个数检测方法及装置
CN109507653A (zh) 一种基于uwb的多信息感知生物雷达系统及其获取目标信息的方法
CN112686094B (zh) 一种基于毫米波雷达的非接触式身份识别方法及系统
CN102058411B (zh) 一种多通道基于uwb雷达式生命探测仪
US20140266860A1 (en) Method and system for activity detection and classification
CN106805940A (zh) 一种连续波生物雷达体征检测装置
CN112998668B (zh) 基于毫米波的非接触式远场多人体呼吸心率监测方法
CN107132512B (zh) 基于多通道hht的uwb雷达人体运动微多普勒特征提取方法
CN112336322A (zh) 一种无接触式呼吸或心跳检测方法
CN109901130B (zh) 一种基于Radon变换和改进2DPCA的旋翼无人机检测与识别方法
CN114847911A (zh) 一种基于毫米波雷达的多人生命体征监测方法
CN106805941A (zh) 一种连续波生物雷达体征检测装置
CN113341392B (zh) 一种基于多站雷达微多普勒运动测向的人体行为分类方法
Qiao et al. Learning-refined integral null space pursuit algorithm for noncontact multisubjects vital signs measurements using SFCW-UWB and IR-UWB radar
KR20210006225A (ko) 레이더 기반 다중 객체 생체신호 측정방법 및 그 장치
CN114732390A (zh) 一种基于调频连续波雷达的非接触式心率变异性监测方法
CN106646466A (zh) 一种基于主成分分析的加权后向投影算法的成像方法
CN113786176B (zh) 精确的毫米波雷达呼吸心跳测量方法、系统和存储介质
Jing et al. Harmonics and intermodulation products-based fuzzy logic (HIPBFL) algorithm for vital sign frequency estimation using a UWB radar
Faisal et al. Human activity recognition from FMCW radar signals utilizing cross-terms free WVD
Xue et al. Accurate multi-target vital signs detection method for FMCW radar
CN107526064A (zh) 基于二维特征的自适应lfm信号参数估计方法
CN115372959A (zh) 毫米波调频连续波雷达探测非视距区域内人体的探测方法
Duong Principal component analysis for heart rate measurement using UWB radar

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21951295

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21951295

Country of ref document: EP

Kind code of ref document: A1