Non-contact vital sign estimation method and system based on FMCW millimeter wave radar

By constructing a breathing and heartbeat model with harmonics in FMCW millimeter wave radar technology and using Newton iterative method, the interference problem of respiratory harmonics on heart rate estimation is solved, and accurate monitoring of breathing and heart rate is achieved, suitable for contactlessness surveillance of vital signs, especially in patients with skin fragility.

CN119025814AActive Publication Date: 2024-11-26SUN YAT SEN UNIV

Patent Information

Application Number
CN202411070237.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2024-11-26
Estimated Expiration
2044-08-06

AI Technical Summary

Technical Problem

In the monitoring of contactless vital signs, especially in breathing and heart rate estimation, there are challenges to heart rate interference from respiratory harmonics, and existing methods such as bandpass filtering and wavelet transformation have limitations, making it difficult to effectively solve the interference problem. , especially in surveillance of patients with fragile skin barriers.

Method used

Using the FMCW millimeter wave radar method, a respiration and heartbeat model with harmonics is constructed to establish a likelihood function about breathing rate and heart rate, and the maximum likelihood estimation is solved by using the Newtonian iterative method to reduce the interference of respiratory harmonics on heart rate estimation. , improve the accuracy and reliability of signal processing.

Benefits of technology

Accurate estimation of respiratory rate and heart rate is achieved, interference is reduced, and monitoring accuracy and reliability is improved. It is suitable for contactless monitoring, especially for long-term monitoring of patients with skin fragility.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025814A_ABST
    Figure CN119025814A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of radars, and particularly relates to a non-contact vital sign estimation method and system based on an FMCW millimeter wave radar, and the method comprises the steps: carrying out the preprocessing of a radar echo signal, and extracting a chest wall displacement signal from the radar echo signal; modeling the chest wall displacement signal into a combination of a fundamental frequency and a harmonic wave thereof, establishing a likelihood function about a respiration rate and a heart rate, constructing an objective function of maximum likelihood estimation of a respiration or heartbeat frequency, and representing the maximum likelihood estimation of the respiration or heartbeat frequency as an optimization problem of the objective function; and solving the optimization problem of the objective function through a Newton iteration method to obtain the maximum likelihood estimation values of the respiratory rate and the heart rate. According to the method, the harmonic energy in the signal is fully utilized, the precision and reliability of signal processing are improved, and the interference of breathing harmonics on heart rate estimation is effectively reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of radar technology, and in particular relates to a non-contact vital sign estimation method and system based on FMCW millimeter wave radar. Background Art

[0002] With the improvement of living standards, people's health awareness is getting higher and higher, and radar-based non-contact vital sign detection technology is increasingly attracting people's attention. Breathing and heartbeat are the main basis for evaluating a person's health status. Research on the technology of monitoring breathing and heartbeat signals based on radar sensors is of great significance in medicine. Compared with other vital sign measurement sensors, radar sensors have unique advantages. Traditional vital sign detection methods mainly include electrocardiograph (ECG), photocapacitance scanning (PCG), airflow sensing and other methods. These methods are mainly used in operating rooms and intensive care units of hospitals. They require direct contact with the skin through electrodes, which will cause tactile discomfort, epidermal damage or pressure necrosis in patients, which is not conducive to patient recovery. In some special scenarios, such as patients with severe burns, premature babies or patients with severe skin diseases, special patients with fragile skin barriers, long-term contact with the equipment will cause damage to their skin and even cause infection. In this case, patients cannot wear the measuring device for a long time, which makes the traditional vital sign detection method unable to meet the needs of all-day detection. For daily vital sign monitoring of bedridden elderly people or paralyzed patients, the discomfort caused by contact equipment makes it difficult for the target group to accept it. Radar measurement systems can monitor for a long time and can be directly applied to special patients (such as premature babies) without harming the patients. Therefore, it is very necessary to study non-contact and accurate vital sign measurement methods based on radar.

[0003] Achieving accurate non-contact vital signs monitoring, especially monitoring related to heart rate estimation, remains a challenging task. Traditional methods usually simply simulate breathing and heartbeat signals as the superposition of two sine waves, ignoring the harmonic components in the actual signal. The accuracy of breathing estimation is usually limited by noise, while heart rate estimation is mainly affected by the strong interference of breathing signals and their higher harmonics. Existing methods for dealing with the interference of respiratory harmonics on heart rate include bandpass filtering and wavelet transform. These methods can effectively suppress the harmonic signals caused by breathing and improve the accuracy of heart rate estimation. However, the effect of the bandpass filter is limited by the close frequency band range of the breathing and heartbeat signals, and cannot effectively solve the interference problem. Wavelet transform can extract breathing and heartbeat signals, but it is necessary to select an appropriate wavelet basis function, which is a difficulty, and after selection, it cannot be flexibly adjusted according to different signals and lacks adaptability.

[0004] In summary, existing technologies have technical challenges and limitations in dealing with the interference of respiratory harmonics on heart rate, and further research and improvement are needed to improve accuracy and practicality. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention proposes a non-contact vital signs estimation method and system based on FMCW millimeter wave radar. By constructing a breathing and heartbeat model with harmonics, a likelihood function about the breathing rate and heart rate is established, and the maximum likelihood estimation of the breathing rate and heart rate is derived; the harmonic energy in the signal is fully utilized, the accuracy and reliability of signal processing are improved, and the interference of respiratory harmonics on heart rate estimation is effectively reduced.

[0006] A non-contact vital sign estimation method based on FMCW millimeter wave radar in an embodiment of the present invention includes the following steps:

[0007] S1, preprocessing the radar echo signal to extract the chest wall displacement signal;

[0008] S2. Modeling the chest wall displacement signal as a combination of the fundamental frequency of breathing and heartbeat and its harmonics, establishing likelihood functions for the breathing rate and heart rate, constructing an objective function for maximum likelihood estimation of the breathing or heartbeat frequency, and expressing the maximum likelihood estimation of the breathing or heartbeat frequency as an optimization problem of the objective function;

[0009] S3. Solve the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

[0010] A non-contact vital sign estimation system based on FMCW millimeter wave radar in an embodiment of the present invention includes the following modules:

[0011] A preprocessing module, used for preprocessing the radar echo signal and extracting the chest wall displacement signal therefrom;

[0012] A model building module, used for modeling the chest wall displacement signal as a combination of the fundamental frequency of breathing and heartbeat and its harmonics, establishing a likelihood function about the breathing rate and heart rate, constructing an objective function of maximum likelihood estimation of the breathing or heartbeat frequency, and expressing the maximum likelihood estimation of the breathing or heartbeat frequency as an optimization problem of the objective function;

[0013] The model solving module solves the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

[0014] Compared with the prior art, the technical effects achieved by the present invention include:

[0015] 1. The present invention utilizes the harmonic structure existing in the respiratory and heartbeat signals, constructs a respiratory and heartbeat model with harmonics, and solves the maximum likelihood estimation function about the respiratory rate and heart rate based on the Newton method. The present invention does not need to use a bandpass filter to separate the respiratory and heart rate signals. It only needs to simply divide the search area and make independent estimates within the respiratory and heart rate ranges to derive the maximum likelihood estimation of the respiratory rate and heart rate. This method makes full use of the harmonic energy in the signal, improves the accuracy and reliability of signal processing, and effectively reduces the interference of respiratory harmonics on heart rate estimation, which is expected to play an important role in contactless vital signs monitoring.

[0016] 2. The present invention first eliminates the interference of static strong reflectors through the phasor cancellation method, thereby realizing the correct estimation of the target distance; secondly, the inverse tangent method and phase unwrapping are used to correctly extract the vital sign signals, and then the energy of the signals is enhanced through slow-time phase correlation; then, a respiration and heartbeat model with harmonics is constructed, the maximum likelihood estimation of the respiration rate and heart rate is derived, and the respiration rate and heart rate are solved through Newton iteration. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of a non-contact vital sign estimation method in an embodiment of the present invention;

[0018] Figure 2 is a distance spectrum diagram in an embodiment of the present invention, wherein (a) is a schematic diagram before phasor cancellation, and (b) is a schematic diagram after phasor cancellation;

[0019] Figure 3 is a curve diagram of a chest wall displacement signal after phase unwrapping in an embodiment of the present invention;

[0020] Figure 4 is a schematic diagram of the polynomial regression method for removing motion trends in an embodiment of the present invention;

[0021] Figure 5 is a graph of polynomial regression removing motion trends in an embodiment of the present invention;

[0022] Figure 6 Schematic diagrams before and after slow time phase correlation processing in an embodiment of the present invention, wherein (a) is a time domain comparison schematic diagram, and (b) is a frequency domain comparison schematic diagram;

[0023] Figure 7 is a schematic diagram comparing the respiratory rate estimation result with the reference sensor;

[0024] Figure 8 It is a schematic diagram comparing the heart rate estimation results with the reference sensor. DETAILED DESCRIPTION

[0025] The present invention is further described in detail below with reference to the embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0026] Example

[0027] This embodiment provides a non-contact vital sign estimation method based on FMCW millimeter wave radar. Figure 1 As shown, the main steps include:

[0028] S1. Preprocess the radar echo signal to extract the chest wall displacement signal.

[0029] In this embodiment, this step specifically includes:

[0030] S11, mixing and filtering the radar echo signal to obtain a radar intermediate frequency signal; the radar intermediate frequency signal is sampled by an analog-to-digital converter (ADC), and the sampled radar intermediate frequency signal can be expressed as:

[0031]

[0032] Among them A b and f IF are the amplitude and frequency of the radar intermediate frequency signal, T f is the frame period, T s is the fast time sampling interval, λ is the wavelength of the radar signal, R0 is the distance from the human target to the radar, r(nT f ) represents nT f The chest displacement at the moment; n represents the nth radar intermediate frequency signal, also known as the slow time (speed dimension) sampling point; m represents the mth ADC sampling, also known as the fast time dimension (distance dimension) sampling point.

[0033] S12. Use the phase cancellation method to eliminate the echo interference of the stationary obstacle, and obtain the radar intermediate frequency signal after the echo interference is eliminated.

[0034] Specifically, all radar intermediate frequency signals are averaged and the average value is used as the reference signal; then the reference signal is subtracted from each received radar intermediate frequency signal to eliminate the echo interference of stationary obstacles and improve the ranging accuracy. The reference signal can be expressed as:

[0035]

[0036] Where R[m,n] is the matrix of radar intermediate frequency signals, m is the fast time dimension (distance dimension) sampling point, n is the slow time dimension (speed dimension) sampling point, and N is the total number of received radar intermediate frequency signals; the formula of the phasor cancellation algorithm is:

[0037] R a [m,n]=R[m,n]-C[m](3)

[0038] Where R a [m,n] is the matrix of radar intermediate frequency signal after echo interference is eliminated.

[0039] The distance spectrum before and after phase cancellation is as follows: Figure 2 As shown, (a) is the distance spectrum before cancellation, and (b) is the distance spectrum after cancellation.

[0040] S13, perform distance dimension fast Fourier transform on the radar intermediate frequency signal after eliminating echo interference, and find the peak value after fast Fourier transform to determine the distance from the human target to the radar. The expression of the distance R0 from the human target to the radar is:

[0041]

[0042] S14. Demodulate the signal at the range gate where the human target is located by using the inverse tangent and phase unwrapping methods to obtain the phase signal of the range gate where the human target is located as the chest wall displacement signal caused by breathing and heartbeat.

[0043] In this step, the phase signal of the range gate where the human target is located is extracted from the one-dimensional range spectrum diagram. The extracted phase signal can be expressed as:

[0044]

[0045] Among them B I (t) is the in-phase branch of the phase signal; B Q (t) is the orthogonal branch of the phase signal, A is the amplitude of the phase signal, θ is the fixed phase shift caused by the absolute distance between the human target and the radar, δ(t) is the chest wall displacement caused by breathing and heartbeat, and φ is the total phase noise generated by the system.

[0046] By using the orthogonal characteristics of the IQ signals, a phase function containing chest displacement can be obtained, thereby extracting phase information:

[0047]

[0048] Since the value range of the inverse tangent function is in the interval [-π, π], the phase value obtained by AD demodulation may be different from the actual phase. When some values ​​in the actual phase Φ(t) exceed this interval, phase wrapping will occur, that is, the demodulated result of the value exceeding this range is still between [-π, π], which is equivalent to a jump of 2π at [-π, π]. This embodiment performs phase unwrapping by comparing the AD demodulated phase values ​​of adjacent sampling points.

[0049]

[0050] Where f(x) represents the phase information after unwrapping, Φ n+1 Indicates the phase value at the n+1th sampling moment, Φ n Indicates the phase value at the nth sampling moment.

[0051] The chest wall displacement signal curve obtained after phase unwrapping is as follows: Figure 3 shown.

[0052] S15. Subsequently, in view of the interference caused by slight shaking of the human body, polynomial regression is used to remove the motion trend (i.e., remove the offset) of the chest wall displacement signal, thereby eliminating the strong low-frequency component interference in the frequency domain and improving the accuracy of subsequent parameter estimation.

[0053] The principle of polynomial regression to remove offset is as follows Figure 4 As shown, the motion trend signal p(t) of the chest wall displacement signal δ(t) is calculated based on the k-order polynomial regression, and then the motion trend signal p(t) is subtracted from the chest wall displacement signal δ(t) to obtain the vital sign signal after the motion trend is removed, that is, the corrected phase signal y(t), which is used as the chest wall displacement signal after the motion trend is removed.

[0054] Figure 5 The signal curves before and after polynomial regression offset removal are shown.

[0055] S16. Perform slow-time phase correlation processing on the chest wall displacement signal after removing the motion trend.

[0056] When performing slow time phase correlation processing in this step, the correlation between the chest wall displacement signals at the range gate where the human target is located and the chest wall displacement signals at the adjacent range gates is analyzed, and their Pearson correlation coefficient is calculated.

[0057]

[0058] The human target is located at the dth range gate, and the i-th range gate is adjacent to the d-th range gate. and are the phase signal vectors (i.e., chest wall displacement signals) at the dth range gate and the ith range gate, respectively. represents the covariance between the phase signal vectors; is the corresponding standard deviation.

[0059] Figure 6 In FIG. 1 , sub-figure (a) illustrates the time domain comparison before and after the slow time phase correlation, and sub-figure (b) illustrates the frequency domain comparison before and after the slow time phase correlation.

[0060] When the Pearson correlation coefficient exceeds a predetermined threshold, the chest wall displacement signals are summed and the signal-to-noise ratio is improved accordingly. This allows efficient use of energy in the vicinity, thus improving the estimation performance without increasing the computational burden.

[0061] S2. Model the chest wall displacement signal as a combination of the fundamental frequency of breathing and heart rate and their harmonics, establish likelihood functions for respiratory rate and heart rate, construct an objective function for maximum likelihood estimation of respiratory rate or heart rate, and express the maximum likelihood estimation of respiratory rate or heart rate as an optimization problem of the objective function.

[0062] In this embodiment, step S2 includes:

[0063] S21. Build a respiratory and heartbeat signal model that includes harmonics.

[0064] In most current studies, for the modeling and analysis of vital signs signals, the simulation of breathing and heartbeat signals is generally set as the superposition of two sinusoidal signals with different amplitudes and frequencies. However, this does not conform to the actual situation. In fact, breathing and heartbeat signals are not single-frequency signals, but contain harmonic components and intermodulation components; that is, breathing and heartbeat signals contain multiple harmonics, which are integer multiples of the fundamental frequencies of breathing and heartbeat.

[0065] In order to utilize the harmonic structure in the respiratory and heartbeat signals, make full use of the energy of the harmonics, and reduce the interference of respiratory harmonics on heart rate estimation, this embodiment constructs a respiratory and heartbeat signal model containing harmonics to model the chest wall displacement signal as a combination of the respiratory and heartbeat fundamental frequencies and their harmonics. Through this signal model, solving within the frequency range of respiratory and heartbeat respectively can more effectively capture the harmonic components of respiratory and heartbeat signals, accumulate the harmonic energy and fundamental frequency energy, improve the accuracy and reliability of signal processing, and thus optimize respiratory rate and heart rate estimation.

[0066] The respiration and heartbeat signal model containing harmonics constructed in this embodiment can be expressed as:

[0067]

[0068] Where f is the breathing or heartbeat frequency, a1, a2 and a3 represent the amplitudes of the fundamental, second harmonic and third harmonic of breathing or heartbeat respectively. and are the phases of the fundamental wave, second harmonic, and third harmonic of breathing or heartbeat, respectively. ω[n] represents a mean of 0 and a variance of σ 2 Gaussian white noise, n represents the nth sampling moment.

[0069] That is to say, in this embodiment, the breathing signal model and the heartbeat signal model are both represented by the same formula (9). When f is the breathing frequency, formula (9) represents the breathing signal model; when f is the heartbeat frequency, formula (9) represents the heartbeat signal model.

[0070] S22. Based on the constructed breathing and heartbeat signal model, a probability density function of the variable to be estimated is obtained.

[0071] According to the above signal model, the probability density function of the variable to be estimated can be obtained:

[0072]

[0073] in is the variable to be estimated; x is the vector form of x[n].

[0074] S23. Obtain an estimated value of the variable Θ to be estimated by maximizing the likelihood function.

[0075] Maximizing the likelihood function is equivalent to minimizing the following formula:

[0076]

[0077] Where A=[a1,a2,a3], Simplify J(A, f, Φ) into the following form:

[0078] J(A, f, Φ)=(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3) T ×(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3)(12)

[0079] in:

[0080] α=[α1, α2, β1, β2, γ1, γ2] T

[0081] H=[c1,s1,c2,s2,c3,s3] T

[0082] α1=a1cos(2πfn), α2=-a1sin(2πfn)

[0083] β1=a2cos(4πfn), β2=-a2sin(4πfn)

[0084] γ1=a3cos(6πfn), γ2=-a3sin(6πfn)

[0085] c1=[1, cos(2πf),…, cos(2πf(N-1))]T

[0086] s1=[1,sin(2πf),…,sin(2πf(N-1))] T

[0087] c2=[1, cos(4πf),…, cos(4πf(N-1))] T

[0088] s2=[1,sin(4πf),…,sin(4πf(N-1))] T

[0089] c3=[1, cos(6πf),…, cos(6πf(N-1)] T

[0090] s3=[1,sin(6πf),…,sin(6πf(N-1))] T

[0091] The solution to minimize formula (11) is:

[0092]

[0093] Among them, IH(H T H) -1 H T is an idempotent matrix, and I represents the identity matrix. Therefore, in order to estimate the breathing or heart rate f, the minimum value of the cost function J based on f should be calculated, that is, equivalently, the maximum value of the following expression should be obtained:

[0094] x T H(H T H) -1 H T x

[0095] When the normalized frequency of f is not near 0 or 1 / 2, the maximum likelihood MLE estimate of f can be obtained:

[0096]

[0097] S24. Construct an objective function for maximum likelihood estimation of respiratory or heart rate, and express the maximum likelihood estimation of respiratory or heart rate as an optimization problem of the objective function.

[0098] make Construct the objective function of maximum likelihood estimation of respiratory rate or heart rate f Then the maximum likelihood estimation (MLE) of the respiratory rate or heart rate f can be expressed as the following optimization problem about the objective function:

[0099]

[0100] Therefore, the maximum likelihood estimation of the respiratory rate and heart rate is obtained by searching for the frequency value that makes the objective function reach the maximum value in the respiratory rate and heart rate ranges respectively.

[0101] In order to provide sufficient signal length while maintaining the update rate of respiratory rate and heart rate estimation, the data are processed using an overlapping sliding window of duration 15 s, with an overlap of 14 s, so that the estimate of respiratory heart rate is updated every Δt=1 s.

[0102] S3. Solve the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

[0103] In this step, the maximum likelihood estimation of respiratory rate and heart rate is solved by Newton iteration method:

[0104]

[0105] in and The objective functions are The first and second partial derivatives of represents the estimated value of the respiratory rate or heart rate obtained at the kth iteration, and the estimated value obtained at the k+1th iteration is taken as the maximum likelihood estimate.

[0106] In this embodiment, the maximum likelihood estimation of the respiratory rate and heart rate shares the same formula (14), and the respiratory rate and heart rate are calculated separately, specifically by searching for the frequency value that makes the objective function reach the maximum value within the range of the respiratory rate and heart rate, respectively, as the maximum likelihood estimation values ​​of the respiratory rate and heart rate. Figure 7 The comparison between the respiratory rate estimation result of this embodiment and the reference sensor is shown; Figure 8 The comparison between the heart rate estimation result of this embodiment and that of the reference sensor is illustrated.

[0107] In addition, this embodiment also provides a non-contact vital sign estimation system based on FMCW millimeter wave radar, which specifically includes the following modules:

[0108] A preprocessing module, used for preprocessing the radar echo signal and extracting the chest wall displacement signal therefrom;

[0109] A model building module, used for modeling the chest wall displacement signal as a combination of the fundamental frequency and its harmonics, establishing a likelihood function about the respiratory rate and the heart rate, constructing an objective function for the maximum likelihood estimation of the respiratory or heart rate, and expressing the maximum likelihood estimation of the respiratory or heart rate as an optimization problem of the objective function;

[0110] The model solving module solves the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

[0111] The above modules of this embodiment are respectively used to implement corresponding steps of the non-contact vital sign estimation method. The detailed implementation process thereof refers to the above steps S1-S3.

[0112] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A non-contact vital sign estimation method based on FMCW millimeter wave radar, characterized in that: The following steps are involved: S1, preprocessing the radar echo signal to extract the chest wall displacement signal; S2. Modeling the chest wall displacement signal as a combination of the fundamental frequency of breathing and heartbeat and its harmonics, establishing likelihood functions for the breathing rate and heart rate, constructing an objective function for maximum likelihood estimation of the breathing or heartbeat frequency, and expressing the maximum likelihood estimation of the breathing or heartbeat frequency as an optimization problem of the objective function; S3. Solve the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

2. The non-contact vital sign estimation method according to claim 1, characterized in that: Step S1 includes: S11, mixing and filtering the radar echo signal to obtain a radar intermediate frequency signal; the radar intermediate frequency signal is sampled by an analog-to-digital converter; S12, using a phasor cancellation method to eliminate echo interference from stationary obstacles, and obtaining a radar intermediate frequency signal after the echo interference is eliminated; S13, performing distance-dimensional fast Fourier transform on the radar intermediate frequency signal after eliminating echo interference, and finding the peak value after the fast Fourier transform to determine the distance from the human target to the radar; S14, demodulating the signal at the range gate where the human target is located by using the inverse tangent and phase unwrapping method to obtain the phase signal of the range gate where the human target is located as the chest wall displacement signal caused by breathing and heartbeat; S15, using polynomial regression to remove motion trend processing on the chest wall displacement signal to eliminate strong low-frequency component interference in the frequency domain; S16. Perform slow-time phase correlation processing on the chest wall displacement signal after removing the motion trend.

3. The non-contact vital sign estimation method according to claim 2, characterized in that: In step S14, the phase signal of the range gate where the human target is located is extracted from the one-dimensional range spectrum diagram, and the extracted phase signal is expressed as: Among them B I (t) is the in-phase branch of the phase signal; B Q (t) is the orthogonal branch of the phase signal, A represents the amplitude of the phase signal, θ represents the fixed phase shift caused by the absolute distance between the human target and the radar, δ(t) represents the chest wall displacement caused by breathing and heartbeat, and φ is the total phase noise generated by the system; By using the orthogonal characteristics of the IQ signals, the phase function containing the chest displacement is obtained, thereby extracting the phase information: Phase unwrapping is performed by comparing the AD demodulation phase values ​​of adjacent sampling points: Where f(x) represents the phase information after unwrapping, Φ n+1 Indicates the phase value at the n+1th sampling moment, Φ n Indicates the phase value at the nth sampling moment.

4. The non-contact vital sign estimation method according to claim 2, characterized in that: When the slow time phase correlation processing is performed in step S16, the correlation between the chest wall displacement signals at the range gate where the human target is located and the chest wall displacement signals at the adjacent range gates is analyzed, and their Pearson correlation coefficient is calculated: The human target is located at the dth range gate, and the i-th range gate is adjacent to the d-th range gate. and are the phase signal vectors at the dth range gate and the ith range gate respectively, represents the covariance between the phase signal vectors; is the corresponding standard deviation.

5. The non-contact vital sign estimation method according to claim 1, characterized in that: Step S2 includes: S21, constructing a respiratory and heartbeat signal model containing harmonics to model the chest wall displacement signal as a combination of the fundamental frequency and its harmonics; using the respiratory and heartbeat signal model, solving within the frequency range of the respiratory and heartbeat signals respectively, capturing the harmonic components of the respiratory and heartbeat signals, and accumulating the harmonic energy and the fundamental frequency energy; The breathing and heartbeat signal model is expressed as: Where f is the breathing or heartbeat frequency, a1, a2 and a3 represent the amplitudes of the fundamental, second harmonic and third harmonic of breathing or heartbeat respectively. and are the phases of the fundamental wave, second harmonic, and third harmonic of breathing or heartbeat, respectively. ω[n] represents a mean of 0 and a variance of σ 2 Gaussian white noise, n represents the nth sampling moment; When f is the respiratory frequency, x[n] represents the respiratory signal model; when f is the heart rate, x[n] represents the heart rate signal model; S22. Based on the constructed breathing and heartbeat signal model, the probability density function of the variable to be estimated is obtained: in represents the variable to be estimated; x is the vector form of x[n]; N is the total number of radar intermediate frequency signals received; S23. Obtain an estimated value of the variable Θ to be estimated by maximizing the likelihood function; maximizing the likelihood function is equivalent to minimizing J(A, f, Φ): Where A=[a1,a2,a3], Simplify J(A,f,Φ) to: J(A,f,Φ)=(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3) T ×(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3) in: α=[α1,α2,β1,β2,γ1,γ2] T <h2 style=";text-align:left;direction:ltr">H = [c1,s1,c2,s2,c3,s3]<h2 style=";text-align:left;direction:ltr"> T α1=α1cos(2πfn),α2=-a1sin(2πfn) β1=a2cos(4πfn),β2=-a2sin(4πfn) γ1=a3cos(6πfn), γ2=-a3sin(6πfn) c1=[1,cos(2πf),…,cos(2πf(N-1))] T s1=[1,sin(2πf),…,sin(2πf(N-1))] T c2=[1,cos(4πf),…,cos(4πf(N-1))] T s2=[1,sin(4πf),…,sin(4πf(N-1))] T c3=[1,cos(6πf),…,cos(6πf(N-1))] T <h2 style=";text-align:left;direction:ltr">s3 = [1, sin(6πf),…, sin(6πf(N-1))]<h2 style=";text-align:left;direction:ltr"> T The solution that minimizes J(A,f,Φ) is: Among them, IH(H T H) -1 H T is an idempotent matrix, I represents the identity matrix; in order to estimate the breathing or heart rate f, the maximum value of the following expression should be obtained: x T H(H T H) -1 H T x When the normalized frequency of the breathing or heart rate f is not near 0 or 1 / 2, the maximum likelihood estimate of f is obtained: S24. Let the maximum likelihood estimate of the breathing or heart rate f be Constructing the objective function for maximum likelihood estimation of breathing or heart rate The maximum likelihood estimation of the breathing or heart rate is formulated as an optimization problem of the objective function.

6. The non-contact vital sign estimation method according to claim 5, characterized in that: Step S3 uses Newton iteration method to solve the maximum likelihood estimation of respiratory rate and heart rate: in and The objective functions are The first and second partial derivatives of represents the estimated value of the respiratory rate or heart rate obtained at the kth iteration, and the estimated value obtained at the k+1th iteration is taken as the maximum likelihood estimate.

7. A non-contact vital sign estimation system based on FMCW millimeter wave radar, characterized in that: Includes the following modules: A preprocessing module, used for preprocessing the radar echo signal and extracting the chest wall displacement signal therefrom; A model building module, used for modeling the chest wall displacement signal as a combination of the fundamental frequency and its harmonics, establishing a likelihood function about the respiratory rate and the heart rate, constructing an objective function for the maximum likelihood estimation of the respiratory or heart rate, and expressing the maximum likelihood estimation of the respiratory or heart rate as an optimization problem of the objective function; The model solving module solves the optimization problem of the objective function by Newton iteration method to obtain the maximum likelihood estimation values ​​of the respiratory rate and the heart rate.

8. The non-contact vital sign estimation system according to claim 7, characterized in that: Model building modules include: A respiratory and heartbeat signal model building module is used to build a respiratory and heartbeat signal model containing harmonics to model the chest wall displacement signal as a combination of the fundamental frequency and its harmonics; through the respiratory and heartbeat signal model, solving in the frequency range of respiratory and heartbeat respectively, capturing the harmonic components of the respiratory and heartbeat signals, and accumulating the harmonic energy and the fundamental frequency energy; The breathing and heartbeat signal model is expressed as: Where f is the breathing or heartbeat frequency, a1, a2 and a3 represent the amplitudes of the fundamental, second harmonic and third harmonic of breathing or heartbeat respectively. and are the phases of the fundamental wave, second harmonic, and third harmonic of breathing or heartbeat, respectively. ω[n] represents a mean of 0 and a variance of σ 2 Gaussian white noise, n represents the nth sampling moment; When f is the respiratory frequency, x[n] represents the respiratory signal model; when f is the heart rate, x[n] represents the heart rate signal model; The module for obtaining the variables to be estimated is used to obtain the probability density function of the variables to be estimated based on the constructed breathing and heartbeat signal model: in represents the variable to be estimated; x is the vector form of x[n]; N is the total number of radar intermediate frequency signals received; The maximum likelihood estimation module obtains the estimated value of the variable Θ by maximizing the likelihood function; maximizing the likelihood function is equivalent to minimizing J(A,f,Φ): Where A=[a1,a2,a3], Simplify J(A,f,Φ) to: J(A,f,Φ)=(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3) T ×(x-α1c1-α2s1-β1c2-β2s2-γ1c3-γ2s3) in: α=[α1,α2,β1,β2,γ1,γ2] T <h2 style=";text-align:left;direction:ltr">H = [c1,s1,c2,s2,c3,s3]<h2 style=";text-align:left;direction:ltr"> T α1=a1cos(2πfn),α2=-a1sin(2πfn) β1=a2cos(4πfn),β2=-a2sin(4πfn) γ1=a3cos(6πfn), γ2=-a3sin(6πfn) c1=[1,cos(2πf),…,cos(2πf(N-1))] T s1=[1,sin(2πf),…,sin(2πf(N-1))] T c2=[1,cos(4πf),…,cos(4πf(N-1))] T <h2 style=";text-align:left;direction:ltr">s2 = [1, sin(4π),…, sin(4π(N-1))]<h2 style=";text-align:left;direction:ltr"> T c3=[1,cos(6π),…,cos(6πf(N-1))] T s3=[1,sin(6π),…,sin(6πf(n-1))] T The solution that minimizes J(A,f,Φ) is: Among them, IH(H T H) -1 H T is an idempotent matrix, I represents the identity matrix; in order to estimate the breathing or heart rate f, the maximum value of the following expression should be obtained: x T H(H T H) -1 H T x When the normalized frequency of the breathing or heart rate f is not near 0 or 1 / 2, the maximum likelihood estimate of f is obtained: The objective function building block is to let the maximum likelihood estimate of the breathing or heart rate f be Constructing the objective function for maximum likelihood estimation of breathing or heart rate The maximum likelihood estimation of the breathing or heart rate is formulated as an optimization problem of the objective function.

9. The non-contact vital sign estimation system according to claim 8, characterized in that: The model solving module uses the Newton iteration method to solve the maximum likelihood estimates of the respiratory rate and heart rate: in and The objective functions are The first and second partial derivatives of represents the estimated value of the respiratory rate or heart rate obtained at the kth iteration, and the estimated value obtained at the k+1th iteration is taken as the maximum likelihood estimate.

10. The non-contact vital sign estimation system according to claim 7, characterized in that: The process of preprocessing radar echo signals by the preprocessing module includes: The radar echo signal is mixed and filtered to obtain a radar intermediate frequency signal; the radar intermediate frequency signal is sampled by an analog-to-digital converter; The echo interference of the stationary obstacle is eliminated by using the phase cancellation method, and the radar intermediate frequency signal after the echo interference is eliminated is obtained; Perform distance-dimensional fast Fourier transform on the radar intermediate frequency signal after eliminating echo interference, and find the peak value after fast Fourier transform to determine the distance from the human target to the radar; The determined distance from the human target to the radar is demodulated by using the inverse tangent and phase unwrapping methods to obtain the phase signal of the range gate where the human target is located as the chest wall displacement signal caused by breathing and heartbeat; Polynomial regression is used to remove the motion trend of the chest wall displacement signal to eliminate strong low-frequency component interference in the frequency domain; The chest wall displacement signal after removing the motion trend was processed by slow time phase correlation.

Citation Information

Patent Citations

  • Respiratory signal detection method and device

    CN105640502A

  • Vital sign signal feature extraction method and device

    CN117349645A

  • Millimeter wave radar-based non-contact real-time vital sign monitoring system and method

    WO2022104868A1

Cited By

  • Respiration rate and heart rate measuring method based on millimeter wave radar amplitude phase information fusion

    CN120959708A