Heart rate estimation method and device based on millimeter wave radar multichannel combination

Abstract

The invention belongs to the technical field of millimeter wave radar non-contact vital sign monitoring, and particularly relates to a millimeter wave radar multi-channel combination-based heart rate estimation method and device, and the specific process is as follows: firstly, a millimeter wave radar based on a multi-channel frequency modulation continuous wave (FMCW) synchronously collects chest echo signals; extracting a distance bin containing a life signal from the echo signal through distance FFT, and performing preprocessing to obtain a multi-channel phase signal; secondly, performing modal decomposition by constructing a multi-channel signal joint optimization objective function and combining multi-channel phase signals, and adaptively determining an initial center frequency of modal decomposition according to a peak value of multi-channel signal spectrum fusion and a peak value with large energy; performing iterative updating based on the target function to obtain an optimal heart rate modal sequence; and finally, screening a final heart rate mode from the heart rate mode sequence according to an energy maximum principle, and outputting the center frequency of the screened heart rate mode as a heart rate value to realize heart rate estimation.

Description

Technical Field

[0001] This invention belongs to the field of millimeter-wave radar non-contact vital sign monitoring technology, specifically relating to a heart rate estimation method and device based on multi-channel joint millimeter-wave radar. Background Technology

[0002] Heart rate (HR) is a core physiological indicator for assessing cardiovascular health, and its long-term continuous monitoring has significant clinical value for early diagnosis of cardiovascular diseases, postoperative rehabilitation assessment, and daily health management. Traditional heart rate monitoring relies on contact devices (such as photoelectric pulse wave (PPG) sensors and electrocardiogram (ECG) devices), which have problems such as poor comfort and weak environmental adaptability. When patients have burns, infectious skin diseases, or need to be bedridden for a long time, contact sensors are prone to causing discomfort or cross-infection.

[0003] With the development of wireless sensing technology, millimeter-wave radar based on frequency modulated continuous wave (FMCW) has become the preferred solution for non-contact vital sign monitoring due to its small size, low power consumption, and non-contact characteristics. Its principle is to transmit FMCW signals and receive the reflected echoes of human chest cavity vibrations (caused by heartbeat and respiration), extracting periodic heartbeat information from the echo signals to achieve heart rate estimation. However, existing millimeter-wave radar heart rate estimation technology has the following key problems: the amplitude of human chest cavity heartbeat vibrations is extremely small (typically only 0.1-1 mm), resulting in a very low signal-to-noise ratio (SNR) in single-channel radar echo signals. After being superimposed with environmental noise and respiratory signal interference, the heart rate estimation error is significant, making it difficult to meet clinical accuracy requirements; traditional single-channel variational mode decomposition (VMD) relies on manually setting the initial center frequency, which is prone to confusion between heartbeat modes and respiratory or noise modes due to frequency initialization deviations, further reducing estimation accuracy. Summary of the Invention

[0004] To address the above problems, this invention proposes a heart rate estimation method and device based on multi-channel joint millimeter-wave radar, which can solve the problems of low signal-to-noise ratio and initial frequency deviation in mode decomposition of existing single-channel millimeter-wave radar.

[0005] The technical solution for implementing the present invention is as follows:

[0006] In a first aspect, the present invention provides a heart rate estimation method based on multi-channel joint millimeter-wave radar, the specific process of which is as follows:

[0007] First, millimeter-wave radar based on multi-channel frequency modulated continuous wave (FMCW) synchronously acquires chest echo signals. Range FFT is used to extract range cells containing vital signs from the echo signals, and preprocessing is performed to obtain multi-channel phase signals.

[0008] Secondly, by constructing a joint optimization objective function for multi-channel signals, mode decomposition is performed by combining multi-channel phase signals. The initial center frequency of mode decomposition is adaptively determined based on the peak value and high-energy peak value of the multi-channel signal spectrum fusion. The optimal heart rate mode sequence is obtained by iterative updating based on the objective function.

[0009] Finally, based on the principle of maximizing energy, the final heart rate mode is selected from the heart rate mode sequence, and the center frequency of the selected heart rate mode is output as the heart rate value to achieve heart rate estimation.

[0010] Optionally, the preprocessing described in this invention is as follows: performing phase extraction, phase unwinding, and phase difference on the signal within the distance chamber. If the signal after phase difference has a phase outlier jump, then the median is used to replace the outlier within the window for correction.

[0011] Optionally, the present invention decomposes the phase signal of each channel into K band-limited modes u k,b (t), the multi-channel joint optimization objective function is constructed as follows:

[0012]

[0013] Where B is the number of channels, K is the number of modes in the decomposition, and α is the penalty factor. It is for signal u k,b (t) is subjected to Hilbert transform, where Ω is the center frequency ω. k The value space of u, where u is u k,b The space of values ​​for (t), y b (t) is the phase signal of the b-th channel.

[0014] Optionally, the present invention obtains the optimal heart rate modality sequence by iteratively updating based on the objective function; the specific process is as follows:

[0015] Heart rate modal update: For each modality k = 1, 2…K and each channel b = 1, 2…B, update in the Fourier domain based on the optimization objective function:

[0016]

[0017] in, The input signal is in frequency domain form. In the frequency domain form of the modes, It is the frequency domain form of the Lagrange multiplier; ω is the frequency;

[0018] Center frequency update: The common center frequency is updated based on the centroid of the power spectrum of all channels.

[0019]

[0020] Iterate through the above update process until the convergence condition is met.

[0021] Optionally, the convergence condition of this invention is: the difference between the modal functions of two adjacent iterations is less than a set threshold ∈, and the final iteration result of the k-th mode and the b-th channel is output.

[0022]

[0023] Secondly, the present invention provides a heart rate estimation device based on multi-channel joint millimeter-wave radar, comprising: a signal processing module, a heart rate mode optimization module, and a heart rate estimation module; wherein...

[0024] The signal processing module synchronously acquires chest echo signals based on millimeter-wave radar using multi-channel frequency modulated continuous wave (FMCW), extracts range cells containing vital signs from the echo signals using range FFT, and performs preprocessing to obtain multi-channel phase signals.

[0025] The heart rate mode optimization module is used to construct a joint optimization objective function for multi-channel signals, perform mode decomposition on multi-channel phase signals, and adaptively determine the initial center frequency of mode decomposition based on the peak values ​​and high-energy spikes of the multi-channel signal spectrum fusion; and iteratively update the objective function to obtain the optimal heart rate mode sequence.

[0026] The heart rate estimation module is used to select the final heart rate mode from the heart rate mode sequence according to the principle of maximum energy, and output the center frequency of the selected heart rate mode as the heart rate value to realize heart rate estimation.

[0027] Beneficial effects:

[0028] Compared with existing millimeter-wave radar heart rate estimation technology, the significant advantages of this invention are:

[0029] First, this invention constructs a multi-channel joint optimization objective function through MVMD, achieving cross-channel information coordination during the decomposition stage, effectively suppressing weak channel noise transmission, and improving the signal-to-noise ratio of the heartbeat echo signal.

[0030] Second, the present invention adaptively determines the initial center frequency of MVMD by selecting the peak values ​​of the multi-channel fused spectrum and the high-energy peak values, thus avoiding the problem of inaccurate mode decomposition caused by manual setting. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0033] Figure 2 This is a schematic diagram of the radar system in the method of the present invention;

[0034] Figure 3 This is a schematic diagram of phase difference outlier processing in the method of the present invention;

[0035] Figure 4 This is a schematic diagram of multi-channel spectrum fusion in the method of the present invention;

[0036] Figure 5 This is a schematic diagram of the experimental scenario in the method of the present invention. Detailed Implementation

[0037] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0038] It should be noted that, in the absence of conflict, the following embodiments and features can be combined with each other; and, based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0039] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0040] This application provides an embodiment of a heart rate estimation method based on multi-channel joint millimeter-wave radar, the specific process of which is as follows:

[0041] First, millimeter-wave radar based on multi-channel frequency modulated continuous wave (FMCW) synchronously acquires chest echo signals. Range FFT is used to extract range cells containing vital signs from the echo signals, and preprocessing is performed to obtain multi-channel phase signals.

[0042] Secondly, by constructing a joint optimization objective function for multi-channel signals, mode decomposition is performed by combining multi-channel phase signals. The initial center frequency of mode decomposition is adaptively determined based on the peak value and high-energy peak value of the multi-channel signal spectrum fusion. The optimal heart rate mode sequence is obtained by iterative updating based on the objective function.

[0043] Finally, based on the principle of maximizing energy, the final heart rate mode is selected from the heart rate mode sequence, and the center frequency of the selected heart rate mode is output as the heart rate value to achieve heart rate estimation.

[0044] The above process will be explained in detail below, and the flowchart is as follows: Figure 1 As shown, it includes the following steps:

[0045] Step 1: Acquire and preprocess radar echo data

[0046] The FMCW radar acquired raw data of human chest echoes from a 1-transmit, 4-receive system. The radar system's transmitted signal diagram is shown below. Figure 2 As shown. A Fast Fourier Transform (FFT) is performed on the original data, and range bins containing heartbeat signals are selected based on the principle of maximizing energy. Phase extraction, phase unwrapping, and phase differencing are then performed on the signals within the range bins. If the phase-differentiated signal exhibits outlier abrupt phase jumps, the median is used to replace the outliers within the window for correction. Figure 3 As shown, a sliding smoothing filter is applied to the signal after processing outlier jumps.

[0047] Step 2: Perform mode decomposition optimization using multi-channel signals.

[0048] 1. Construct a multi-channel joint optimization objective function

[0049] Let the preprocessed multi-channel phase signal vector be... It is caused by respiratory signals s r (t), heartbeat signal s h It consists of (t) and noise n(t), that is:

[0050] y(t)=s r (t)+s h (t)+n(t)

[0051] Construct an MVMD variational optimization problem, decomposing the phase signal of each channel in y(t) into K band-limited modes u. k,b (t), k = 1, 2…K, b = 1, 2…B, each mode has the same center frequency ω in the multi-channel. k The amplitudes can be different. The multi-channel signal is jointly optimized to minimize the sum of the bandwidths of all channels and all modes, and the sum of the modes of each variable is required to be equal to the original signal of the corresponding channel.

[0052] The objective function to be optimized is:

[0053]

[0054] Where B is the number of channels, K is the number of decomposed modes, and α is a penalty factor used to adjust the weight between bandwidth constraints and reconstruction error. It is for signal u k,b (t) is subjected to Hilbert transform to construct its analytic signal, where Ω is the center frequency ω. k The value space of u, where u is u k,b The space of values ​​for (t), y b (t) is the original input signal of the b-th channel.

[0055] 2. Adaptively determine the initial center frequency of mode decomposition.

[0056] The spectra of the preprocessed multi-channel signals are superimposed to obtain a fused spectrum. Within the heart rate range (0.8-2Hz), K peak points and high-energy spikes from the fused spectrum are selected as the initial center frequency ω for mode decomposition. k ,like Figure 4 As shown.

[0057] 3. Perform MVMD iterative solution.

[0058] The heart rate mode u is updated iteratively using the alternating direction of multipliers (ADMM) method. k Center frequency ω k The process continues until convergence, specifically as follows:

[0059] First, modal updates are performed. For each mode k = 1, 2…K and each channel b = 1, 2…B, updates are performed in the Fourier domain based on the optimization objective function:

[0060]

[0061] in, The frequency domain form of the original input signal of the b-th channel. It is the frequency domain form of the Lagrange multiplier, used to constrain the reconstruction error of the original signal and the sum of all modal components.

[0062] Secondly, the center frequency is updated by updating the common center frequency based on the centroid of the power spectrum of all channels:

[0063]

[0064] Iterate until the convergence condition is met: the difference in mode functions between two adjacent iterations is less than the threshold ∈ = 10. -6 Output the final iteration result of the k-th mode and the b-th channel.

[0065]

[0066] Step 3: Heartbeat mode screening and heart rate value estimation

[0067] Modes that typically contain heartbeat signals have high energy. Within the heart rate range, heart rate modes are selected based on the principle of maximizing energy, and the center frequency value of the heart rate mode is finally output as the heart rate value.

[0068] Another embodiment of this application discloses a heart rate estimation device based on multi-channel joint millimeter-wave radar, comprising: a signal processing module, a heart rate mode optimization module, and a heart rate estimation module; wherein...

[0069] The signal processing module synchronously acquires chest echo signals based on millimeter-wave radar using multi-channel frequency modulated continuous wave (FMCW), extracts range cells containing vital signs from the echo signals using range FFT, and performs preprocessing to obtain multi-channel phase signals.

[0070] The heart rate mode optimization module is used to construct a joint optimization objective function for multi-channel signals, perform mode decomposition on multi-channel phase signals, and adaptively determine the initial center frequency of mode decomposition based on the peak values ​​and high-energy spikes of the multi-channel signal spectrum fusion; and iteratively update the objective function to obtain the optimal heart rate mode sequence.

[0071] The heart rate estimation module is used to select the final heart rate mode from the heart rate mode sequence according to the principle of maximum energy, and output the center frequency of the selected heart rate mode as the heart rate value to realize heart rate estimation.

[0072] The following experiment will verify this.

[0073] To verify the proposed multi-channel combined heart rate estimation method based on millimeter-wave radar, a field experiment was designed and analyzed. A 77GHz millimeter-wave radar, AWR1642, was used in the experiment, employing one Tx antenna and four Rx antennas. The parameters corresponding to the FMCW radar settings are listed in Table 1. The true heart rate values ​​were measured using a medical five-lead multi-parameter monitor. The experimental scenario is as follows. Figure 5 As shown, the radar was located at the head of the bed, at a horizontal distance of 0.55 / 0.65 / 0.75m from the chest cavity and at a vertical height of 0.6 / 0.7m. It collected 12 sets of heart rate data from 2 different people, with each set of data collected for 5 minutes.

[0074] Table 1

[0075]

[0076] To verify the performance of the proposed method, it was compared with four single-channel methods that performed VMD decomposition to estimate heart rate. Experimental results show that in 90% or more of the results, the absolute error of the proposed algorithm for heart rate estimation is 7.2 BPM; the absolute error of the traditional single-channel VMD decomposition is greater than or equal to 14.9 BPM. The experimental results are shown in Table 2.

[0077] Table 2

[0078]

[0079] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A heart rate estimation method based on multi-channel joint millimeter-wave radar, characterized in that, The specific process is as follows: First, millimeter-wave radar based on multi-channel frequency modulated continuous wave (FMCW) synchronously acquires chest echo signals. Range FFT is used to extract range cells containing vital signs from the echo signals, and preprocessing is performed to obtain multi-channel phase signals. Secondly, by constructing a joint optimization objective function for multi-channel signals, mode decomposition is performed by combining multi-channel phase signals, and the initial center frequency of mode decomposition is adaptively determined based on the peak value and high-energy peak value of the multi-channel signal spectrum fusion. The optimal heart rate modality sequence is obtained by iteratively updating based on the objective function. Finally, based on the principle of maximizing energy, the final heart rate mode is selected from the heart rate mode sequence, and the center frequency of the selected heart rate mode is output as the heart rate value to achieve heart rate estimation.

2. The heart rate estimation method based on millimeter-wave radar multi-channel joint according to claim 1, characterized in that, The preprocessing involves: extracting the phase of the signal within the distance chamber, unwinding the phase, and performing phase differentiation. If the signal after phase differentiation has abnormal phase jumps, the median is used to replace the abnormal values ​​within the window for correction.

3. The heart rate estimation method based on millimeter-wave radar multi-channel joint according to claim 1, characterized in that, The phase signal of each channel is decomposed into K band-limited modes u k,b (t), the multi-channel joint optimization objective function is constructed as follows: Where B is the number of channels, K is the number of modes in the decomposition, and α is the penalty factor. It is for signal u k,b (t) is subjected to Hilbert transform, where Ω is the center frequency ω. k The space of values, is u k,b The space of values ​​for (t), y b (t) is the phase signal of the b-th channel.

4. The heart rate estimation method based on millimeter-wave radar multi-channel joint according to claim 3, characterized in that, The optimal heart rate modality sequence is obtained by iterative updating based on the objective function; the specific process is as follows: Heart rate modal update: For each modality k = 1, 2…K and each channel b = 1, 2…B, update in the Fourier domain based on the optimization objective function: in, The input signal is in frequency domain form. In the frequency domain form of the modes, It is the frequency domain form of the Lagrange multiplier; ω is the frequency; Center frequency update: The common center frequency is updated based on the centroid of the power spectrum of all channels. Iterate through the above update process until the convergence condition is met.

5. The heart rate estimation method based on millimeter-wave radar multi-channel joint according to claim 4, characterized in that, The convergence condition is: the difference between the modal functions of two adjacent iterations is less than a set threshold ∈, and the final iteration result of the k-th mode and the b-th channel is output.

6. A heart rate estimation device based on millimeter-wave radar multi-channel joint operation, characterized in that, include: Signal processing module, heart rate mode optimization module, and heart rate estimation module; in The signal processing module synchronously acquires chest echo signals based on millimeter-wave radar using multi-channel frequency modulated continuous wave (FMCW), extracts range cells containing vital signs from the echo signals using range FFT, and performs preprocessing to obtain multi-channel phase signals. The heart rate mode optimization module is used to construct a joint optimization objective function for multi-channel signals, perform mode decomposition on multi-channel phase signals, and adaptively determine the initial center frequency of mode decomposition based on the peak value and high-energy peak value of the multi-channel signal spectrum fusion. The optimal heart rate modality sequence is obtained by iteratively updating based on the objective function. The heart rate estimation module is used to select the final heart rate mode from the heart rate mode sequence according to the principle of maximum energy, and output the center frequency of the selected heart rate mode as the heart rate value to realize heart rate estimation.

7. The heart rate estimation device based on millimeter-wave radar multi-channel joint according to claim 6, characterized in that, The preprocessing involves: extracting the phase of the signal within the distance chamber, unwinding the phase, and performing phase differentiation. If the signal after phase differentiation has abnormal phase jumps, the median is used to replace the abnormal values ​​within the window for correction.

8. The heart rate estimation device based on millimeter-wave radar multi-channel joint according to claim 6, characterized in that, The phase signal of each channel is decomposed into K band-limited modes u k,b (t), the multi-channel joint optimization objective function is constructed as follows: Where B is the number of channels, K is the number of modes in the decomposition, and α is the penalty factor. It is for signal u k,b (t) is subjected to Hilbert transform, where Ω is the center frequency ω. k The space of values, is u k,b The space of values ​​for (t), y b (t) is the phase signal of the b-th channel.

9. The heart rate estimation device based on millimeter-wave radar multi-channel joint according to claim 8, characterized in that, The optimal heart rate modality sequence is obtained by iterative updating based on the objective function; the specific process is as follows: Heart rate modal update: For each modality k = 1, 2…K and each channel b = 1, 2…B, update in the Fourier domain based on the optimization objective function: in, The input signal is in frequency domain form. In the frequency domain form of the modes, It is the frequency domain form of the Lagrange multiplier; ω is the frequency; Center frequency update: The common center frequency is updated based on the centroid of the power spectrum of all channels. Iterate through the above update process until the convergence condition is met.

10. The heart rate estimation device based on millimeter-wave radar multi-channel joint according to claim 9, characterized in that, The convergence condition is: the difference between the modal functions of two adjacent iterations is less than a set threshold ∈, and the final iteration result of the k-th mode and the b-th channel is output.

Images