Respiratory heart rate monitoring method based on variational mode decomposition and electronic device

Abstract

This application provides a respiratory and heart rate monitoring method and electronic device based on variational mode decomposition, applicable to the field of vital sign monitoring technology. The method includes: processing acquired raw data to obtain a raw phase sequence; the raw data is obtained from the echo signal returned by the detection signal of a non-contact detection module through the target area of ​​the target object; performing variational mode decomposition based on the raw phase sequence to obtain multiple target intrinsic mode functions; determining respiratory data and / or heart rate data based on the multiple target intrinsic mode functions; displaying the respiratory data and / or heart rate data, and / or sending the respiratory data and / or heart rate data to a terminal device for display. The embodiments of this application achieve blind separation of respiratory and heart rate signals, applicable to high-precision non-contact monitoring of infant respiration and heart rate.

Description

Technical Field

[0001] This application relates to the field of vital sign monitoring technology, and in particular to a respiratory and heart rate monitoring method and electronic device based on variational mode decomposition. Background Technology

[0002] Currently, ECG electrodes, chest strap respiratory sensors, and wearable pulse oximeters are typically attached to the infant's body surface to collect ECG (electrocardiogram) and chest impedance signals. After amplification and analog-to-digital conversion, the heart rate is calculated through threshold filtering and R-wave peak detection, and the respiratory rate is calculated through the chest wall movement cycle.

[0003] An infant's respiratory rate can reach up to 60 breaths per minute (1.0 Hz). Its second harmonic (2.0 Hz) and third harmonic (3.0 Hz) completely overlap with the infant's heart rate frequency band (1.2-3.0 Hz). Existing fixed filtering methods cannot achieve effective blind separation, and respiratory harmonics are easily misidentified as heart rate signals. Summary of the Invention

[0004] In view of this, embodiments of this application provide a respiratory and heart rate monitoring method and electronic device based on variational mode decomposition to solve the problem that the prior art cannot effectively separate respiratory signals and heart rate signals.

[0005] A first aspect of this application provides a respiratory heart rate monitoring method based on variational mode decomposition, comprising: The acquired raw data is processed to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal. Based on the original phase sequence, variational mode decomposition is performed to obtain multiple target intrinsic mode functions; Based on multiple target intrinsic mode functions, determine respiratory data and / or heart rate data; Display respiratory data and / or heart rate data, and / or send respiratory data and / or heart rate data to a terminal device for display of respiratory data and / or heart rate data on the terminal device.

[0006] This application embodiment can acquire raw data through a non-contact detection module, then process the raw data to obtain a raw phase sequence, and then perform variational mode decomposition based on the raw phase sequence to obtain multiple target intrinsic mode functions. Based on the multiple target intrinsic mode functions, respiratory data and / or heart rate data are determined. That is, the variational mode decomposition method can be used to achieve blind separation of respiratory data and heart rate data, and then the respiratory data and / or heart rate data can be displayed to caregivers, so that caregivers can understand the vital signs of the target object in a timely manner. It is suitable for high-precision non-contact monitoring of infant breathing and heart rate.

[0007] In one possible implementation, respiratory data and / or heart rate data are determined based on multiple target intrinsic mode functions, including: Multiple target intrinsic mode functions are filtered and separated to reconstruct respiratory signals and / or heart rate signals; Based on respiratory signals and / or heart rate signals, corresponding respiratory data and / or heart rate data are determined.

[0008] This application embodiment achieves blind separation of respiratory and / or heart rate signals by screening, separating and reconstructing multiple target intrinsic mode functions. Then, based on the respiratory and / or heart rate signals, corresponding respiratory and / or heart rate data are determined, thus achieving effective blind separation of respiratory and heart rate data.

[0009] In one possible implementation, variational mode decomposition is performed based on the original phase sequence to obtain multiple target intrinsic mode functions, including: The original phase sequence is subjected to time-series differential calculation to obtain the differential phase signal; Variational mode decomposition is performed on the differential phase signal to obtain multiple target intrinsic mode functions.

[0010] After the original phase sequence in this embodiment of the application is subjected to time differential, the signal becomes stable and the energy is concentrated on the two main peaks of respiration and heart rate. Then, variational mode decomposition is performed, which is beneficial for the subsequent clean and efficient screening and separation of respiratory and heart rate signals.

[0011] In one possible implementation, variational mode decomposition is performed on the differential phase signal to obtain multiple target intrinsic mode functions, including: The target variational mode decomposition algorithm is used to perform variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions; The objective variational mode decomposition algorithm is obtained as follows: An initial variational mode decomposition algorithm is used to perform variational mode decomposition on the differential phase signal to obtain multiple candidate intrinsic mode functions; The Grey Wolf optimization algorithm is used to adaptively determine the number of modes and the penalty factor of the initial variational mode decomposition algorithm until the total envelope entropy of multiple candidate intrinsic mode functions is minimized, thus obtaining the target variational mode decomposition algorithm.

[0012] This application employs the Grey Wolf optimization algorithm to adaptively determine variational mode decomposition (MODE) parameters, solving the over- or under-decomposition problems caused by manually setting parameters in existing technologies, and achieving adaptive optimization of variational mode decomposition parameters. This application uses the optimized variational mode decomposition parameters to perform variational mode decomposition on the differential phase signal, resulting in more accurate multiple target intrinsic mode functions.

[0013] In one possible implementation, multiple target intrinsic mode functions are filtered and separated to reconstruct respiratory signals and / or heart rate signals, including: Determine the center frequency of the intrinsic mode function of each target; Based on the center frequency of each target intrinsic mode function, the intrinsic mode functions of each target are screened and separated to reconstruct respiratory signals and / or heart rate signals.

[0014] The embodiments of this application can filter and separate the target intrinsic mode functions by the center frequency of each target intrinsic mode function. This not only solves the problem of individual differences, but also ensures the absolute accuracy of the reconstructed respiratory and heart rate signals on the time axis, providing a high-fidelity data source for further data processing of respiratory and heart rate signals.

[0015] In one possible implementation, based on the center frequency of each target intrinsic mode function, the target intrinsic mode functions are filtered and separated to reconstruct the respiratory signal and / or heart rate signal, including: The respiratory signal is obtained by summing and reconstructing multiple target intrinsic mode functions corresponding to the center frequency within the first frequency range; Multiple target intrinsic mode functions corresponding to the center frequency in the second frequency range are determined as multiple candidate intrinsic mode functions; the frequency in the second frequency range is greater than the frequency in the first frequency range; Determine the mutual information between each candidate intrinsic mode function and the respiratory signal; The heart rate signal is obtained by accumulating and reconstructing candidate intrinsic mode functions whose mutual information is below the first threshold.

[0016] In this embodiment, multiple target intrinsic mode functions are screened and separated based on frequency range and mutual information constraints. Candidate intrinsic mode functions with mutual information below a first threshold are accumulated and reconstructed into a heart rate signal, completely eliminating respiratory harmonic components and achieving accurate separation of respiratory and heart rate signals.

[0017] In one possible implementation, respiratory data and / or heart rate data are determined based on respiratory signals and / or heart rate signals, including: Filter the respiratory signal and / or heart rate signal to obtain the respiratory time-domain waveform and / or heart rate time-domain waveform; Based on the respiratory time-domain waveform and / or heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension, respectively. If the difference between two candidate respiratory rates is less than the first preset error, then the average of the two candidate respiratory rates is taken as the respiratory rate. If the difference between two candidate heart rates is less than the second preset error, then the average of the two candidate heart rates will be used as the heart rate.

[0018] The embodiments of this application can determine respiratory rate and heart rate by calculating parameters from both the frequency domain and time domain dimensions, thereby improving the accuracy of respiratory rate and heart rate calculation and avoiding misjudgments from single-dimensional detection.

[0019] In one possible implementation, based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension, respectively, including: Based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, spectral peak search is performed in the physiological frequency bands corresponding to the respiratory time-domain waveform and the physiological frequency bands corresponding to the heart rate time-domain waveform, respectively, to determine the first candidate respiratory rate and / or the first candidate heart rate. Normalize the respiratory time-domain waveform and / or heart rate time-domain waveform, and determine the candidate respiratory peak and / or the candidate heart rate peak of the normalized respiratory time-domain waveform. If the candidate respiratory peak value is greater than the first preset peak value, then the candidate respiratory peak value is determined to be a valid respiratory peak value; If the candidate heart rate peak value is greater than the second preset peak value, then the candidate heart rate peak value is determined to be the effective heart rate peak value; A second candidate respiratory rate and / or a second candidate heart rate are determined based on the time interval between adjacent effective respiratory peaks and / or the time interval between adjacent effective heart rate peaks.

[0020] This application embodiment can use spectral peak search to find the highest frequency point of the physiological frequency band to determine the first candidate respiratory rate and / or the first candidate heart rate. At the same time, this application embodiment can use normalization and peak threshold to determine the second candidate respiratory rate and / or the second candidate heart rate, combining the frequency domain dimension and the time domain dimension, which facilitates the calculation and cross-validation of multi-dimensional physiological parameters.

[0021] In one possible implementation, the acquired raw data is processed to obtain the raw phase sequence, including: The raw data is preprocessed to obtain the target complex matrix; the target complex matrix is ​​a two-dimensional complex matrix consisting of the total number of chirps and the total number of distance cells in a single frame. For the target complex matrix, extract the complex number corresponding to each Chirp under the target range cell, and decompose the extracted complex number into in-phase time components and quadrature time components; Based on in-phase and quadrature timing components, arctangent operation is used to solve for each Chirp to obtain the original phase sequence.

[0022] This application embodiment can obtain the target complex matrix by preprocessing the original data, and then extract the complex number corresponding to each chirp of the target range cell from the two-dimensional target complex matrix, compressing the two-dimensional matrix into a one-dimensional time vector. By accurately locating the target range cell, the target can be locked and tracked, effectively shielding static clutter from other ranges. The extracted complex number is then split into in-phase time components and orthogonal time components of the current frame, which facilitates the solution of the original phase sequence. While maximizing the preservation of micro-motion phase details, the computational complexity is reduced to a minimum.

[0023] In one possible implementation, the original data undergoes data preprocessing to obtain the target complex matrix, including: The raw data is preprocessed to generate a target three-dimensional echo matrix. The target three-dimensional echo matrix is ​​a three-dimensional complex matrix consisting of the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp. The non-contact detection module includes multiple antennas. The target three-dimensional echo matrix is ​​processed by the antenna's weighted coherent accumulation to obtain the first two-dimensional complex matrix; the first two-dimensional complex matrix is ​​a complex matrix of the total number of chirps and the total number of range cells in a single frame; The first two-dimensional complex matrix is ​​subjected to two-stage clutter suppression processing to obtain the target complex matrix.

[0024] The embodiments of this application can enhance the signal-to-noise ratio of weak, micro-motion signals at the 0.05mm level in infants by using multi-antenna weighted coherent accumulation and two-stage adaptive clutter suppression preprocessing, and effectively suppress static and dynamic clutter interference.

[0025] In one possible implementation, the raw data is preprocessed to generate the target three-dimensional echo matrix, including: The raw data is parsed according to the frame timing rules to generate the first three-dimensional echo matrix; the first three-dimensional echo matrix is ​​a three-dimensional complex matrix with the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp; The first three-dimensional echo matrix is ​​normalized to obtain the second three-dimensional echo matrix; The target three-dimensional echo matrix is ​​obtained by performing a Hamming windowing operation on each Chirp of each antenna on the second three-dimensional echo matrix.

[0026] The embodiments of this application can convert the original data into a first three-dimensional echo matrix, and then perform quantization and normalization processing on the first three-dimensional echo matrix to obtain a second three-dimensional echo matrix. This facilitates the performance of Hamming windowing operation on each chirp data of each receiving antenna to suppress spectral leakage in subsequent range dimension transformations.

[0027] In one possible implementation, the target three-dimensional echo matrix is ​​subjected to antenna-weighted coherent accumulation processing to obtain a first two-dimensional complex matrix, including: Perform a range-dimensional transformation on the target's three-dimensional echo matrix for each antenna to obtain the range-dimensional complex spectrum of each antenna; Based on the complex spectrum of each distance dimension, the echo signal-to-noise ratio of each antenna is determined; The weight of each antenna is determined based on the echo signal-to-noise ratio of each antenna; The first two-dimensional complex matrix is ​​determined based on the distance dimension complex spectrum of each antenna and the weight of each antenna.

[0028] In this embodiment, the distance dimension complex spectrum of each antenna is obtained by performing a distance dimension transformation on each antenna, and the signal-to-noise ratio of the echo of a single receiving antenna is calculated. The signal-to-noise ratio of each antenna is used as a weighting coefficient to coherently accumulate the complex signals of the same distance unit. In this way, the antenna with the higher signal-to-noise ratio is assigned a larger weight, the receiving antenna dimension is eliminated, and the first two-dimensional complex matrix can maximize the gain of the weak chest cavity micro-motion signal.

[0029] In one possible implementation, the first two-dimensional complex matrix undergoes two-stage clutter suppression processing to obtain the target complex matrix, including: Subtract a preset complex number from each complex number in the first two-dimensional complex matrix to obtain the second two-dimensional complex matrix; the preset complex number is obtained by averaging the complex numbers corresponding to each distance unit in the first two-dimensional complex matrix for a preset number of frames. The target complex matrix is ​​obtained by filtering the second two-dimensional complex matrix using a least mean square adaptive filter.

[0030] This embodiment of the application achieves first-level static clutter cancellation by subtracting a preset complex number from each complex number in the first two-dimensional complex matrix, which can remove static reflection signals with a fixed background. Then, this embodiment of the application uses a minimum mean square adaptive filter to filter the second two-dimensional complex matrix to achieve variable step size minimum mean square adaptive filtering, which can track and suppress non-static clutter caused by bed vibration and airflow disturbance in real time to obtain an accurate target complex matrix.

[0031] In one possible implementation, after determining the respiratory data and / or heart rate data based on multiple target intrinsic mode functions, the following is also included: The respiratory abnormality type is determined based on respiratory data, and an early warning message is generated based on the respiratory abnormality type; and / or, the heart rate abnormality type is determined based on heart rate data, and an early warning message is generated based on the heart rate abnormality type. Display warning information, and / or send warning information to terminal devices for display on the terminal devices.

[0032] The embodiments of this application can provide timely warnings and display warning information to remind caregivers when the breathing and / or heartbeat of the target object are abnormal. It can provide real-time warnings for the unique physiological abnormalities of infants and accurately identify high-risk symptoms such as apnea, and is suitable for daily home monitoring.

[0033] A second aspect of this application provides a respiratory heart rate monitoring device based on variational mode decomposition, comprising: The first data processing module is used to process the acquired raw data to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal. The second data processing module is used to perform variational mode decomposition based on the original phase sequence to obtain multiple target intrinsic mode functions; The respiratory and heart rate determination module is used to determine respiratory data and / or heart rate data based on multiple target intrinsic mode functions; The display processing module is used to display respiratory data and / or heart rate data, and / or send respiratory data and / or heart rate data to a terminal device for display of respiratory data and / or heart rate data on the terminal device.

[0034] A third aspect of this application provides an electronic device, including: Memory, used to store computer programs; A processor is used to implement the steps of the method as described in the first aspect when executing a computer program.

[0035] In one possible implementation, the electronic device is any one of a monitoring device, a display device, and a cloud processing device.

[0036] In one possible implementation, the display device is a mobile terminal or a monitoring terminal with a display screen.

[0037] A fourth aspect of this application provides a computer-readable storage medium storing a computer program, the computer program being executed by a processor of the steps of the method of the first aspect.

[0038] The fifth aspect of this application provides a computer program product containing instructions that, when executed by a cluster of computer devices, cause the cluster of computer devices to perform the steps of the method of the first aspect.

[0039] Compared with the prior art, the embodiments of this application have at least the following technical effects: The first aspect of this application provides a respiratory and heart rate monitoring method based on variational mode decomposition. This method processes acquired raw data to obtain a raw phase sequence. The raw data is obtained from the echo signal returned by the detection signal of a non-contact detection module through the target area of ​​the target object. This method monitors vital signs in a non-contact manner. The raw phase sequence is obtained by solving the complex numbers corresponding to the chirps of each linear frequency modulated signal under the target distance unit, where the target distance unit is the distance unit corresponding to the energy peak of the echo signal. Based on the raw phase sequence, variational mode decomposition is performed to obtain multiple target intrinsic mode functions (IMFs). Based on these IMFs, respiratory data and / or heart rate data are determined. Therefore, this application uses variational mode decomposition to achieve blind separation of respiratory and heart rate data, thereby displaying the respiratory and / or heart rate data to caregivers. This facilitates timely understanding of the target object's vital signs and is applicable to high-precision non-contact monitoring of infant respiration and heart rate.

[0040] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of a respiratory heart rate monitoring method based on variational mode decomposition provided in an embodiment of this application; Figure 2 This is a flowchart of a step for processing acquired raw data to obtain a raw phase sequence, as provided in an embodiment of this application. Figure 3 This is a flowchart of another respiratory heart rate monitoring method based on variational mode decomposition provided in the embodiments of this application; Figure 4 This is a flowchart of another respiratory heart rate monitoring method based on variational mode decomposition provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of a respiratory heart rate monitoring device based on variational mode decomposition provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0043] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0044] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0045] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0046] In the description of this application, unless otherwise stated, the " / " used in this specification and appended claims indicates that the related objects are in an "or" relationship. For example, A / B can mean A or B. The "and / or" in this application merely describes the relationship between the related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. Furthermore, in the description of this application, unless otherwise stated, "multiple" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, a and b, a and c, b and c, or a, b, and c. Here, a, b, and c can be single or multiple.

[0047] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0048] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0049] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0050] Research has found that existing methods for monitoring infant respiratory and heart rate can be divided into three categories, as follows: (1) Baseline Scheme for Contact-Based Infant Monitoring This is the mainstream approach used in hospital neonatal departments. It involves attaching ECG electrodes, a chest strap respiratory sensor, and a wearable pulse oximeter to the infant's skin to collect ECG and chest impedance signals. After amplification and analog-to-digital conversion, the heart rate is calculated using threshold filtering and R-wave peak detection, and the respiratory rate is calculated using the chest wall motion cycle. The system outputs physiological parameters and abnormal alarms. This approach is considered the clinical gold standard, but it is only suitable for medical settings and cannot be used for continuous daily monitoring at home.

[0051] (2) General solution for monitoring adult vital signs with millimeter-wave radar As the current mainstream technology for non-contact monitoring, it uses commercial radar modules such as the TI IWR6843 as the hardware carrier, and the core process is as follows: 1. The radar acquires raw ADC data and performs a 1D range-dimensional FFT (Fast Fourier Transform) to obtain the range-frame matrix; 2. Based on the peak amplitude, the target distance unit of the human chest cavity is locked, and the phase information of the corresponding complex signal is extracted; 3. Phase entanglement is removed by arctangent and phase unwrapping, and DC offset is eliminated by differential to obtain the phase signal related to thoracic cavity micromotion; 4. The respiratory and heart rate frequency band signals are separated by a fixed-order IIR bandpass filter, and the respiratory rate and heart rate are finally obtained by FFT frequency domain peak detection or time domain peak statistics.

[0052] (3) Improved millimeter-wave radar scheme for infant scenarios Existing research specifically targeting infants focuses on adjusting frequency band parameters based on adult protocols, with the core improvement being: 1. Reduced target distance detection range to 0.3-2m, suitable for crib scenarios; 2. Adjust the respiratory bandpass range to 0.2-1.0Hz (12-60 beats / min) and the heart rate bandpass range to 1.2-3.0Hz (72-180 beats / min) to match the infant's physiological range.

[0053] The main problems in infant vital sign monitoring are as follows: (1) The ability to extract weak vital signs signals in infants is severely insufficient. The amplitude of infants' chest breathing micro-movements is only 0.05-0.2mm, which is 1 / 5-1 / 10 of that of adults. Existing simple coherent accumulation and fixed clutter cancellation methods cannot effectively suppress noise caused by crib vibration, air conditioning airflow, and bedding obstruction. The signal-to-noise ratio is generally lower than 5dB, making it very easy for signal loss and monitoring interruption to occur.

[0054] (2) Harmonic interference in infant respiration cannot be effectively eliminated An infant's respiratory rate can reach up to 60 breaths per minute (1.0 Hz). Its second harmonic (2.0 Hz) and third harmonic (3.0 Hz) completely overlap with the infant's heart rate frequency band (1.2-3.0 Hz). Existing fixed filtering methods cannot achieve effective blind separation, and respiratory harmonics are easily misidentified as heart rate signals.

[0055] (3) There is no abnormal early warning and result verification mechanism for infant physiological characteristics. The existing solution only outputs respiratory rate and heart rate values, and does not design discrimination logic for specific abnormal symptoms such as infant apnea, tachycardia / bradycardia, and periodic breathing. At the same time, it only uses single-dimensional frequency domain peak detection results and has no cross-validation mechanism, which is prone to false alarms and false negatives, and cannot meet the monitoring reliability requirements of home and primary healthcare.

[0056] The respiratory and heart rate monitoring method and electronic device based on variational mode decomposition provided in this application are intended to solve the above-mentioned technical problems of the prior art.

[0057] The technical solution of this application and how it solves the above-mentioned technical problems are described in detail below with specific embodiments. It should be noted that the following embodiments can be referenced, borrowed, or combined with each other, and the same terms, similar features, and similar implementation steps in different embodiments will not be described again.

[0058] See Figure 1 As shown, this application provides a flowchart of a respiratory and heart rate monitoring method based on variational mode decomposition. Figure 1 As shown, the respiratory heart rate monitoring method based on variational mode decomposition includes steps S101 to S105.

[0059] S101. The acquired raw data is processed to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal.

[0060] This application embodiment can obtain an original phase sequence by performing a series of preset data processing procedures on the original data. The original phase sequence is a one-dimensional phase sequence, which is the phase value obtained by solving a complex number through an arctangent operation.

[0061] Optionally, the target object is an infant, the target area is the infant's chest cavity, and the raw data can be obtained by analyzing the echo signal.

[0062] Optionally, the non-contact detection module can be a radar module that transmits millimeter-wave radar as a detection signal.

[0063] Optionally, the radar module can adopt a domestically produced 1-transmit 1-receive or 1-transmit 3-receive module, which can be expanded to adapt to multi-antenna millimeter-wave radars such as 2-transmit 2-receive or 4-receive, as well as radar hardware in the 24GHz and 77GHz frequency bands.

[0064] S102. Based on the original phase sequence, perform variational mode decomposition to obtain multiple target intrinsic mode functions.

[0065] In some embodiments, variational mode decomposition is performed based on the original phase sequence to obtain multiple target intrinsic mode functions, including: performing time-series difference calculation on the original phase sequence to obtain a differential phase signal; and performing variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions.

[0066] Since the original phase sequence has DC offset and linear phase drift, it cannot be directly fed into VMD (Variational Mode Decomposition) for decomposition. Instead, the differential phase signal can be obtained by performing time-series differential calculations on a chirp-by-chirp basis according to the target distance cell.

[0067] After the original phase sequence in this embodiment of the application is subjected to time-series differential, the signal becomes stable and the energy is concentrated on the two main peaks of respiration and heart rate. Then, variational mode decomposition is performed, which is beneficial for the subsequent clean and efficient screening and separation of respiratory signals and / or heart rate signals.

[0068] S103. Determine respiratory data and / or heart rate data based on multiple target intrinsic mode functions.

[0069] Optionally, the respiratory data includes at least one of a respiratory dynamic waveform and a respiratory rate, and the heart rate data includes at least one of a heart rate dynamic waveform and a heart rate.

[0070] In some embodiments, respiratory data and / or heart rate data are determined based on multiple target intrinsic mode functions, including: Multiple target intrinsic mode functions are filtered and separated to reconstruct respiratory signals and / or heart rate signals; Based on respiratory signals and / or heart rate signals, corresponding respiratory data and / or heart rate data are determined.

[0071] This application embodiment achieves blind separation of respiratory and / or heart rate signals by screening, separating and reconstructing multiple target intrinsic mode functions. Then, based on the respiratory and / or heart rate signals, corresponding respiratory and / or heart rate data are determined, thus achieving effective blind separation of respiratory and heart rate data.

[0072] S104. Display respiratory data and / or heart rate data, and / or send respiratory data and / or heart rate data to a terminal device to display respiratory data and / or heart rate data through the terminal device.

[0073] The respiratory and heart rate monitoring method based on variational mode decomposition provided in this application can process the acquired raw data to obtain a raw phase sequence. The raw data is obtained from the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object, which is a non-contact method for monitoring vital signs. The raw phase sequence is obtained by solving the complex number corresponding to the chirp of each linear frequency modulated signal under the target distance unit, where the target distance unit is the distance unit corresponding to the energy peak of the echo signal. Therefore, based on the raw phase sequence, variational mode decomposition is performed to obtain multiple target intrinsic mode functions. Based on the multiple target intrinsic mode functions, respiratory data and / or heart rate data are determined. Therefore, this application uses variational mode decomposition to achieve blind separation of respiratory data and heart rate data, and then displays the respiratory data and / or heart rate data to caregivers, so that caregivers can understand the vital signs of the target object in a timely manner. It is applicable to high-precision non-contact monitoring of infant respiration and heart rate.

[0074] In some embodiments, variational mode decomposition is performed on the differential phase signal to obtain multiple target intrinsic mode functions, including: The target variational mode decomposition algorithm is used to perform variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions; The objective variational mode decomposition algorithm is obtained as follows: An initial variational mode decomposition algorithm is used to perform variational mode decomposition on the differential phase signal to obtain multiple candidate intrinsic mode functions; The Grey Wolf optimization algorithm is used to update the number of modes and the penalty factor of the initial variational mode decomposition algorithm until the total envelope entropy of multiple candidate intrinsic mode functions is minimized, thus obtaining the target variational mode decomposition algorithm.

[0075] In this embodiment, the Grey Wolf optimization algorithm is used to adaptively determine the variational mode decomposition parameters, and then the differential phase signal is subjected to variational mode decomposition to obtain multiple target intrinsic mode functions.

[0076] In this embodiment, VMD decomposition is performed on the differential phase signal to obtain multiple IMFs (Intrinsic Mode Functions). Then, the improved Grey Wolf Optimizer (GWO) algorithm is used, with the minimum total envelope entropy of all IMFs as the optimization criterion, to automatically optimize the number of VMD modes K and the penalty factor α. This solves the problem of over-decomposition or under-decomposition caused by manually setting parameters in the prior art, and realizes adaptive optimization of VMD parameters.

[0077] Furthermore, embodiments of this application can adaptively decompose VMD, using optimized variational mode decomposition parameters to perform VMD decomposition on the differential phase signal to obtain multiple IMFs.

[0078] Furthermore, the embodiments of this application, based on the improved GWO-VMD for respiratory harmonic suppression and blind separation of physiological signals, can completely solve the problem of overlap between infant respiratory harmonics and heart rate frequency bands, and achieve precise separation of respiratory and heart rate signals.

[0079] In some embodiments, filtering and separating multiple target intrinsic mode functions to reconstruct respiratory signals and / or heart rate signals includes: Determine the center frequency of the intrinsic mode function of each target; Based on the center frequency of each target intrinsic mode function, the intrinsic mode functions of each target are screened and separated to reconstruct respiratory signals and / or heart rate signals.

[0080] The embodiments of this application can filter and separate target intrinsic mode functions by the center frequency of each target intrinsic mode function, thereby reconstructing respiratory signals and / or heart rate signals.

[0081] The embodiments of this application can filter and separate multiple target intrinsic mode functions based on frequency range and mutual information constraints to reconstruct respiratory signals and / or heart rate signals.

[0082] In some embodiments, based on the center frequency of each target intrinsic mode function, the target intrinsic mode functions are screened and separated to reconstruct respiratory signals and / or heart rate signals, including: The respiratory signal is obtained by summing and reconstructing multiple target intrinsic mode functions corresponding to the center frequency within the first frequency range; Multiple target intrinsic mode functions corresponding to the center frequency in the second frequency range are determined as multiple candidate intrinsic mode functions; the frequency in the second frequency range is greater than the frequency in the first frequency range; Determine the mutual information between each candidate intrinsic mode function and the respiratory signal; The heart rate signal is obtained by accumulating and reconstructing the candidate intrinsic mode functions whose mutual information is below the first threshold.

[0083] Optionally, the first frequency range can be selected as 0.2-1.0Hz, and the second frequency range can be selected as 1.2-3.0Hz. As an example, this application provides an automatic component filtering method based on mutual information: calculating the mutual information between each IMF and the differential phase signal, and simultaneously calculating the center frequency of each IMF, and automatically filtering according to the infant's physiological frequency band. For example: IMFs with a center frequency of 0.2-1.0Hz as candidate intrinsic mode functions are accumulated and reconstructed into a respiratory signal; IMFs with a center frequency of 1.2-3.0Hz as candidate intrinsic mode functions, and whose mutual information with the respiratory signal is lower than a first threshold, are accumulated and reconstructed into a heart rate signal, completely eliminating respiratory harmonic components.

[0084] See Figure 2 As shown, this application embodiment provides a flowchart of the steps for processing acquired raw data to obtain a raw phase sequence. For example... Figure 2 As shown, the process of processing the acquired raw data to obtain the raw phase sequence includes steps S201 to S203.

[0085] S201. Perform data preprocessing on the original data to obtain the target complex matrix; the target complex matrix is ​​a two-dimensional complex matrix consisting of the total number of chirps and the total number of distance cells in a single frame.

[0086] In some embodiments, data preprocessing is performed on the original data to obtain the target complex matrix, including: The raw data is preprocessed to generate a target three-dimensional echo matrix. The target three-dimensional echo matrix is ​​a three-dimensional complex matrix consisting of the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp. The non-contact detection module includes multiple antennas. The target three-dimensional echo matrix is ​​processed by the antenna's weighted coherent accumulation to obtain the first two-dimensional complex matrix; the first two-dimensional complex matrix is ​​a complex matrix of the total number of chirps and the total number of range cells in a single frame; The first two-dimensional complex matrix is ​​subjected to two-stage clutter suppression processing to obtain the target complex matrix.

[0087] The embodiments of this application can enhance the signal-to-noise ratio of weak, micro-motion signals at the 0.05mm level in infants by using multi-antenna weighted coherent accumulation and two-stage adaptive clutter suppression preprocessing, and effectively suppress static and dynamic clutter interference.

[0088] In this application embodiment, the signal enhancement method based on Singular Value Decomposition (SVD) can be used to replace the weighted coherent accumulation method to obtain the first two-dimensional complex matrix.

[0089] In some embodiments, data preprocessing is performed on the raw data to generate a target three-dimensional echo matrix, including: The raw data is parsed according to the frame timing rules to generate the first three-dimensional echo matrix; the first three-dimensional echo matrix is ​​a three-dimensional complex matrix with the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp; The first three-dimensional echo matrix is ​​normalized to obtain the second three-dimensional echo matrix; The target three-dimensional echo matrix is ​​obtained by performing a Hamming windowing operation on each Chirp of each antenna on the second three-dimensional echo matrix.

[0090] As an example, the radar module outputs a binary I / Q raw data stream; this stream is parsed according to frame timing rules to form a first three-dimensional echo matrix, with dimensions of: number of receiving antennas × total number of chirps per frame × number of sampling points per chirp. Each sampling point in the first three-dimensional echo matrix contains two integers, a real part and an imaginary part, corresponding to the orthogonal echo components of the same radar sampling. The first three-dimensional echo matrix is ​​then quantized and normalized, mapping all sampled values ​​to the interval [-1, 1]. A Hamming windowing operation is performed on each chirp data from each receiving antenna to suppress spectral leakage in the range-dimensional FFT.

[0091] In some embodiments, the target three-dimensional echo matrix is ​​subjected to antenna weighted coherent accumulation processing to obtain a first two-dimensional complex matrix, including: Perform a range-dimensional transformation on the target's three-dimensional echo matrix for each antenna to obtain the range-dimensional complex spectrum of each antenna; Based on the complex spectrum of each distance dimension, the echo signal-to-noise ratio of each antenna is determined; The weight of each antenna is determined based on the echo signal-to-noise ratio of each antenna; The first two-dimensional complex matrix is ​​determined based on the distance dimension complex spectrum of each antenna and the weight of each antenna.

[0092] As an example, this application employs adaptive weighted coherent accumulation: a range-dimensional FFT is performed on each chirp of the windowed target three-dimensional echo matrix to complete the conversion from time-domain sampling to range-domain frequency points, resulting in independent range-dimensional complex spectra for each antenna. Here, a range cell is the smallest range-resolution unit obtained by mapping a single frequency point after the radar undergoes range-dimensional FFT. Each range cell corresponds to a fixed physical distance interval in front of the radar; the index, from smallest to largest, represents the detection distance from near to far. Each range cell stores a set of complex I / Q echo data, carrying the echo amplitude and phase information at that location. Subsequent clutter suppression, target detection, and phase demodulation all use the range cell as the basic processing unit.

[0093] Then, in this embodiment, the signal-to-noise ratio of the echo of a single receiving antenna is calculated for each distance cell. The signal-to-noise ratio of each antenna is used as a weighting coefficient to coherently accumulate the complex signals of the same distance cell. The antenna with the higher signal-to-noise ratio is assigned a greater weight. The receiving antenna dimension is eliminated, and the output dimension is a single-frame integrated distance-Chirp complex matrix (i.e., the first two-dimensional complex matrix) with the total number of Chirps in a single frame × the total number of distance cells, which maximizes the gain of the weak chest cavity micro-motion signal.

[0094] In some embodiments, a two-stage clutter suppression process is performed on the first two-dimensional complex matrix to obtain the target complex matrix, including: Subtract a preset complex number from each complex number in the first two-dimensional complex matrix to obtain the second two-dimensional complex matrix; the preset complex number is obtained by averaging the complex numbers corresponding to each distance unit in the first two-dimensional complex matrix for a preset number of frames. The target complex matrix is ​​obtained by filtering the second two-dimensional complex matrix using a least mean square adaptive filter.

[0095] Optionally, the preset number of frames can be the first 32 frames.

[0096] As an example, this application embodiment employs a two-stage clutter suppression: the first stage employs static clutter cancellation, calculating the average complex number of each distance unit in the first 32 frames, subtracting the average complex number from the current frame signal to remove the static reflection signal of the fixed background, and the static reflection components of walls, bed frames, fixed bedding, etc., which remain unchanged for a long time, are completely canceled out, leaving only time-varying components such as human body micro-movement, bed shaking, and airflow.

[0097] The second stage employs variable step size LMS (Least Mean Square Adaptive Filter) adaptive filtering to track and suppress non-static clutter caused by bed vibration and airflow disturbance in real time. The output dimension is the pure distance-frame complex matrix (i.e., the target complex matrix) which is the total number of chirps in a single frame × the total number of distance cells.

[0098] S202. For the target complex matrix, extract the complex number corresponding to each Chirp under the target distance cell, and split the extracted complex number into in-phase time components and quadrature time components.

[0099] Optionally, the in-phase timing component corresponds to the real part of the complex number, and the quadrature timing component corresponds to the imaginary part of the complex number.

[0100] S203. Based on the in-phase and quadrature timing components, the arctangent operation is used to solve each Chirp to obtain the original phase sequence.

[0101] As an example, this embodiment of the application uses a target range cell with a fixed Kalman filter output, extracts the complex numbers corresponding to all chirps under the target range cell, splits the complex numbers into two time-series components: in-phase I and quadrature Q, and solves the original phase sequence one chirp by one through arctangent operation; finally, it outputs three sets of one-dimensional time-series vectors (in-phase time-series component, quadrature time-series component, and original phase sequence) as input data for phase demodulation and physiological signal separation.

[0102] The embodiments of this application can obtain raw phase data through data processing processes such as preprocessing and target locking of the raw data.

[0103] As an example, the embodiments of this application can realize automatic locking and tracking of the infant target distance unit, accurately lock the infant's chest cavity target, shield interference from bedding, bed rails, etc., and adapt to the scenario of the infant turning over and moving.

[0104] After clutter suppression, the target complex matrix is ​​fixed in range cells. The sum of squared amplitudes along the frame dimension yields a one-dimensional range-dimensional energy spectrum. Within the effective range of the crib (0.3-2.0m), the range cells corresponding to the energy peaks are searched and used as the initial target range cells. A sliding-window Kalman filter is employed to track the target range cells in real time. When the baby moves, causing target displacement, the locked range cells are automatically updated to prevent target loss.

[0105] This application embodiment uses the automatic locking and tracking of the infant target distance unit to accurately lock the infant's chest cavity target, shielding it from interference from bedding, bed rails, etc., and adapting to scenarios where the infant rolls over and moves.

[0106] In some embodiments, determining respiratory data and / or heart rate data based on respiratory signals and / or heart rate signals includes: The respiratory signal and / or heart rate signal are filtered separately to obtain the respiratory time-domain waveform and / or heart rate time-domain waveform. Based on the respiratory time-domain waveform and / or heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension, respectively. If the difference between two candidate respiratory rates is less than the first preset error, then the average of the two candidate respiratory rates is taken as the respiratory rate. If the difference between two candidate heart rates is less than the second preset error, then the average of the two candidate heart rates will be used as the heart rate.

[0107] The embodiments of this application can determine respiratory rate and heart rate by calculating parameters from both the frequency domain and time domain dimensions, thereby improving the accuracy of respiratory rate and heart rate calculation and avoiding misjudgments from single-dimensional detection.

[0108] See Figure 3 As shown, this application provides a flowchart of another respiratory and heart rate monitoring method based on variational mode decomposition. Figure 3 As shown, the respiratory heart rate monitoring method based on variational mode decomposition includes steps S301 to S309.

[0109] S301. Obtain raw data; the raw data is obtained from the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object.

[0110] S302. Process the raw data to obtain the raw phase sequence. The raw phase sequence is obtained by solving the complex number corresponding to each linear frequency modulated signal Chirp under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal.

[0111] S303. Perform time-series differential calculation on the original phase sequence to obtain the differential phase signal.

[0112] S304. Perform variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions.

[0113] S305. Screen and separate multiple target intrinsic mode functions to reconstruct respiratory signals and / or heart rate signals.

[0114] S306. Filter the respiratory signal and / or heart rate signal respectively to obtain the respiratory time-domain waveform and / or heart rate time-domain waveform.

[0115] The embodiments of this application can use SG filtering to smooth and denoise respiratory signals and / or heart rate signals, and output clean respiratory time-domain waveforms and / or heart rate time-domain waveforms.

[0116] S307. Based on the respiratory time-domain waveform and / or heart rate time-domain waveform, determine two candidate respiratory rates and / or two candidate heart rates from the frequency domain dimension and the time domain dimension, respectively.

[0117] In some embodiments, based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension, respectively, including: Based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, spectral peak search is performed in the physiological frequency bands corresponding to the respiratory time-domain waveform and the physiological frequency bands corresponding to the heart rate time-domain waveform, respectively, to determine the first candidate respiratory rate and / or the first candidate heart rate. Normalize the respiratory time-domain waveform and / or heart rate time-domain waveform, and determine the candidate respiratory peak and / or the candidate heart rate peak of the normalized respiratory time-domain waveform. If the candidate respiratory peak value is greater than the first preset peak value, then the candidate respiratory peak value is determined to be a valid respiratory peak value; If the candidate heart rate peak value is greater than the second preset peak value, then the candidate heart rate peak value is determined to be the effective heart rate peak value; A second candidate respiratory rate and / or a second candidate heart rate are determined based on the time interval between adjacent effective respiratory peaks and / or the time interval between adjacent effective heart rate peaks.

[0118] As an example, frequency domain dimension calculation: Perform a 1024-point FFT on the respiratory time-domain waveform and the heart rate time-domain waveform, search for the spectral peaks in the corresponding physiological frequency bands, and convert them into respiratory rate and heart rate values, with a frequency resolution of 0.0488Hz.

[0119] As an example, time-domain dimension calculation: An improved peak detection algorithm is used to locate peaks in the time-domain waveform. The specific steps are as follows: (1) Normalize the time-domain waveforms of respiration and heart rate respectively, and calculate the local maximum values ​​as candidate respiratory peaks and candidate heart rate peaks; (2) The first preset peak value and the second preset peak value are respectively set to 1.5 times the standard deviation of the signal to filter out the effective peak values; (3) Calculate the time interval between adjacent effective peaks, take the average value as the cardiac cycle / respiratory cycle, and convert it into heart rate / respiratory rate.

[0120] S308. If the difference between two candidate respiratory rates is lower than the first preset error, then the average of the two candidate respiratory rates shall be taken as the respiratory rate.

[0121] S309. If the difference between the two candidate heart rates is lower than the second preset error, then the average of the two candidate heart rates shall be taken as the heart rate.

[0122] Optionally, the first preset error is set to 5. In practical applications, the relative error between two candidate respiratory rates can also be set to be lower than the first error threshold, and the relative error between two candidate heart rates can be set to be lower than the second error threshold. The first and second error thresholds can both be set to a relative error threshold of 5%.

[0123] As an example, this application employs multiple three-level cross-validation mechanisms, the technical solutions of which are as follows: Level 1 Time-Frequency Cross-Check: Set a 5% error threshold. When the relative error between the frequency domain and time domain calculation results is lower than the error threshold, take the average of the two as the final output. Second-level time-frequency domain wavelet transform verification: When the error exceeds the threshold, the db4 wavelet is used to decompose the respiratory time-domain waveform and the heart rate time-domain waveform into 5 levels, extract the wavelet coefficients corresponding to respiratory and heart rate, and calculate the time-frequency peak value for secondary verification. Third-level historical data verification: If the second verification still fails, the historical physiological parameter data of the predetermined time (e.g., the previous 10 seconds) is referenced. If the current result deviates from the historical average by more than 20%, it is marked as an invalid frame and no error value is output.

[0124] The embodiments of this application improve the accuracy of respiratory rate and heart rate calculations by calculating and cross-validating multi-dimensional physiological parameters, and avoid misjudgments caused by single-dimensional detection.

[0125] In some embodiments, after determining respiratory data and / or heart rate data based on a plurality of said target intrinsic mode functions, the method further includes: The respiratory abnormality type is determined based on respiratory data, and an early warning message is generated based on the respiratory abnormality type; and / or, the heart rate abnormality type is determined based on heart rate data, and an early warning message is generated based on the heart rate abnormality type.

[0126] Display warning information, and / or send warning information to terminal devices for display on the terminal devices.

[0127] Optionally, the warning information can be displayed to caregivers via voice prompts or by showing the warning information on a screen.

[0128] The embodiments of this application can also determine the type of respiratory abnormality based on respiratory signals and the type of heart rate abnormality based on heart rate signals.

[0129] This application's embodiments can provide real-time early warning for specific physiological abnormalities in infants. Specifically, this application's embodiments can be a complete algorithm for non-contact respiratory and heart rate monitoring and abnormal early warning of infants using 60GHz band FMCW millimeter-wave radar.

[0130] See Figure 4 As shown, this application provides a flowchart of another respiratory heart rate monitoring method based on variational mode decomposition. Figure 4 As shown, the respiratory heart rate monitoring method based on variational mode decomposition includes steps S401 to S406.

[0131] S401. Obtain raw data; the raw data is obtained from the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object.

[0132] S402. Process the raw data to obtain the raw phase sequence. The raw phase sequence is obtained by solving the complex number corresponding to each linear frequency modulated signal Chirp under the target range cell. The target range cell is the range cell corresponding to the energy peak of the echo signal.

[0133] S403. Perform time-series differential calculation on the original phase sequence to obtain the differential phase signal.

[0134] S404. Perform variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions.

[0135] S405. Filter and separate multiple target intrinsic mode functions to reconstruct respiratory signals and / or heart rate signals.

[0136] S406. Determine the type of respiratory abnormality based on respiratory data and generate early warning information based on the type of respiratory abnormality, and / or determine the type of heart rate abnormality based on heart rate data and generate early warning information based on the type of heart rate abnormality.

[0137] Optionally, physiological abnormalities are determined based on respiratory signals and / or heart rate signals, including: determining physiological abnormalities based on respiratory rate and / or heart rate.

[0138] Optionally, the types of respiratory abnormalities include apnea and respiratory disorders, and the types of heart rate abnormalities include tachycardia, bradycardia, and periodic breathing.

[0139] The embodiments of this application can provide real-time early warning for specific physiological abnormalities in infants.

[0140] As an example, this application embodiment sets infant-specific anomaly detection rules based on the output results of a continuous sliding window: (1) Apnea: respiratory rate less than 10 breaths / min and lasting for more than 5 seconds; (2) Abnormal breathing: respiratory rate consistently higher than 60 breaths / min or lower than 20 breaths / min for more than 30 seconds; (3) Tachycardia: Heart rate sustained above 180 beats / min for more than 30 seconds; (4) Bradycardia: Heart rate consistently below 90 beats / min for more than 30 seconds; (5) Periodic breathing: The amplitude of breathing increases and decreases periodically, accompanied by brief pauses in breathing, lasting for more than 1 minute; Optionally, when an abnormal state is detected, the time of occurrence, duration, and type of the abnormality are automatically marked, and an early warning message is output. At the same time, the original signals and waveforms before and after the abnormality are saved for retrospective analysis.

[0141] Based on the above technical solution, the main technical contents of the embodiments of this application include: (1) Based on the improved GWO-VMD method for the suppression of infant respiratory harmonics and the blind separation of physiological signals, the core solution is to address the industry pain point of overlap between infant respiratory harmonics and heart rate frequency bands, and to achieve accurate extraction of heart rate signals.

[0142] (2) Based on the signal-to-noise ratio weighted multi-antenna multi-chirp coherent accumulation and two-stage adaptive clutter suppression method, the signal-to-noise ratio of weak and micro-motion signals at the level of 0.05mm for infants is enhanced, solving the core problem of the difficulty in weak signal extraction.

[0143] (3) The multi-dimensional cross-validation mechanism of infant physiological parameters and the exclusive abnormal early warning method significantly reduce the false alarm rate of parameters, realize the accurate identification of newborn-specific high-risk symptoms, and improve the clinical usability of the algorithm.

[0144] The respiratory and heart rate monitoring method based on variational mode decomposition applied in the embodiments of this application can achieve at least the following technical effects: (1) Significantly improved weak signal extraction capability: Through weighted coherent accumulation and two-stage clutter suppression, the signal-to-noise ratio of infant micro-movement signal is improved from less than 5dB in the existing technology to more than 15dB, which can stably detect chest micro-movements of 0.05mm level, and the signal loss rate under the occlusion of bedding is reduced from 30% to less than 10%.

[0145] (2) Breakthrough improvement in heart rate measurement accuracy: By improving the respiratory harmonic suppression of GWO-VMD, the interference problem of the overlap between harmonics and heart rate frequency bands is completely solved, effectively reducing the average error of infant heart rate measurement.

[0146] (3) The reliability and clinical value of monitoring are significantly improved: the multi-dimensional cross-validation mechanism reduces the false alarm rate of parameters from 20% of the existing technology to less than 5%; the infant-specific abnormal warning mechanism can accurately identify high-risk symptoms such as apnea and is suitable for daily family monitoring.

[0147] See Figure 5 As shown, this application provides a schematic diagram of the structure of a respiratory heart rate monitoring device 50 based on variational mode decomposition. Figure 5 As shown, the respiratory and heart rate monitoring device 50 based on variational mode decomposition includes: a first data processing module 501, a second data processing module 502, a respiratory and heart rate determination module 503, and a display processing module 504.

[0148] The first data processing module 501 is used to process the acquired raw data to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal. The second data processing module 502 is used to perform variational mode decomposition based on the original phase sequence to obtain multiple target intrinsic mode functions; The respiratory and heart rate determination module 503 is used to determine respiratory data and / or heart rate data based on multiple target intrinsic mode functions; The display processing module 504 is used to display respiratory data and / or heart rate data, and / or send respiratory data and / or heart rate data to a terminal device for display of respiratory data and / or heart rate data through the terminal device.

[0149] Optionally, the respiratory and heart rate determination module 503 is used to filter and separate multiple target intrinsic mode functions, reconstruct respiratory signals and / or heart rate signals, and determine respiratory data and / or heart rate data based on the respiratory signals and / or heart rate signals.

[0150] Optionally, the second data processing module 502 is used to perform time-series differential calculation on the original phase sequence to obtain a differential phase signal; and to perform variational mode decomposition on the differential phase signal to obtain multiple target intrinsic mode functions.

[0151] Optionally, the second data processing module 502 is used to perform variational mode decomposition on the differential phase signal using a target variational mode decomposition algorithm to obtain multiple target intrinsic mode functions; The objective variational mode decomposition algorithm is obtained as follows: An initial variational mode decomposition algorithm is used to perform variational mode decomposition on the differential phase signal to obtain multiple candidate intrinsic mode functions; The Grey Wolf optimization algorithm is used to adaptively determine the number of modes and the penalty factor of the initial variational mode decomposition algorithm until the total envelope entropy of multiple candidate intrinsic mode functions is minimized, thus obtaining the target variational mode decomposition algorithm.

[0152] Optionally, the respiratory and heart rate determination module 503 is used to determine the center frequency of each target intrinsic mode function; based on the center frequency of each target intrinsic mode function, the target intrinsic mode functions are screened and separated to reconstruct the respiratory signal and / or heart rate signal.

[0153] Optionally, the respiratory heart rate determination module 503 is used to accumulate and reconstruct multiple target intrinsic mode functions corresponding to the center frequency in a first frequency range to obtain a respiratory signal; determine multiple target intrinsic mode functions corresponding to the center frequency in a second frequency range as multiple candidate intrinsic mode functions; the frequency in the second frequency range is greater than the frequency in the first frequency range; determine the mutual information between each candidate intrinsic mode function and the respiratory signal; and accumulate and reconstruct candidate intrinsic mode functions with mutual information lower than a first threshold to obtain a heart rate signal.

[0154] Optionally, the respiratory and heart rate determination module 503 is used to filter the respiratory signal and / or heart rate signal respectively to obtain a respiratory time-domain waveform and / or a heart rate time-domain waveform; based on the respiratory time-domain waveform and / or heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension respectively; if the difference between the two candidate respiratory rates is lower than a first preset error, the average of the two candidate respiratory rates is taken as the respiratory rate; if the difference between the two candidate heart rates is lower than a second preset error, the average of the two candidate heart rates is taken as the heart rate.

[0155] Optionally, the respiratory and heart rate determination module 503 is used to perform spectrum peak search in the physiological frequency band corresponding to the respiratory time-domain waveform and the physiological frequency band corresponding to the heart rate time-domain waveform, respectively, based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, to determine the first candidate respiratory rate and / or the first candidate heart rate. Normalize the respiratory time-domain waveform and / or heart rate time-domain waveform, and determine the candidate respiratory peak and / or the candidate heart rate peak of the normalized respiratory time-domain waveform. If the candidate respiratory peak value is greater than the first preset peak value, then the candidate respiratory peak value is determined to be a valid respiratory peak value; If the candidate heart rate peak value is greater than the second preset peak value, then the candidate heart rate peak value is determined to be the effective heart rate peak value; A second candidate respiratory rate and / or a second candidate heart rate are determined based on the time interval between adjacent effective respiratory peaks and / or the time interval between adjacent effective heart rate peaks.

[0156] Optionally, the first data processing module 501 is used to preprocess the original data to obtain a target complex matrix; the target complex matrix is ​​a two-dimensional complex matrix consisting of the total number of chirps and the total number of distance cells in a single frame; for the target complex matrix, the complex number corresponding to each chirp under the target distance cell is extracted, and the extracted complex number is split into in-phase timing components and quadrature timing components; based on the in-phase timing components and quadrature timing components, arctangent operation is used to solve for each chirp to obtain the original phase sequence.

[0157] Optionally, the first data processing module 501 is used to preprocess the raw data to generate a target three-dimensional echo matrix; the target three-dimensional echo matrix is ​​a three-dimensional complex matrix of the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp, and the non-contact detection module includes multiple antennas; The target three-dimensional echo matrix is ​​processed by the antenna's weighted coherent accumulation to obtain the first two-dimensional complex matrix; the first two-dimensional complex matrix is ​​a complex matrix of the total number of chirps and the total number of range cells in a single frame; The first two-dimensional complex matrix is ​​subjected to two-stage clutter suppression processing to obtain the target complex matrix.

[0158] Optionally, the first data processing module 501 is used to parse the original data according to the frame timing rules to generate a first three-dimensional echo matrix; the first three-dimensional echo matrix is ​​a three-dimensional complex matrix with the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp; the first three-dimensional echo matrix is ​​normalized to obtain a second three-dimensional echo matrix; the second three-dimensional echo matrix is ​​subjected to a Hamming windowing operation for each chirp of each antenna to obtain a target three-dimensional echo matrix.

[0159] Optionally, the first data processing module 501 is used to perform a range dimension transformation on the target three-dimensional echo matrix according to each antenna to obtain the range dimension complex spectrum of each antenna; determine the echo signal-to-noise ratio of each antenna based on the range dimension complex spectrum; determine the weight of each antenna based on the echo signal-to-noise ratio of each antenna; and determine the first two-dimensional complex matrix based on the range dimension complex spectrum of each antenna and the weight of each antenna.

[0160] Optionally, the first data processing module 501 is used to subtract a preset complex number from each complex number in the first two-dimensional complex matrix to obtain a second two-dimensional complex matrix; the preset complex number is obtained by the average value of the complex number corresponding to each distance unit of the first two-dimensional complex matrix based on the preset number of frames; and the second two-dimensional complex matrix is ​​filtered by a least mean square adaptive filter to obtain a target complex matrix.

[0161] Optionally, the respiratory heart rate monitoring device 50 based on variational mode decomposition also includes an early warning module.

[0162] The early warning module is used to determine the type of respiratory abnormality based on respiratory data and generate early warning information based on the type of respiratory abnormality, and / or to determine the type of heart rate abnormality based on heart rate data and generate early warning information based on the type of heart rate abnormality.

[0163] The display processing module 504 is used to display warning information and / or send warning information to a terminal device for display of the warning information on the terminal device.

[0164] In applications, the modules in the variational mode decomposition-based respiratory and heart rate monitoring device 50 can be software program modules, or they can be implemented through different logic circuits integrated in the processor, or they can be implemented through multiple distributed processors.

[0165] The respiratory and heart rate monitoring device 50 based on variational mode decomposition in this application embodiment can execute the method provided in this application embodiment. The implementation principle is similar. The actions performed by each module in the respiratory and heart rate monitoring device 50 based on variational mode decomposition in each embodiment of this application correspond to the steps in the method of each embodiment of this application. For detailed functional descriptions of each module of the respiratory and heart rate monitoring device 50 based on variational mode decomposition, please refer to the descriptions in the corresponding methods shown above, which will not be repeated here.

[0166] This application provides an electronic device, including: Memory, used to store computer programs; A processor for executing a computer program to implement the steps of the method as described in the embodiments of this application.

[0167] In some embodiments, the electronic device is any one of a monitoring device, a display device, and a cloud processing device.

[0168] As an example scenario, the electronic device in this application embodiment can be a device that integrates monitoring and display, such as a monitoring device that can both collect and display data. Further, the monitoring device can be a smart surveillance camera with a display screen.

[0169] The example scenarios described above in this application are suitable for scenarios with high privacy and small space (such as a family bedroom), and are the simplest and require no additional network configuration.

[0170] As another example scenario, the electronic device in this application embodiment can be a display device, the original data acquisition adopts a monitoring device, the monitoring and display hardware are separated, and the display device displays respiratory data and / or heart rate data.

[0171] The above-described example scenarios in this application are suitable for scenarios with multiple beds that require centralized management (such as hospitals / nursing homes), which can greatly improve management efficiency. The monitoring equipment and display equipment can be located in the same local area network to achieve low-latency transmission.

[0172] As another example scenario, the electronic device in this application embodiment can be a cloud processing device. The original data acquisition is carried out by the monitoring device, and the data processing process is carried out by the cloud processing device. The cloud processing device sends the respiratory data and / or heart rate data to the display device, and the display device displays the respiratory data and / or heart rate data as a terminal device.

[0173] The above-described example scenarios in this application are suitable for remote, fragmented, and low-cost scenarios (such as family members providing care in different locations). They have low front-end hardware requirements but rely on stable network bandwidth and cloud server resources, and personal privacy data can be transmitted in encrypted form.

[0174] In some embodiments, the display device is a mobile terminal or a monitoring terminal with a display screen.

[0175] Optionally, the mobile terminal can be a desktop computer, laptop, handheld computer, or other device with a display screen, and the display screen of the care terminal can be a large screen monitor, which transmits data in real time and displays it centrally on the large screen at the nurse station in waveform and / or digital form.

[0176] See Figure 6 As shown, this application provides a schematic diagram of the structure of an electronic device 60. Figure 6 As shown, the electronic device 60 of this application embodiment includes: a memory 62, a processor 61, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the processor 61 executes the computer program, it implements the steps of the methods of the various embodiments of this application.

[0177] Electronic device 60 may include, but is not limited to, processor 61 and memory 62. Those skilled in the art will understand that electronic device 60 may also include more or fewer components, or combinations of certain components, or different components, such as input / output devices, network access devices, etc.

[0178] Processor 61 can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0179] In some embodiments, memory 62 may be an internal storage unit, such as a hard disk or RAM. Memory 62 may be a removable / non-removable, volatile / non-volatile computer system storage medium; for example, memory 62 may be a non-volatile memory used for reading and writing non-volatile magnetic media. In other embodiments, memory 62 may be an external storage device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD) card, flash card, etc., provided on electronic device 60. Memory 62 is used to store operating systems, applications, bootloaders, data, and other programs, such as program code for computer programs. Memory 62 may also be used to temporarily store data that has been output or will be output.

[0180] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0181] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0182] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0183] If the integrated units described above are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a device / terminal equipment, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0184] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc. The storage medium can also include combinations of the above types of memory.

[0185] This application provides a computer program product containing instructions that, when executed by a cluster of computer devices, cause the cluster of computer devices to perform the steps of the method described in this application.

[0186] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0187] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0188] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0189] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0190] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A respiratory and heart rate monitoring method based on variational mode decomposition, characterized in that, include: The acquired raw data is processed to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal. Based on the original phase sequence, variational mode decomposition is performed to obtain multiple target intrinsic mode functions; Based on multiple target intrinsic mode functions, respiratory data and / or heart rate data are determined; Display the respiratory data and / or the heart rate data, and / or send the respiratory data and / or the heart rate data to a terminal device for display of the respiratory data and / or heart rate data via the terminal device.

2. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 1, characterized in that, The determination of respiratory data and / or heart rate data based on multiple target intrinsic mode functions includes: The multiple target intrinsic mode functions are filtered and separated to reconstruct respiratory signals and / or heart rate signals; Based on the respiratory signal and / or the heart rate signal, corresponding respiratory data and / or heart rate data are determined.

3. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 1, characterized in that, Based on the original phase sequence, variational mode decomposition is performed to obtain multiple target intrinsic mode functions, including: The original phase sequence is subjected to time-series differential calculation to obtain a differential phase signal; Variational mode decomposition is performed on the differential phase signal to obtain multiple target intrinsic mode functions.

4. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 3, characterized in that, The variational mode decomposition of the differential phase signal yields multiple target intrinsic mode functions, including: The differential phase signal is subjected to variational mode decomposition using a target variational mode decomposition algorithm to obtain multiple target intrinsic mode functions; The target variational mode decomposition algorithm is obtained in the following way: The differential phase signal is subjected to variational mode decomposition using an initial variational mode decomposition algorithm to obtain multiple candidate intrinsic mode functions; The Grey Wolf optimization algorithm is used to adaptively determine the number of modes and the penalty factor of the initial variational mode decomposition algorithm until the total envelope entropy of the multiple candidate intrinsic mode functions is minimized, thereby obtaining the target variational mode decomposition algorithm.

5. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 2, characterized in that, The step of filtering and separating multiple target intrinsic mode functions to reconstruct respiratory signals and / or heart rate signals includes: Determine the center frequency of each of the target intrinsic mode functions; Based on the center frequency of each of the target intrinsic mode functions, the target intrinsic mode functions are screened and separated to reconstruct respiratory signals and / or heart rate signals.

6. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 5, characterized in that, The step of filtering and separating the target intrinsic mode functions based on their center frequencies to reconstruct respiratory signals and / or heart rate signals includes: The respiratory signal is obtained by summing and reconstructing multiple target intrinsic mode functions corresponding to the center frequency within the first frequency range; Multiple target intrinsic mode functions corresponding to the center frequency in the second frequency range are determined as multiple candidate intrinsic mode functions; the frequency in the second frequency range is greater than the frequency in the first frequency range; Determine the mutual information between each of the candidate intrinsic mode functions and the respiratory signal; The heart rate signal is obtained by accumulating and reconstructing the candidate intrinsic mode functions whose mutual information is lower than the first threshold.

7. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 2, characterized in that, The step of determining respiratory data and / or heart rate data based on the respiratory signal and / or the heart rate signal includes: The respiratory signal and / or the heart rate signal are filtered to obtain the respiratory time-domain waveform and / or the heart rate time-domain waveform. Based on the respiratory time-domain waveform and / or heart rate time-domain waveform, two candidate respiratory rates and / or two candidate heart rates are determined from the frequency domain dimension and the time domain dimension, respectively. If the difference between two candidate respiratory rates is less than a first preset error, then the average of the two candidate respiratory rates is taken as the respiratory rate. If the difference between two candidate heart rates is less than the second preset error, then the average of the two candidate heart rates is taken as the heart rate.

8. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 7, characterized in that, The step of determining two candidate respiratory rates and / or two candidate heart rates from the frequency domain and time domain dimensions, respectively, based on the respiratory time-domain waveform and / or heart rate time-domain waveform, includes: Based on the respiratory time-domain waveform and / or the heart rate time-domain waveform, spectral peak search is performed on the physiological frequency bands corresponding to the respiratory time-domain waveform and the physiological frequency bands corresponding to the heart rate time-domain waveform to determine the first candidate respiratory rate and / or the first candidate heart rate. The respiratory time-domain waveform and / or the heart rate time-domain waveform are normalized, and candidate respiratory peaks of the normalized respiratory time-domain waveform and / or candidate heart rate peaks of the normalized heart rate time-domain waveform are determined. If the candidate respiratory peak value is greater than the first preset peak value, then the candidate respiratory peak value is determined to be a valid respiratory peak value; If the candidate heart rate peak value is greater than the second preset peak value, then the candidate heart rate peak value is determined to be an effective heart rate peak value; A second candidate respiratory rate and / or a second candidate heart rate are determined based on the time interval between adjacent effective respiratory peaks and / or the time interval between adjacent effective heart rate peaks.

9. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 1, characterized in that, The step of processing the acquired raw data to obtain the raw phase sequence includes: The original data is preprocessed to obtain the target complex matrix; the target complex matrix is ​​a two-dimensional complex matrix consisting of the total number of chirps and the total number of distance cells in a single frame. For the target complex matrix, extract the complex number corresponding to each Chirp under the target distance cell, and split the extracted complex number into in-phase time components and quadrature time components; Based on the in-phase timing components and the quadrature timing components, the arctangent operation is used to solve for each Chirp to obtain the original phase sequence.

10. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 9, characterized in that, The step of preprocessing the original data to obtain the target complex matrix includes: The raw data is preprocessed to generate a target three-dimensional echo matrix; the target three-dimensional echo matrix is ​​a three-dimensional complex matrix consisting of the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp; the non-contact detection module includes multiple antennas. The target three-dimensional echo matrix is ​​subjected to antenna weighted coherent accumulation processing to obtain a first two-dimensional complex matrix; the first two-dimensional complex matrix is ​​a complex matrix of the total number of chirps and the total number of range cells in a single frame; The first two-dimensional complex matrix is ​​subjected to two-stage clutter suppression processing to obtain the target complex matrix.

11. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 10, characterized in that, The step of preprocessing the raw data to generate the target three-dimensional echo matrix includes: The original data is parsed according to frame timing rules to generate a first three-dimensional echo matrix; the first three-dimensional echo matrix is ​​a three-dimensional complex matrix with the number of receiving antennas, the total number of chirps in a single frame, and the number of sampling points in a single chirp; The first three-dimensional echo matrix is ​​normalized to obtain the second three-dimensional echo matrix; The target three-dimensional echo matrix is ​​obtained by performing a Hamming windowing operation on the second three-dimensional echo matrix for each chirp of each antenna.

12. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 10, characterized in that, The step of performing weighted coherent accumulation processing on the target three-dimensional echo matrix using an antenna to obtain a first two-dimensional complex matrix includes: Perform a range-dimensional transformation on the target three-dimensional echo matrix for each antenna to obtain the range-dimensional complex spectrum of each antenna; Based on the complex spectrum of each distance dimension, the echo signal-to-noise ratio of each antenna is determined; The weight of each antenna is determined based on the echo signal-to-noise ratio of each antenna; The first two-dimensional complex matrix is ​​determined based on the distance dimension complex spectrum of each antenna and the weight of each antenna.

13. The respiratory and heart rate monitoring method based on variational mode decomposition according to claim 10, characterized in that, The step of performing two-stage clutter suppression processing on the first two-dimensional complex matrix to obtain the target complex matrix includes: Subtract a preset complex number from each complex number in the first two-dimensional complex matrix to obtain a second two-dimensional complex matrix; the preset complex number is obtained by averaging the complex numbers corresponding to each distance unit of the first two-dimensional complex matrix for a preset number of frames. The second two-dimensional complex matrix is ​​filtered using a least mean square adaptive filter to obtain the target complex matrix.

14. The respiratory and heart rate monitoring method based on variational mode decomposition according to any one of claims 1-13, characterized in that, After determining the respiratory data and / or heart rate data based on multiple target intrinsic mode functions, the method further includes: Based on the respiratory data, determine the type of respiratory abnormality and generate early warning information based on the type of respiratory abnormality; and / or, based on the heart rate data, determine the type of heart rate abnormality and generate early warning information based on the type of heart rate abnormality. Display the warning information, and / or send the warning information to a terminal device so that the warning information can be displayed on the terminal device.

15. A respiratory and heart rate monitoring device based on variational mode decomposition, characterized in that, include: The first data processing module is used to process the acquired raw data to obtain the raw phase sequence. The raw data is obtained by the echo signal returned by the detection signal of the non-contact detection module through the target area of ​​the target object. The raw phase sequence is obtained by complex calculation based on the Chirp of each linear frequency modulation signal under the target distance unit. The target distance unit is the distance unit corresponding to the energy peak of the echo signal. The second data processing module is used to perform variational mode decomposition based on the original phase sequence to obtain multiple target intrinsic mode functions; A respiratory and heart rate determination module is used to determine respiratory data and / or heart rate data based on multiple target intrinsic mode functions; The display processing module is used to display the respiratory data and / or the heart rate data, and / or send the respiratory data and / or the heart rate data to a terminal device for displaying the respiratory data and / or heart rate data through the terminal device.

16. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the method as claimed in any one of claims 1 to 14.

17. The electronic device according to claim 16, characterized in that, The electronic device is any one of a monitoring device, a display device, and a cloud processing device.

18. The electronic device according to claim 17, characterized in that, The display device is a mobile terminal or a monitoring terminal with a display screen.

19. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 14.

20. A computer program product containing instructions, characterized in that, When the instructions are executed by a cluster of computer devices, the cluster of computer devices causes the cluster of computer devices to perform the steps of the method as described in any one of claims 1 to 14.

Images