Multi-target vital sign detection method of TDM-MIMO millimeter wave radar

Abstract

The invention discloses a multi-target vital sign detection method for a TDM-MIMO millimeter wave radar, and the method comprises the steps: controlling a transmitting antenna to transmit a signal to a target region in a TDM-MIMO mode of the millimeter wave MIMO radar, controlling a receiving antenna to receive an echo signal reflected by the target region, and carrying out the extraction and recombination of the echo signal to form an array signal; carrying out distance FFT (Fast Fourier Transform) and carrying out static clutter filtering by using a phasor mean cancellation algorithm; performing forward and backward spatial smoothing and spatial spectrum estimation, and performing two-dimensional constant false alarm rate detection and clustering to realize two-dimensional plane positioning of a human body target; carrying out phase unwrapping and smoothing processing to obtain a phase difference value sequence; and finally, using an improved sine algorithm to guide a dung beetle optimization algorithm to carry out iterative optimization on parameters of the variational mode decomposition algorithm, and using the parameters after iterative optimization to carry out variational mode decomposition on the phase difference value sequence to obtain vital sign parameters of each human body target. The method has high environmental adaptability, and the detection reliability and automation degree are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar technology and vital sign detection, and in particular relates to a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar. Background Art

[0002] With the continuous progress of society and science and technology, human vital signs detection technology is also developing. Vital sign parameters are the main criteria for evaluating human physiological conditions, including basic physiological parameters such as heartbeat, respiration, blood pressure and body temperature, among which respiration and heartbeat are the most direct key indicators reflecting vital signs. Human activity detection mainly monitors human respiration and heartbeat, and detection technologies are divided into contact and non-contact types. Among them, contact technology requires direct contact with the human body, but sometimes it may be inconvenient, such as discomfort when wearing or inability to wear under special circumstances; in contrast, non-contact vital signs detection technology mainly uses radar, infrared sensors and computer vision to achieve remote detection, without direct contact with the human body, and has obvious convenience advantages. Millimeter wave radar is increasingly widely used to detect human vital signs due to its advantages such as strong adaptability to the environment, high anti-interference ability and superior measurement accuracy; this technology has shown wide practicality in many fields, including driver fatigue detection, hospital patient monitoring and sleep monitoring.

[0003] In addition, compared with traditional vital signs detection instruments, millimeter wave radar has the advantages of small size, low power consumption, and good comfort. It can protect the privacy of the subject to a certain extent, and can realize all-weather monitoring, and its application prospects are relatively broad. However, the existing millimeter wave radar is not ideal when performing vital signs detection, and it is difficult to automatically separate the heartbeat and breathing signals of the target. The Chinese patent document with application number 202111548515.1 provides a method for detecting vital signs of adjacent multi-targets of millimeter wave MIMO radar under a clutter background, which uses space-time coding, two-dimensional smoothing, and spatial spectrum estimation to achieve two-dimensional plane positioning of multiple targets, and performs variational multimodal decomposition on the vital signs signals and clutter scattering point phase signals corresponding to the separated multiple human targets, thereby obtaining the vital signs parameters corresponding to each human target. However, the size of the penalty factor of the variational modal decomposition of the invention and the number of modal decompositions need to be manually specified. When an inappropriate number of decomposition layers is selected, it is easy to cause over-decomposition and under-decomposition, which interferes with the extraction of breathing and heartbeat signals. Summary of the invention

[0004] The main purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide a multi-target vital sign detection method of TDM-MIMO millimeter wave radar, which has strong environmental adaptability and reliable detection accuracy.

[0005] In order to achieve the above object, the present invention adopts a multi-target vital sign detection method of TDM-MIMO millimeter wave radar, comprising the following steps:

[0006] In the TDM-MIMO mode of the millimeter wave MIMO radar, the transmitting antenna is controlled to transmit a signal to the target area, and the receiving antenna is controlled to receive the echo signal reflected back from the target area, and the echo signal is extracted and reassembled to form an array signal; the target area includes multiple human targets;

[0007] Perform range FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal;

[0008] The filtered signal is smoothed in forward and backward space and the spatial spectrum is estimated, and a two-dimensional constant false alarm rate detection is performed to separate the valid point cloud in the target area; the valid point cloud is clustered to obtain multiple human targets in the target area, and the two-dimensional plane positioning of the human targets is achieved;

[0009] The phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target is selected for phase unwrapping and smoothing to obtain the phase difference sequence;

[0010] The improved sine algorithm is used to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational mode decomposition algorithm, and the iteratively optimized parameters are used to perform variational mode decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

[0011] As a preferred technical solution, the signal transmitted by the transmitting antenna is expressed as:

[0012]

[0013] Among them, x m (t) is the signal transmitted by the mth transmitting antenna, t is the signal transmission time, f c is the starting frequency, T r is the switching time of the transmitting antenna alternately working, β is the linear frequency increment, λ is the wavelength, d m =(m-1)d TX , d TX is the distance between the two transmitting antennas, θ m (TX) is the deviation angle from the mth transmitting antenna to the target area, θ m (t) is the phase noise of the mth transmitting antenna;

[0014] The echo signal reflected from the target area received by the receiving antenna is expressed as:

[0015]

[0016] Among them, x mn(t) is the signal transmitted by the mth transmitting antenna and received by the nth receiving antenna, is the reflection coefficient of the nth receiving antenna, Δt is the delay time for the same transmitted signal to be received by different receiving antennas, and d n =(n-1)d RX , d RX is the distance between the two receiving antennas, θ n (RX) is the arrival angle from the target to the nth receiving antenna, θ mn (t) is the phase noise of the signal transmitted by the mth transmitting antenna at the nth receiving antenna;

[0017] The signal received by the receiving antenna and the baseband signal are mixed through a mixer to obtain an intermediate frequency signal, which is expressed as:

[0018]

[0019] Among them, T x is a random transmitting antenna x, Δθ mn (t) is the phase difference of the signal transmitted by the mth transmitting antenna at the nth receiving antenna.

[0020] As a preferred technical solution, the two-dimensional plane positioning of the human body target is specifically achieved as follows:

[0021] Perform forward and backward spatial smoothing on the filtered signal to obtain the subarray signal Z;

[0022] The spectral function P of the subarray signal Z is obtained by using the spatial spectrum estimation MUSIC (r,θ), expressed as:

[0023]

[0024] Where r is the position corresponding to the target, θ is the azimuth angle corresponding to the target, a(r,θ) is the steering vector of the subarray signal Z, and a(r,θ) H is the conjugate transposed matrix of the sub-array signal Z, U N is the noise subspace obtained after singular value decomposition of the subarray signal Z, For U N The conjugate transposed matrix of ;

[0025] The distance angle is estimated for the result after the spatial spectrum estimation to obtain a rough point cloud map;

[0026] Use two-dimensional constant false alarm rate detection to filter out the noise in the rough point cloud image and select the valid point cloud;

[0027] The filtered valid point clouds are clustered, and each cluster center is taken as a human target to separate multiple human targets and obtain their distances and orientations.

[0028] As a preferred technical solution, the phase difference sequence is obtained as follows:

[0029] The phase signal sequence of the maximum value of the spatial spectrum estimation point in each cluster target is selected and phase unwrapping is performed using differential cross multiplication to obtain the phase information of each human target. It is expressed as:

[0030]

[0031] Wherein, d is the d-th human target, n is the total number of phase signal sequences of the d-th human target, ω[k] is the frequency of the k-th phase signal sequence, Δt is the interval time between phase signal sequences, I[k] is the real part of the k-th phase signal sequence, Q[k] is the imaginary part of the k-th phase signal sequence, I[k-1] is the real part of the k-1-th phase signal sequence, and Q[k-1] is the imaginary part of the k-1-th phase signal sequence;

[0032] Phase information of each human target Perform first-order phase difference to obtain the difference sequence of each target It is expressed as:

[0033]

[0034] in, is the phase information of the n+1th phase signal sequence of the dth human target;

[0035] The difference sequence for each target Savitzky-Golay smoothing filter processing is performed to obtain the phase difference value sequence of each target.

[0036] As a preferred technical solution, the vital sign parameters of each human target are obtained as follows:

[0037] Initialize the parameters of the improved sinusoidal algorithm-guided dung beetle optimization algorithm, the modal decomposition number K, and the range of the penalty factor α; the parameters include the population size, the maximum number of iterations, the upper bound of the optimized dung beetle position, the lower bound of the optimized dung beetle position, and the distribution ratio of different individuals in the population;

[0038] The Bernoulli chaos mapping algorithm is used to initialize the initial position of each dung beetle, which is expressed as:

[0039]

[0040] Among them, z n+1 is the n+1th number in the generated chaotic random sequence, z n is the nth number in the generated chaotic random sequence, β is the mapping parameter;

[0041] The phase difference sequence of each target is subjected to variational mode decomposition using the parameters of the initial position of the dung beetle, and multiple modal components of each individual in the population are obtained.

[0042] Updated the locations of rolling ball dung beetles, breeding ball dung beetles, small dung beetles and stealing dung beetles;

[0043] After updating, the fuzzy entropy of each individual in the population is calculated as the optimized fitness value, and the minimum fuzzy entropy is taken as the current optimal solution X b (t);

[0044] Use adaptive Gauss-Cauchy variation perturbation to perturb the current optimal solution to generate the current new solution H b (t), the perturbation formula is:

[0045]

[0046] in, is the weight coefficient of the Gaussian mutation operator, T max is the maximum number of iterations, Gauss(σ) is the Gaussian mutation operator, is the weight coefficient of the Cauchy mutation operator, cauchy(σ) is the Cauchy variation operator;

[0047] Compare the current new solution H according to the greedy rule b (t) and the current optimal solution X b (t) Perform position update, repeat the above steps until the maximum number of iterations is reached and the optimal position is output; the position update formula is:

[0048]

[0049] Among them, f[H b (t)] is the current new solution H b (t) fitness value, f[X b (t)] is the current optimal solution X b (t) Adaptation value;

[0050] The parameters at the optimal position are used to perform variational mode decomposition to obtain the vital sign parameters of each human target.

[0051] As a preferred technical solution, the phase difference sequence of each target is subjected to variational mode decomposition using the parameters of the initial position of the dung beetle, specifically:

[0052] Initialize the modal components μ k , center frequency ω k , Lagrange multiplier λ and n = 0, select the appropriate modal decomposition number K and penalty factor α;

[0053] Update the modal component μ based on the following formula k , center frequency ω k and the Lagrange multiplier λ, the formula is:

[0054]

[0055] Where n is the number of iterations, ω is the frequency, is the kth modal component in the n+1th iteration, is the ith modal component in the n+1th iteration, is the ith modal component in the nth iteration, is the frequency of the kth modal component at the n+1th iteration, is the Lagrange multiplier of the n+1th iteration, is the Lagrange multiplier of the nth iteration, γ is the noise tolerance, Original signal;

[0056] Repeat the iterative update until the iterative constraint condition is satisfied; the iterative constraint condition is expressed as:

[0057]

[0058] Among them, ε is the set accuracy.

[0059] As a preferred technical solution, the updating formula of the rolling ball dung beetle position is:

[0060]

[0061] Among them, x i (t+1) is the position of the ith rolling ball beetle after the tth iteration update, x i (t) is the position of the ith rolling ball beetle before the tth iteration update, α∈{1,-1} is the direction selection factor, k is the search step length, x i (t-1) is the position of the ith dung beetle before the t-1th iteration update, b is a constant greater than 0 and less than 1, Δx is the difference between the position of the ith dung beetle and the position of the dung beetle with the worst fitness in the current iteration, δ=rand(1) is a random number in the range of 0-1, ST∈(0.5,1] is the global search probability, r 1 is a nonlinear decreasing function, ω t is the inertia weight of the tth iteration, r 2 is a random number in the interval [0,2π], r 3 is a random number in the interval [-2,2], p i (t) is the position of the rolling ball beetle with the minimum fuzzy entropy in the tth iteration, ω max is the maximum value of frequency ω, ωmin is the minimum value of ω, T max is the maximum number of iterations;

[0062] The updating formula of the breeding ball position is:

[0063] B i (t+1)=X * +b 1 ×(B i (t)-Lb * )+b 2 ×(B i (t)-Ub * ),

[0064] Lb * =max(X * ×(1-R),Lb),

[0065] Ub * =min(X * ×(1-R),Ub),

[0066] Among them, B i (t+1) is the position of the i-th breeding ball after the t-th iteration update, X * is the current local optimal position, b 1 and b 2 are two independent random vectors of size 1×D, where D is the dimension, and B i (t) is the position of the i-th breeding ball before updating at the t-th iteration, Lb * is the lower boundary of the female dung beetle's egg-laying area, Ub * The upper boundary of the female dung beetle's egg-laying area. Lb is the lower bound of the optimized dung beetle position, and Ub is the upper bound of the optimized dung beetle position;

[0067] The updating formula of the position of the small dung beetle is:

[0068] xs i (t+1)=xs i (t)+C 1 ×(xs i (t)-Lb b )+C 2 ×(xs i (t)-Ub b ),

[0069] Lb b =max(X b ×(1-R),Lb),

[0070] Ub b =mib(X b×(1-R),Ub),

[0071] Among them, xs i (t+1) is the updated position of the i-th dung beetle in the t-th iteration, xs i (t) is the position of the i-th dung beetle before the t-th iteration update, C 1 is a random number that follows a normal distribution, C 2 is a random vector of (0,1), Lb b is the lower bound of the optimal foraging area, Ub b is the upper bound of the optimal foraging area, X b represents the global optimal position;

[0072] The updating formula of the stealing dung beetle position is:

[0073] xt i (t+1)=X b +S×g×(|xt i (t)-X * |+|xt i (t)-X b |),

[0074] Among them, xt i (t+1) is the updated position of the i-th stealing dung beetle in the t-th iteration, X b is the global optimal position, S is open, g is a normally distributed random vector, xt i (t) is the position of the i-th stealing dung beetle before the update in the t-th iteration, X * is the current local optimal position;

[0075] The current local optimal position is the position where the rolling dung beetle has the best fitness after being updated in the current iteration number;

[0076] The global optimal position is the position of the individual with the best fitness before all individuals in the population are updated in the current iteration.

[0077] As a preferred technical solution, the calculation method of the fuzzy entropy is:

[0078] Reconstruct multiple modal components of each individual in the population to obtain an m-dimensional time series. The reconstruction formula is:

[0079]

[0080] Among them, each modal component is an N-dimensional time series {μ(j), 1≤j≤N}, X(i) is the vector sequence reconstructed from the i-th modal component, μ 0 (i) is the mean of the vector sequence reconstructed from the i-th modal component;

[0081] Calculate the distance between the two modal components of each individual in the population after reconstruction of the m-dimensional time series The formula is:

[0082]

[0083] Where n is the length of the N-dimensional time series, μ(i+k)-μ 0 (i) is the difference between the original value and the average value of the i-th dimension time series, (μ(j+k)-μ 0 (j)) is the difference between the original value and the average value of the j-th dimension time series;

[0084] Introducing membership function Calculate the membership between any two modal components of each individual in the population after reconstructing the m-dimensional time series And the average value of all memberships of the m-dimensional time series reconstructed from any modal component of each individual in the population except itself φ m (z,r), the formula is:

[0085]

[0086] Among them, z is the similarity measurement index, r is the standard deviation of the N-dimensional time series;

[0087] The fuzzy entropy of multiple modal components of each individual in the population is obtained and expressed as:

[0088] FuzzyEn(m,z,r)=lnφ m (z,r)-lnφ m+1 (z,r)

[0089] Among them, φ m+1 (z,r) is the average of all memberships of the m+1-dimensional time series reconstructed from any modal component of each individual in the population except itself.

[0090] On the other hand, the present invention provides a multi-target vital sign detection system of TDM-MIMO millimeter wave radar, including a signal acquisition module, a signal filtering module, a two-dimensional positioning module, a phase calculation module and a parameter acquisition module;

[0091] The signal acquisition module is used to control the transmitting antenna to transmit signals to the target area in the TDM-MIMO mode of the millimeter wave MIMO radar, and control the receiving antenna to receive the echo signal reflected back from the target area, and extract and reassemble it to form an array signal; the target area includes multiple human targets;

[0092] The signal filtering module is used to perform distance FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal;

[0093] The two-dimensional positioning module is used to perform forward and backward spatial smoothing and spatial spectrum estimation on the filtered signal, and perform two-dimensional constant false alarm rate detection to separate the effective point cloud in the target area; cluster the effective point cloud to obtain multiple human targets in the target area, and realize two-dimensional plane positioning of the human targets;

[0094] The phase calculation module is used to select the phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target, perform phase unwrapping and smoothing processing to obtain a phase difference sequence;

[0095] The parameter acquisition module is used to use the improved sine algorithm to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational modal decomposition algorithm, and use the iteratively optimized parameters to perform variational modal decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

[0096] Another aspect of the present invention provides a computer-readable storage medium storing a program, which, when executed by a processor, implements a multi-target vital sign detection method using a TDM-MIMO millimeter wave radar.

[0097] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0098] The present invention provides a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar, which can realize multi-human target positioning and decompose vital sign signals, wherein the variational mode decomposition algorithm of the improved sinusoidal algorithm-guided dung beetle optimization algorithm uses fuzzy entropy as an optimization index, and can automatically iterate the optimal VMD decomposition parameters suitable for separating heartbeat and respiratory signals according to the optimization parameters, thereby making up for the disadvantage that artificially presetting the modal decomposition number and the size of the penalty factor in the variational mode decomposition easily leads to over-decomposition or under-decomposition of the signal, so that the millimeter wave radar human life feature detection has stronger environmental adaptability and effectively improves the reliability and automation of the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0100] Figure 1 The figure is an overall flow chart of a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar in an embodiment of the present invention.

[0101] Figure 2This is an application scenario diagram of a multi-target vital signs detection method of a TDM-MIMO millimeter wave radar in an embodiment of the present invention.

[0102] Figure 3 4 is a diagram showing the effect of filtering a signal in an embodiment of the present invention.

[0103] Figure 4 This is a rendering of two human targets detected by the millimeter-wave MIMO radar in an embodiment of the present invention.

[0104] Figure 5 This is a diagram showing the optimization effect of the target one variational mode decomposition parameter in an embodiment of the present invention after the improved sinusoidal algorithm guides the dung beetle optimization algorithm.

[0105] Figure 6 This is a respiratory and heartbeat parameter diagram of target one in an embodiment of the present invention.

[0106] Figure 7 This is a diagram showing the optimization effect of the target two-variation modal decomposition parameters in an embodiment of the present invention after the improved sinusoidal algorithm guides the dung beetle optimization algorithm.

[0107] Figure 8 This is a respiratory and heartbeat parameter diagram of target two in an embodiment of the present invention.

[0108] Fig. 9 The figure is a schematic diagram of the structure of a multi-target vital sign detection system of a TDM-MIMO millimeter wave radar in an embodiment of the present invention.

[0109] Fig.10 Schematic diagram of the structure of a computer-readable storage medium in an embodiment of the present invention. DETAILED DESCRIPTION

[0110] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.

[0111] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0112] like Figure 1As shown, this embodiment provides a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar, comprising the following steps:

[0113] S1. In the TDM (time division multiplexing)-MIMO mode of the millimeter wave MIMO (multiple input multiple output) radar, control the transmitting antenna to transmit signals to the target area (such as Figure 2 As shown), the receiving antenna is controlled to receive the echo signal reflected back from the target area, and the echo signal is extracted and reassembled to form an array signal; wherein the target area contains multiple human targets to be detected;

[0114] Specifically, the signal transmitted by the transmitting antenna is shown in the following formula (1):

[0115]

[0116] In formula (1), x m (t) is the signal transmitted by the mth transmitting antenna, t is the signal transmission time, f c is the starting frequency, T r is the switching time of the transmitting antenna alternately working, β is the linear frequency increment, λ is the wavelength, d m =(m-1)d TX , d TX is the distance between the two transmitting antennas, θ m (TX) is the deviation angle from the mth transmitting antenna to the target area, θ m (t) is the phase noise of the mth transmitting antenna;

[0117] The echo signal reflected from the target area received by the receiving antenna is shown in formula (2):

[0118]

[0119] In formula (2), x mn (t) is the signal transmitted by the mth transmitting antenna and received by the nth receiving antenna, is the reflection coefficient of the nth receiving antenna, Δt is the delay time for the same transmitted signal to be received by different receiving antennas, and d n =(n-1)d RX , d RX is the distance between the two receiving antennas, θ n (RX) is the arrival angle from the target to the nth receiving antenna, θ mn (t) is the phase noise of the signal transmitted by the mth transmitting antenna at the nth receiving antenna;

[0120] The signal received by the receiving antenna and the baseband signal are mixed through a mixer to obtain an intermediate frequency signal, as shown in formula (3):

[0121]

[0122] In formula (3), T x is a random transmitting antenna x, Δθ mn (t) is the phase difference of the signal transmitted by the mth transmitting antenna at the nth receiving antenna. After obtaining the intermediate frequency signal, it needs to be sorted and compensated to form an array signal.

[0123] S2, performing range FFT on the array signal and using the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal;

[0124] The effect after distance FFT and static clutter filtering is as follows Figure 3 shown.

[0125] S3, performing forward and backward spatial smoothing and spatial spectrum estimation on the filtered signal, and performing two-dimensional constant false alarm rate detection to separate the valid point cloud in the target area; clustering the valid point cloud to obtain multiple human targets in the target area, and realizing two-dimensional plane positioning of the human targets;

[0126] Specifically, the operation of step S3 is:

[0127] Perform forward and backward spatial smoothing on the filtered signal to obtain the subarray signal Z;

[0128] The spectral function P of the subarray signal Z is obtained by using the spatial spectrum estimation MUSIC (r,θ), as shown in formula (4):

[0129]

[0130] In formula (4), r is the position corresponding to the target, θ is the azimuth angle corresponding to the target, a(r,θ) is the steering vector of the subarray signal Z, and a(r,θ) H is the conjugate transposed matrix of the sub-array signal Z, U N is the noise subspace obtained after singular value decomposition of the subarray signal Z, For U N The conjugate transposed matrix of ;

[0131] The distance angle is estimated for the result after the spatial spectrum estimation to obtain a rough point cloud map;

[0132] Use two-dimensional constant false alarm rate detection to filter out the noise in the rough point cloud image and select the valid point cloud;

[0133] The filtered valid point clouds are clustered, and each cluster center is taken as a human target to separate multiple human targets and obtain their distances and orientations. Figure 4 A schematic diagram showing the effect of detecting two human targets in the target area.

[0134] S4, selecting the phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target, performing phase unwrapping and smoothing processing to obtain a phase difference sequence;

[0135] Specifically, the operation of step S4 is:

[0136] The phase signal sequence of the maximum point of the spatial spectrum estimation in each cluster target is selected and phase unwrapped using differential cross multiplication (DACM) to obtain the phase information of each human target. Such as formula (5):

[0137]

[0138] In formula (5), d is the d-th human target, n is the total number of phase signal sequences of the d-th human target, ω[k] is the frequency of the k-th phase signal sequence, Δt is the interval time between phase signal sequences, I[k] is the real part of the k-th phase signal sequence, Q[k] is the imaginary part of the k-th phase signal sequence, I[k-1] is the real part of the k-1-th phase signal sequence, and Q[k-1] is the imaginary part of the k-1-th phase signal sequence;

[0139] Phase information of each human target Perform first-order phase difference to obtain the difference sequence of each target As shown in formula (6):

[0140]

[0141] In formula (6), is the phase information of the n+1th phase signal sequence of the dth human target;

[0142] The difference sequence for each target Savitzky-Golay smoothing filter processing is performed to obtain the phase difference value sequence of each target.

[0143] S5. Use the improved sinusoidal algorithm guided dung beetle optimization algorithm (MSADBO) to iteratively optimize the parameters of the variational mode decomposition algorithm (VMD), and use the iteratively optimized parameters to perform variational mode decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

[0144] Specifically, the specific optimization steps of the improved sinusoidal algorithm guided dung beetle optimization algorithm (MSADBO) in step S5 are:

[0145] S501, initializing the parameters of the improved sinusoidal algorithm-guided dung beetle optimization algorithm, the modal decomposition number K and the range of the penalty factor α; wherein the parameters include the population size, the maximum number of iterations, the upper bound of the optimized dung beetle position, the lower bound of the optimized dung beetle position and the distribution ratio of different individuals in the population; the individuals in the population include four types: rolling dung beetles, female egg-laying dung beetles, small dung beetles and stealing dung beetles.

[0146] S502, using the Bernoulli chaos mapping algorithm (Bernoulli algorithm) to initialize the initial position of each dung beetle, as shown in formula (7):

[0147]

[0148] In formula (7), z n+1 is the n+1th number in the generated chaotic random sequence, z n is the nth number in the generated chaotic random sequence, β is the mapping parameter;

[0149] S503, using the parameters of the initial position of the dung beetle to perform variational modal decomposition on the phase difference sequence of each target, to obtain multiple modal components of each individual in the population;

[0150] Among them, the specific process of variational mode decomposition is:

[0151] (1) Initialization modal component (IMF) μ k , center frequency ω k , Lagrange multiplier λ and n = 0, select the appropriate modal decomposition number K and penalty factor α, etc.;

[0152] (2) The following formula updates the modal component (IMF) μ k , center frequency ω k and the Lagrange multiplier λ, as shown in (8), (9), and (10):

[0153]

[0154] In formulas (8), (9), and (10), n is the number of iterations, ω is the frequency, is the kth modal component in the n+1th iteration, is the ith modal component in the n+1th iteration, is the ith modal component in the nth iteration, is the frequency of the kth modal component at the n+1th iteration, is the Lagrange multiplier of the n+1th iteration, is the Lagrange multiplier of the nth iteration, γ is the noise tolerance, Original signal;

[0155] (3) Repeat the iterative update until the iterative constraint condition is met, as shown in the following formula (11):

[0156]

[0157] Among them, ε is the set accuracy.

[0158] S504, updating the position of the rolling ball dung beetle, the breeding ball position, the small dung beetle position and the stealing dung beetle position;

[0159] Among them, the update formula of the rolling dung beetle position is shown in formula (12):

[0160]

[0161]

[0162] Among them, x i (t+1) is the position of the ith rolling ball beetle after the tth iteration update, x i (t) is the position of the ith rolling ball beetle before the tth iteration update, α∈{1,-1} is the direction selection factor, k is the search step length, x i (t-1) is the position of the ith dung beetle before the t-1th iteration update, b is a constant greater than 0 and less than 1, Δx is the difference between the position of the ith dung beetle and the position of the dung beetle with the worst fitness in the current iteration, δ=rand(1) is a random number in the range of 0-1, ST∈(0.5,1] is the global search probability; r 1 is a nonlinear decreasing function, which decreases linearly with the increase of iteration number; ω t is the inertia weight of the tth iteration, r 2 is a random number in the interval [0,2π], r 3 is a random number in the interval [-2,2], p i (t) is the position of the rolling ball beetle with the minimum fuzzy entropy in the tth iteration; ω max is the maximum value of ω, a larger ω value is helpful for global search; ω min is the minimum value of frequency ω, and a smaller ω is conducive to local development; T max is the maximum number of iterations;

[0163] The updating formula of the breeding ball position is shown in formula (15):

[0164] B i (t+1)=X * +b 1 ×(B i (t)-Lb * )+b 2 ×(B i (t)-Ub* )(15),

[0165] Lb * =max(X * ×(1-R),Lb)(16),

[0166] Ub * =min(X*×(1-R),Ub)(17),

[0167] Among them, B i (t+1) is the position of the i-th breeding ball after the t-th iteration update, X * is the current local optimal position, b 1 and b 2 are two independent random vectors of size 1×D, where D is the dimension, and B i (t) is the position of the i-th breeding ball before updating at the t-th iteration. Equations (16) and (17) are used to simulate the egg-laying area of ​​female dung beetles. Lb * is the lower boundary of the female dung beetle's egg-laying area, Ub * The upper boundary of the female dung beetle's egg-laying area. Lb is the lower bound of the optimized dung beetle position, and Ub is the upper bound of the optimized dung beetle position;

[0168] The update formula of the dung beetle position is shown in formula (18):

[0169] xs i (t+1)=xs i (t)+C 1 ×(xs i (t)-Lb b )+C 2 ×(xs i (t)-Ub b )(18),

[0170] Lb b =max(X b ×(1-R),Lb)(19),

[0171] Ub b =min(X b ×(1-R),Ub)(20),

[0172] Among them, xs i (t+1) is the updated position of the i-th dung beetle in the t-th iteration, xs i (t) is the position of the i-th dung beetle before the t-th iteration update, C 1 is a random number that follows a normal distribution, C 2is a random vector of (0,1), and equations (19) and (20) are used to define the boundary of the optimal foraging area. b is the lower bound of the optimal foraging area, Ub b is the upper bound of the optimal foraging area, X b represents the global optimal position;

[0173] The updating formula of stealing dung beetle position is shown in formula (21):

[0174] xt i (t+1)=X b +S×g×(|xt i (t)-X * |+|xt i (t)-X b |)(21),

[0175] Among them, xt i (t+1) is the updated position of the i-th stealing dung beetle in the t-th iteration, X b is the global optimal position, S is open, g is a normally distributed random vector, xt i (t) is the position of the i-th stealing dung beetle before the update in the t-th iteration, X * is the current local best position.

[0176] Among them, the global best position refers to the individual position with the best fitness value (minimum fuzzy quotient) before updating among all individuals (four types of dung beetles) in the population, and the local best position refers to the position with the best fitness after updating of the rolling ball dung beetle in the current iteration number.

[0177] S505: After updating, calculate the fuzzy entropy of each individual in the population as the optimized fitness value, and take out the minimum fuzzy entropy as the current optimal solution X b (t);

[0178] Fuzzy entropy is a parameter to measure the complexity of a sequence. The more complex the sequence, the more components it contains, and the higher the fuzzy entropy. On the contrary, the simpler the sequence, the fewer components it contains, and the smaller the fuzzy entropy. Therefore, the minimum fuzzy entropy is used as the optimization index for VMD decomposition. The fuzzy entropy is calculated for the respiratory and heart rate modal components after VMD decomposition. The parameters of VMD decomposition are continuously iterated and optimized, and the parameters that obtain the minimum fuzzy entropy result are selected, so as to obtain the modal components of respiration and heartbeat with fewer components, so as to obtain a cleaner and clearer heartbeat and respiration spectrum, and better separate the vital sign signals. The calculation method of fuzzy entropy is as follows:

[0179] Reconstruct the multiple modal components of each individual in the population, as shown in formula (22), and obtain the m-dimensional time series:

[0180]

[0181] Wherein, each modal component is an N-dimensional time series {μ(j), 1≤j≤N}, X(i) is the vector sequence reconstructed from the i-th modal component, and the specific operation is to start from the first point of the N-dimensional time series, take m consecutive points as the first vector, and then repeat the above operation from the second point, and so on; μ 0 (i) The mean of the new vector formed by each continuous m points is the mean of the vector sequence reconstructed from the i-th modal component, minus μ 0 (i) The purpose is to decentralize;

[0182] Calculate the distance between the two modal components of each individual in the population after reconstruction of the m-dimensional time series It is used to represent the maximum absolute value of the original difference between the two, as shown in formula (23):

[0183]

[0184] Where n is the length of the N-dimensional time series, μ(i+k)-μ 0 (i) is the difference between the original value and the average value of the i-th dimension time series, (μ(j+k)-μ 0 (j)) is the difference between the original value and the average value of the j-th dimension time series;

[0185] Introducing membership function Calculate the membership between any two modal components of each individual in the population after reconstructing the m-dimensional time series And the average value of all memberships of the m-dimensional time series reconstructed from any modal component of each individual in the population except itself φ m (z, r), as shown in formulas (24) and (25):

[0186]

[0187]

[0188] Among them, z is the similarity measurement index, r is the standard deviation of the N-dimensional time series;

[0189] The fuzzy entropy of multiple modal components of each individual in the population is obtained, as shown in formula (26):

[0190] FuzzyEn(m,z,r)=lnφ m (z,r)-lnφ m+1 (z,r) (26),

[0191] Among them, φ m+1(z,r) is the average of all memberships of the m+1-dimensional time series reconstructed from any modal component of each individual in the population except itself.

[0192] S506, using adaptive Gauss-Cauchy variation disturbance to perturb the current optimal solution to generate the current new solution H b (t), the perturbation formula is shown in formula (27):

[0193]

[0194] in, is the weight coefficient of the Gaussian mutation operator, T max is the maximum number of iterations, Gauss(σ) is the Gaussian mutation operator, is the weight coefficient of the Cauchy mutation operator, cauchy(σ) is the Cauchy variation operator;

[0195] S507, compare the current new solution H according to the greedy rule b (t) and the current optimal solution X b (t) Update the position and repeat the above steps until the maximum number of iterations is reached and the optimal position is output; the position update is shown in formula (28):

[0196]

[0197] Among them, f[H b (t)] is the current new solution H b (t) fitness value, f[X b (t)] is the current optimal solution X b (t) Adaptation value;

[0198] S508. Perform variational mode decomposition using the parameters at the optimal position to obtain vital sign parameters of each human target.

[0199] Figure 5 Shown Figure 4 The change curve of the fitness value (i.e. the size of fuzzy entropy), the number of modal decompositions and the penalty factor of the target-variational modal decomposition parameters in the iterative process of the dung beetle optimization algorithm guided by the improved sine algorithm. It can be seen from the figure that during the 15 iterative optimization of the target-variational modal decomposition parameters, the fitness curve (fuzzy entropy) gradually decreases, and then the optimal solution of the VMD decomposition parameters with fuzzy entropy as the optimization indicator is obtained, the number of modal decompositions is 8, and the penalty factor is 1272.

[0200] Figure 6The optimal solution of VMD decomposition parameters is applied to perform VMD decomposition on the time phase sequence of target one after phase unwrapping to obtain the vital sign parameters of target one. From the figure, we can get a relatively clean and clear heartbeat and breathing time domain waveform and spectrum, and clearly distinguish the breathing frequency of 0.25Hz and the heart rate of 1.21Hz.

[0201] Similarly, Figure 7 and Figure 8 for Figure 4 The processing result of target 2; among them, Figure 7 It shows that in the process of 15 iterative optimizations of VMD parameters for target 2, the fitness curve (fuzzy entropy) gradually decreases, and then the optimal solution of VMD decomposition parameters with fuzzy entropy as the optimization index is obtained, the modal decomposition number is 5, and the penalty factor is 2976; based on the optimal solution of VMD decomposition parameters, VMD decomposition is performed on the time phase sequence after phase unwrapping of target 2 to obtain the vital sign parameters of target 2. From the figure, a clearer heartbeat and breathing time domain waveform and spectrum can be obtained, and the breathing frequency is clearly distinguished as 0.23Hz and the heart rate frequency is 1.23Hz, as shown in Figure 8 shown.

[0202] In summary, in order to overcome the disadvantages that the penalty factor of variational mode decomposition and the number of modal decomposition layers need to be manually selected, the present invention provides a multi-target vital sign detection method of TDM-MIMO millimeter wave radar: first, the target is detected by TDM-MIMO, and the received array signal is subjected to distance dimension FFT at the receiving end and static clutter is filtered out by using the phasor mean cancellation method; then, the signal after static clutter filtering is subjected to forward and backward spatial smoothing and spatial spectrum estimation to reduce noise and coherent interference, and the spatial spectrum estimation result is subjected to two-dimensional CFAR and clustering to realize two-dimensional plane positioning of human targets; then, the phase signal sequence of the maximum point of spatial spectrum estimation in each clustered target is selected, and the vital sign signal is separated by the MSADBO-VMD method, and the modal decomposition layer number and the penalty factor are iteratively optimized by MSADBO (improved sine algorithm guided dung beetle optimization algorithm), and the iteratively optimized parameters are used for VMD decomposition to avoid over-decomposition and under-decomposition caused by inappropriate decomposition layers, thereby accurately extracting vital sign signals.

[0203] It should be noted that, for the sake of convenience, the aforementioned method embodiments are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited to the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously.

[0204] Based on the same idea as the multi-target vital signs detection method of a TDM-MIMO millimeter wave radar in the above-mentioned embodiment, the present invention also provides a multi-target vital signs detection system of a TDM-MIMO millimeter wave radar, which can be used to execute the multi-target vital signs detection method of the TDM-MIMO millimeter wave radar. For ease of explanation, in the structural schematic diagram of an embodiment of a multi-target vital signs detection system of a TDM-MIMO millimeter wave radar, only the parts related to the embodiment of the present invention are shown. Those skilled in the art can understand that the illustrated structure does not constitute a limitation on the device, and may include more or fewer components than shown in the diagram, or combine certain components, or arrange the components differently.

[0205] like Fig. 9 As shown, another embodiment of the present invention provides a multi-target vital sign detection system of a TDM-MIMO millimeter wave radar, including a signal acquisition module, a signal filtering module, a two-dimensional positioning module, a phase calculation module and a parameter acquisition module;

[0206] Among them, the signal acquisition module is used to control the transmitting antenna to transmit signals to the target area in the TDM-MIMO mode of the millimeter wave MIMO radar, and at the same time control the receiving antenna to receive the echo signal reflected back from the target area, extract and reassemble to form an array signal; the target area includes multiple human targets;

[0207] The signal filtering module is used to perform distance FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal;

[0208] The two-dimensional positioning module is used to perform forward and backward spatial smoothing and spatial spectrum estimation on the filtered signal, and to perform two-dimensional constant false alarm rate detection to separate the effective point cloud in the target area; cluster the effective point cloud to obtain multiple human targets in the target area, and realize the two-dimensional plane positioning of the human target;

[0209] The phase calculation module is used to select the phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target, perform phase unwrapping and smoothing processing to obtain the phase difference sequence;

[0210] The parameter acquisition module is used to use the improved sine algorithm to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational mode decomposition algorithm, and use the iteratively optimized parameters to perform variational mode decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

[0211] It should be noted that a multi-target vital signs detection system of a TDM-MIMO millimeter-wave radar of the present invention corresponds one-to-one to a multi-target vital signs detection method of a TDM-MIMO millimeter-wave radar of the present invention. The technical features and beneficial effects described in the embodiment of the multi-target vital signs detection method of the above-mentioned TDM-MIMO millimeter-wave radar are applicable to the embodiment of a multi-target vital signs detection system of a TDM-MIMO millimeter-wave radar. For specific contents, please refer to the description in the embodiment of the method of the present invention. No further details will be given here. This is hereby declared.

[0212] In addition, in the implementation of a multi-target vital signs detection system of a TDM-MIMO millimeter-wave radar in the above-mentioned embodiment, the logical division of each program module is only an example. In actual applications, the above-mentioned functions can be assigned to different program modules as needed, for example, for the convenience of corresponding hardware configuration requirements or software implementation. That is, the internal structure of the multi-target vital signs detection system of the TDM-MIMO millimeter-wave radar is divided into different program modules to complete all or part of the functions described above.

[0213] like Fig.10 As shown, in one embodiment, a computer-readable storage medium is provided, in which a program is stored in a memory. When the program is executed by a processor, a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar is implemented, specifically:

[0214] In the TDM-MIMO mode of the millimeter wave MIMO radar, the transmitting antenna is controlled to transmit signals to the target area, and the receiving antenna is controlled to receive the echo signals reflected back from the target area, and extract and reassemble them to form an array signal; the target area includes multiple human targets;

[0215] Perform range FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal;

[0216] The filtered signal is smoothed in forward and backward space and the spatial spectrum is estimated, and a two-dimensional constant false alarm rate detection is performed to separate the valid point cloud in the target area; the valid point cloud is clustered to obtain multiple human targets in the target area, and the two-dimensional plane positioning of the human targets is achieved;

[0217] The phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target is selected for phase unwrapping and smoothing to obtain the phase difference sequence;

[0218] The improved sine algorithm is used to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational mode decomposition algorithm, and the iteratively optimized parameters are used to perform variational mode decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

[0219] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0220] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0221] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A multi-target vital sign detection method using TDM-MIMO millimeter wave radar, characterized in that: The steps include: In the TDM-MIMO mode of the millimeter wave MIMO radar, the transmitting antenna is controlled to transmit a signal to the target area, and the receiving antenna is controlled to receive the echo signal reflected back from the target area, and the echo signal is extracted and reassembled to form an array signal; the target area includes multiple human targets; Perform range FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal; The filtered signal is smoothed in forward and backward space and the spatial spectrum is estimated, and a two-dimensional constant false alarm rate detection is performed to separate the valid point cloud in the target area; the valid point cloud is clustered to obtain multiple human targets in the target area, and the two-dimensional plane positioning of the human targets is achieved; The phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target is selected for phase unwrapping and smoothing to obtain the phase difference sequence; The improved sine algorithm is used to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational mode decomposition algorithm, and the iteratively optimized parameters are used to perform variational mode decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

2. The multi-target vital sign detection method of TDM-MIMO millimeter wave radar according to claim 1, characterized in that: The signal transmitted by the transmitting antenna is expressed as: Among them, x m (t) is the signal transmitted by the mth transmitting antenna, t is the signal transmission time, f c is the starting frequency, T r is the switching time of the transmitting antenna alternately working, β is the linear frequency increment, λ is the wavelength, d m =(m-1)d TX , d TX is the distance between the two transmitting antennas, θ m (TX) is the deviation angle from the mth transmitting antenna to the target area, θ m (t) is the phase noise of the mth transmitting antenna; The echo signal reflected from the target area received by the receiving antenna is expressed as: Among them, x mn (t) is the signal transmitted by the mth transmitting antenna and received by the nth receiving antenna, is the reflection coefficient of the nth receiving antenna, Δt is the delay time for the same transmitted signal to be received by different receiving antennas, and d n =(n-1)d RX , d RX is the distance between the two receiving antennas, θ n (RX) is the arrival angle from the target to the nth receiving antenna, θ mn (t) is the phase noise of the signal transmitted by the mth transmitting antenna at the nth receiving antenna; The signal received by the receiving antenna and the baseband signal are mixed through a mixer to obtain an intermediate frequency signal, which is expressed as: Among them, T x is a random transmitting antenna x, Δθ mn (t) is the phase difference of the signal transmitted by the mth transmitting antenna at the nth receiving antenna.

3. The multi-target vital sign detection method of a TDM-MIMO millimeter wave radar according to claim 1, characterized in that: The two-dimensional plane positioning of the human body target is specifically achieved as follows: Perform forward and backward spatial smoothing on the filtered signal to obtain the subarray signal Z; The spectral function P of the subarray signal Z is obtained by using the spatial spectrum estimation MUSIC (r,θ), expressed as: Among them, r is the position corresponding to the target, r is the azimuth corresponding to the target, a(e,θ) is the steering vector of the subarray signal Z, and a(r,θ) H is the conjugate transposed matrix of the sub-array signal Z, U N is the noise subspace obtained after singular value decomposition of the subarray signal Z, For U N The conjugate transposed matrix of ; The distance angle is estimated for the result after the spatial spectrum estimation to obtain a rough point cloud map; Use two-dimensional constant false alarm rate detection to filter out the noise in the rough point cloud image and select the valid point cloud; The filtered valid point clouds are clustered, and each cluster center is taken as a human target to separate multiple human targets and obtain their distances and orientations.

4. The multi-target vital sign detection method of a TDM-MIMO millimeter wave radar according to claim 1, characterized in that: The phase difference sequence is obtained as follows: The phase signal sequence of the maximum value of the spatial spectrum estimation point in each cluster target is selected and phase unwrapping is performed using differential cross multiplication to obtain the phase information of each human target. It is expressed as: Wherein, d is the d-th human target, n is the total number of phase signal sequences of the d-th human target, ω[k] is the frequency of the k-th phase signal sequence, Δt is the interval time between phase signal sequences, I[k] is the real part of the k-th phase signal sequence, Q[k] is the imaginary part of the k-th phase signal sequence, I[k-1] is the real part of the k-1-th phase signal sequence, and Q[k-1] is the imaginary part of the k-1-th phase signal sequence; Phase information of each human target Perform first-order phase difference to obtain the difference sequence of each target It is expressed as: in, is the phase information of the n1th phase signal sequence of the dth human target; The difference sequence for each target The Savitzky-Golay smoothing filter is performed to obtain the phase difference value sequence of each target.

5. The multi-target vital sign detection method of TDM-MIMO millimeter wave radar according to claim 1, characterized in that: The vital sign parameters of each human target are obtained as follows: Initialize the parameters of the improved sinusoidal algorithm-guided dung beetle optimization algorithm, the modal decomposition number K, and the range of the penalty factor α; the parameters include the population size, the maximum number of iterations, the upper bound of the optimized dung beetle position, the lower bound of the optimized dung beetle position, and the distribution ratio of different individuals in the population; The Bernoulli chaos mapping algorithm is used to initialize the initial position of each dung beetle, which is expressed as: Among them, z n+1 is the n+1th number in the generated chaotic random sequence, z n is the nth number in the generated chaotic random sequence, β is the mapping parameter; The phase difference sequence of each target is subjected to variational mode decomposition using the parameters of the initial position of the dung beetle, and multiple modal components of each individual in the population are obtained. Updated the locations of rolling ball dung beetles, breeding ball dung beetles, small dung beetles and stealing dung beetles; After updating, the fuzzy entropy of each individual in the population is calculated as the optimized fitness value, and the minimum fuzzy entropy is taken as the current optimal solution X b (t); Use adaptive Gauss-Cauchy variation perturbation to perturb the current optimal solution to generate the current new solution H b (t), the perturbation formula is: in, is the weight coefficient of the Gaussian mutation operator, is the maximum number of iterations, Gauss(σ) is the Gaussian mutation operator, is the weight coefficient of the Cauchy mutation operator, is the Cauchy mutation operator; Compare the current new solution H according to the greedy rule b (t) and the current optimal solution X b (t) Perform position update, repeat the above steps until the maximum number of iterations is reached and the optimal position is output; the position update formula is: Among them, f[H b (t)] is the current new solution H b (t) fitness value, f[X b (t)] is the current optimal solution X b (t) Adaptation value; The parameters at the optimal position are used to perform variational mode decomposition to obtain the vital sign parameters of each human target.

6. The multi-target vital sign detection method of TDM-MIMO millimeter wave radar according to claim 5, characterized in that: The variational mode decomposition of the phase difference sequence of each target using the parameters of the initial position of the dung beetle is specifically as follows: Initialize the modal components μ k , center frequency ω k , Lagrange multiplier λ and n = 0, select the appropriate modal decomposition number K and penalty factor α; Update the modal component μ based on the following formula k , center frequency ω k and the Lagrange multiplier λ, the formula is: Where n is the number of iterations, ω is the frequency, is the kth modal component in the n+1th iteration, is the i-th modal component in the nn1-th iteration, is the ith modal component in the nth iteration, is the frequency of the kth modal component at the n+1th iteration, is the Lagrange multiplier of the n+1th iteration, is the Lagrange multiplier of the nth iteration, γ is the noise tolerance, Original signal; Repeat the iterative update until the iterative constraint condition is satisfied; the iterative constraint condition is expressed as: Among them, ε is the set accuracy.

7. The multi-target vital sign detection method of TDM-MIMO millimeter wave radar according to claim 5, characterized in that: The updating formula of the rolling ball dung beetle position is: Among them, x i (t+1) is the position of the ith rolling ball beetle after the tth iteration update, x i (t) is the position of the ith rolling ball beetle before the tth iteration update, α∈{1,-1} is the direction selection factor, k is the search step length, x i (t-1) is the position of the ith dung beetle before the t-1th iteration update, b is a constant greater than 0 and less than 1, Δx is the difference between the position of the ith dung beetle and the position of the dung beetle with the worst fitness in the current iteration, δ=rand(1 is a random number in the range of 0-1, ST∈(0.5,1 is the global search probability, r1 is a nonlinear decreasing function, ω t is the inertia weight of the t-th iteration, r2 is a random number in the interval [0,2π], r3 is a random number in the interval [-2,2], p i (t) is the position of the rolling ball beetle with the minimum fuzzy entropy in the tth iteration, ω max is the maximum value of frequency ω, ω min is the minimum value of ω, T max is the maximum number of iterations; The updating formula of the breeding ball position is: B i (t+1)=X * +b1×(B i (t)-Lb * )+b2×(B i (t)-Ub * ), Lb * =max(X * ×(1),b), Ub * =min(X * ×(1),b), Among them, B i (t+1) is the position of the i-th breeding ball after the t-th iteration update, X * is the current local optimal position, b1 and b2 are two independent random vectors of size 1, D is the dimension, B i (t) is the position of the i-th breeding ball before updating at the t-th iteration, Lb * is the lower boundary of the female dung beetle's egg-laying area, Ub * The upper boundary of the female dung beetle's egg-laying area. Lb is the lower bound of the optimized dung beetle position, and Ub is the upper bound of the optimized dung beetle position; The updating formula of the position of the small dung beetle is: xs i (t+1)=xs i (t)+C1×(xs i (t)-Lb b )+C2×(xs i (t)-Ub b ), Lb b =max(X b ×(1),b), Ub b =min(X b ×(1),b), Among them, xs i (t+1) is the updated position of the i-th dung beetle in the t-th iteration, xs i (t) is the position of the i-th dung beetle before the t-th iteration update, C1 is a random number that obeys the normal distribution, C2 is a random vector of (0,1), Lb b is the lower bound of the optimal foraging area, Ub b is the upper bound of the optimal foraging area, X b represents the global optimal position; The updating formula of the stealing dung beetle position is: xt i (t+1)=X b +S×g×(|xt i (t)-X * |+|xt i (t)-X b |, Among them, xt i (t+1) is the updated position of the i-th stealing dung beetle in the t-th iteration, X b is the global optimal position, S is open, g is a normally distributed random vector, xt i (t) is the position of the i-th stealing dung beetle before the update in the t-th iteration, X * is the current local optimal position; The current local optimal position is the position where the rolling dung beetle has the best fitness after being updated in the current iteration number; The global optimal position is the position of the individual with the best fitness before all individuals in the population are updated in the current iteration.

8. The multi-target vital sign detection method of TDM-MIMO millimeter wave radar according to claim 5, characterized in that: The calculation method of the fuzzy entropy is: Reconstruct multiple modal components of each individual in the population to obtain an m-dimensional time series. The reconstruction formula is: Wherein, each modal component is an N-dimensional time series {μ(j), 1≤j≤N}, X(i) is the vector sequence reconstructed from the i-th modal component, and μ0(i) is the mean of the vector sequence reconstructed from the i-th modal component; Calculate the distance between the two modal components of each individual in the population after reconstruction of the m-dimensional time series The formula is: Where n is the length of the N-dimensional time series, μ(i+k)-μ0(i) is the difference between the original value and the average value of the i-dimensional time series, and (μ(j+k)-μ0(j)) is the difference between the original value and the average value of the j-dimensional time series; Introducing membership function Calculate the membership between any two modal components of each individual in the population after reconstructing the m-dimensional time series And the average value of all memberships of the m-dimensional time series reconstructed from any modal component of each individual in the population except itself φ m (z,r), the formula is: Among them, z is the similarity measurement index, r is the standard deviation of the N-dimensional time series; The fuzzy entropy of multiple modal components of each individual in the population is obtained and expressed as: FuzzyEn(m,z,r)=lnφ m (z,r)-lnφ m+1 (z,r), Among them, φ m+1 (z,r) is the average of all memberships of the m+1-dimensional time series reconstructed from any modal component of each individual in the population except itself.

9. A multi-target vital sign detection system using TDM-MIMO millimeter wave radar, characterized in that: A multi-target vital sign detection method of a TDM-MIMO millimeter wave radar applied to any one of claims 1-8, comprising a signal acquisition module, a signal filtering module, a two-dimensional positioning module, a phase calculation module and a parameter acquisition module; The signal acquisition module is used to control the transmitting antenna to transmit signals to the target area in the TDM-MIMO mode of the millimeter wave MIMO radar, and control the receiving antenna to receive the echo signal reflected back from the target area, and extract and reassemble it to form an array signal; the target area includes multiple human targets; The signal filtering module is used to perform distance FFT on the array signal and use the phasor mean cancellation algorithm to filter out static clutter to obtain a filtered signal; The two-dimensional positioning module is used to perform forward and backward spatial smoothing and spatial spectrum estimation on the filtered signal, and perform two-dimensional constant false alarm rate detection to separate the effective point cloud in the target area; cluster the effective point cloud to obtain multiple human targets in the target area, and realize two-dimensional plane positioning of the human target; The phase calculation module is used to select the phase signal sequence of the maximum value point of the spatial spectrum estimation in each cluster target, perform phase unwrapping and smoothing processing to obtain a phase difference sequence; The parameter acquisition module is used to use the improved sine algorithm to guide the dung beetle optimization algorithm to iteratively optimize the parameters of the variational modal decomposition algorithm, and use the iteratively optimized parameters to perform variational modal decomposition on the phase difference sequence to obtain the vital sign parameters of each human target.

10. A computer-readable storage medium storing a program, characterized in that: When the program is executed by a processor, a multi-target vital sign detection method of a TDM-MIMO millimeter wave radar as described in any one of claims 1 to 8 is implemented.

Images