Vital sign detection method based on MIMO rotary radar
Through MIMO rotary radar, the original data is processed in multi-dimensionally and angle correction, which solves the problem of insufficient radar detection range, realizes all-round dynamic target positioning and vital sign detection, and improves detection accuracy.
Patent Information
- Application Number
- CN202411160091.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The radar detection range is insufficient, resulting in the problem of low accuracy of dynamic target positioning in the omnidirectional range.
Using a method based on MIMO rotary radar, the original radar data matrix is processed in multi-dimensionally, the position information of the dynamic target is obtained, and the angle of the dynamic target is corrected by the angle of the static reference object to achieve all-round detection of vital signs.
All-round detection around the radar is achieved, which reduces the problem of less radar directional detection range, avoids interference from static objects and multipaths on dynamic target positioning, and improves the accuracy of life signal detection.
Smart Images

Figure CN119014844A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of radar technology, and more specifically, to a vital sign detection method based on MIMO rotating radar. Background Art
[0002] With the rapid development of the economy, people pay more and more attention to their own health. In the medical field, breathing and heart rate are two important indicators of physical health. Abnormal breathing patterns may be caused by certain diseases, such as pneumonia, asthma and other diseases, and sleep breathing disorders caused by complications such as heart disease and stroke are a more serious problem. The breathing disorder may cause sleep apnea and the best rescue time may be lost due to untimely rescue. Heart rate variability is an important evaluation criterion for doctors to evaluate the physical condition of patients. It can reflect a person's heart health. For example, a fast heartbeat can be caused by excitement, exercise, drinking, etc., or by shock, heart failure and other diseases.
[0003] At present, common vital sign detection is mainly divided into two methods: contact and non-contact. Compared with non-contact detection, contact vital sign detection has limitations when measuring patients with skin diseases, and long-term vital sign monitoring will make patients feel uncomfortable. Among non-contact sensors, radar is very sensitive to the slight vibration of objects, and compared with other non-contact sensors such as optical sensors and infrared sensors, radar is not affected by light and temperature. Therefore, radar plays a vital role in the field of non-contact vital sign detection.
[0004] Different from other radars, the frequency-modulated continuous wave (FMCW) radar obtains the target's breathing and heartbeat data by extracting the phase changes caused by breathing and heartbeat. However, before that, the target's position needs to be obtained first in order to more accurately extract the target's phase changes and avoid interference from other targets and environmental noise. However, when the human target is not within the detection range, the radar will not be able to locate the human body and obtain vital signs, so the detection of the target is affected by the radar's detection range.
[0005] In existing research, methods to solve the limited detection range of radar can be divided into the following categories: The first category is to increase the radar's transmission power or increase the antenna's wave width. This work can increase the radar's detection range by increasing the radar's detection distance and angle, but this method will increase the hardware complexity. The second category is to use distributed radar for measurement. This radar system can break through the limitations of a single radar in power, antenna size and coverage. However, compared with a single radar, distributed radars must also avoid mutual interference between multiple radars while considering the increase in hardware costs. The third category is to fix the radar on a drone. This method can make the detection range of a single radar cover the entire room through the rotation characteristics of the drone, and realize imaging and positioning of objects in the room. However, due to the limitations of the drone's rotation, this method cannot simultaneously realize vital sign monitoring of all targets in the room. Summary of the invention
[0006] In order to overcome the problems of insufficient radar detection range and low positioning accuracy of dynamic targets within the omnidirectional range in the above-mentioned radar life detection technology, the present invention provides a vital sign detection method based on MIMO rotating radar.
[0007] In order to solve the above technical problems, the technical solution of the present invention is as follows:
[0008] The present invention proposes a vital sign detection method based on MIMO rotating radar, comprising:
[0009] S1: Detect all targets in the omnidirectional range based on the MIMO rotating radar to obtain the original radar data matrix;
[0010] S2: Perform multi-dimensional processing on the original radar data matrix to obtain a distance data matrix, a Doppler dimension data matrix and an angle dimension data matrix;
[0011] S3: Obtaining the dynamic target omnidirectional matrix according to the distance data matrix, the Doppler dimension data matrix and the angle dimension data matrix;
[0012] S4: Acquire the position information of all dynamic targets according to the dynamic target omnidirectional matrix, wherein the position information includes distance and angle;
[0013] S5: Obtain a three-dimensional full-target data matrix according to the distance data matrix; obtain an all-target omnidirectional matrix according to the three-dimensional full-target data matrix; obtain an angle of a static reference object according to the all-target omnidirectional matrix;
[0014] S6: Correcting the angles of all dynamic targets using the angles of the static reference objects to obtain corrected angles of all dynamic targets;
[0015] S7: randomly selecting a detection target from the dynamic targets, and performing vital sign detection on the detection target based on the distances and correction angles of all the dynamic targets in the same distance and non-same distance situations.
[0016] Preferably, the S1 includes:
[0017] Set the radar frame signal transmission frequency to N SFRA , the motor speed is w per minute RPM Turn, calculate the number of virtual radars N according to the radar frame signal transmission frequency and motor speed SEG , the specific calculation method is:
[0018]
[0019] in, represents a positive integer;
[0020] Get the original radar data matrix M PRI :
[0021]
[0022] Among them, N ANT Indicates the total number of antenna channels, For N AFRA Nth frame ANT Each channel receives all complex sampling point data of the original signal.
[0023] Preferably, the S2 includes:
[0024] S2.1: Setting k seg =1;
[0025] S2.2: The original radar data matrix M PRI The frame dimension is divided into N SEG Virtual radar data matrix kth seg Virtual radar data matrix Specifically include:
[0026]
[0027] Among them, N DFRA Indicates the total number of frames in the virtual radar, N AFRA is the total number of frames actually transmitted by the radar;
[0028] S2.3: Yes Perform distance-dimensional Fourier transform to obtain distance-dimensional data matrix Specifically include:
[0029]
[0030] Among them, FFT R (·) represents the fast time-dimensional Fourier transform operation on the matrix, where:
[0031]
[0032] S2.4: Performing a Doppler dimension Fourier transform on the distance dimension data matrix to obtain a Doppler dimension data matrix, specifically including:
[0033] Set the total number of Doppler dimension frames tfv and the Doppler dimension start frame sv, where sv≥1 and sv≤N DFRA -tfv, calculates the Doppler dimension data matrix based on the distance dimension data matrix, including:
[0034]
[0035] in, Indicates the kth seg The Doppler dimension data matrix corresponding to the virtual radar data matrix, FFT V (·) represents the Doppler Fourier transform operation, express The data matrix from row sv to row sv+tfv-1 in , where:
[0036]
[0037] in, N ANT The transposed matrix of the column data matrix in, express The Nth frame of tfv ANT Channel data, express In the svth row, the Nth ANT Column data;
[0038] S2.5: Performing an angle-dimensional Fourier transform operation on the Doppler-dimensional data matrix to obtain an angle-dimensional data matrix, specifically including:
[0039]
[0040] Among them, FFT A (·) represents the angle-dimensional Fourier transform operation, express Data in row sa:
[0041]
[0042] in, express The Nth ANT Channel data, N RB Indicates the maximum distance index, N A Indicates the number of channel dimensions after zero padding, express N A Angle index and Nth RB Data of distance index;
[0043] S2.6: Determine whether k seg ≥N SEG If yes, execute step S2.7; otherwise, set k s ' eg =k seg +1, and k s ' eg Re-update k seg , return to step S2.2;
[0044] S2.7: Get N SEG The distance data matrix, Doppler dimension data matrix and angle dimension data matrix corresponding to the virtual radar data matrix specifically include:
[0045]
[0046] Among them, M 1D 、M 2D and M 3D,DO Respectively represent N SEG The distance data matrix, Doppler dimension data matrix and angle dimension data matrix corresponding to the virtual radar data matrix.
[0047] Preferably, S3 includes:
[0048] The dynamic target omnidirectional matrix is denoted as M DO,C :
[0049]
[0050] in, Indicates M DO,C k seg The dynamic target omnidirectional matrix corresponding to the virtual radar data matrix includes:
[0051]
[0052] in, Indicates acquisition From N SA1 Go to N SA2 The data matrix of the row, the dynamic target omnidirectional matrix is denoted by M DO,C Further expressed as:
[0053]
[0054] in,
[0055] Preferably, the S4 includes:
[0056] S4.1: Obtaining a dynamic target detection threshold according to the dynamic target omnidirectional matrix, specifically including:
[0057]
[0058] Among them, TH AIM represents the dynamic target detection threshold, N θB Represents the maximum angle index in the dynamic target omnidirectional matrix, N RB Represents the maximum distance index in the dynamic target omnidirectional matrix, N θB =(N SA2 -N SA1 +1)×N SEG , M DO,C (j RB ,j θB ) indicates M DO,C J θB Row j RB Column data;
[0059] S4.2: Set N AIM =1;
[0060] S4.3: Get the Nth AIM The position information of a dynamic target is calculated as follows:
[0061]
[0062] in, Indicates the Nth AIM The position information of dynamic targets, argmin seg,R,θ (·) represents the position of the target in the search matrix, and the position information includes distance, angle and corresponding virtual radar index, and Respectively represent the Nth AIM The distance, angle and virtual radar index of a dynamic target, seg, R and θ represent the virtual radar index variable, distance variable and angle variable respectively;
[0063] S4.4: Calculate the Nth AIM A dynamic target in M DO,C The normalized magnitude of And determine whether If yes, execute step S4.6, otherwise, execute step S4.5;
[0064] S4.5: Set N A ′ IM =N AIM +1, And N A ′ IM 、M ′ DO,C Re-update N AIM 、M DO,C , return to step S4.3, where Indicates the Nth AIM A dynamic target in M DO,C submatrix of ;
[0065] S4.6: Get the location information of all dynamic targets
[0066] Preferably, the specific process of S4.3 includes:
[0067] S4.3.1: Set j R =1,j θ =1, NUM MAX =0;
[0068] S4.3.2: Get the jth omnidirectional matrix of the dynamic target R distance index, the jth θ The value of angle index NUM T =M DO,C (j R ,j θ ), where M DO,C (j R ,j θ ) indicates M DO,C J θB The jth row RB Column data;
[0069] S4.3.3: Determine whether NUM T ≥NUM MAX , if yes, go to step S4.3.4, otherwise go to step S4.3.5;
[0070] S4.3.4: Update Nth AIM The distance index and angle index of dynamic targets: and update NUM MAX =NUM T ;
[0071] S4.3.5: Determine whether j θ ≥N θB If yes, then execute step S4.3.6, otherwise execute j ′ θ =j θ +1, and j ′ θ Re-update θ , return to step S4.3.2;
[0072] S4.3.6: Set j θ =1,j ′ R =j R +1, j ′ R Re-update R ;
[0073] S4.3.7: Determine whether j R ≥N RB , if yes, execute step S4.3.8, otherwise return to step S4.3.2;
[0074] S4.3.8: Get the Nth AIM Virtual radar index corresponding to dynamic targets Specifically include:
[0075]
[0076] Preferably, the S5 includes:
[0077] S5.1: Setting j seg =1;
[0078] S5.2: Get the jth distance matrix seg A three-dimensional full target data matrix with virtual radar index in:
[0079]
[0080] Among them, si represents the number of selected angle dimension frames, si∈[1,N DFRA ], express N A Angle index Nth RB Data of distance index;
[0081] S5.3: Based on Calculate the jth omnidirectional matrix of all targets seg The data matrix of the parts
[0082] S5.4: Determine whether j seg ≥N SEG If yes, execute step S5.6, otherwise execute step S5.5;
[0083] S5.5: Setting j s ′ eg =j seg +1, j s ′ eg Re-update seg , return to step S5.2;
[0084] S5.6: Get all-target omnidirectional matrix
[0085] S5.7: Obtain the angle θ of the static reference object based on the omnidirectional matrix of the omni-target REF , including:
[0086]
[0087] Among them, argmax θ (·) represents the angle information of the target in the search matrix, for:
[0088]
[0089] in, Indicates the Nth SEG A three-dimensional full target data matrix with virtual radar index N SA2 Angle index R REF distance index data, R REF Indicates the distance index corresponding to the static reference object.
[0090] Preferably, the S6 includes:
[0091]
[0092] in, represents the correction angle of all dynamic targets, Represents the angle information of all dynamic targets,
[0093] Preferably, the S7 includes:
[0094] S7.1: Determine whether If yes, then execute step S7.2, otherwise execute step S7.4, where aim represents the detection target index, represents the virtual radar index of the aim-th detected target, and Oaim represents the index of other targets except aim;
[0095] S7.2: Determine whether Indicates the distance index of the aim-th detected target. If yes, it is the same distance situation, execute step S7.3, otherwise, it is not the same distance situation, execute step S7.4;
[0096] S7.3: Obtain the vital signs of the aim-th detected target at the same distance based on the virtual radar data matrix Specifically include:
[0097]
[0098] in, Indicates A virtual radar data matrix, Indicates that the signal is processed for the vital signs of the target at the same distance. Indicates the angle index of the aim-th detected target at the same distance;
[0099] S7.3.1: Obtain the raw data matrix after beamforming, including:
[0100]
[0101] in, represents the beamforming matrix, specifically:
[0102]
[0103] Among them, d rx represents the distance between the receiving antennas, and λ represents the radar wavelength;
[0104] S7.3.2: Obtain a distance matrix after beamforming under the same distance condition based on the original data matrix after beamforming, and obtain a distance matrix after static filtering under the same distance condition based on the distance matrix after beamforming under the same distance condition, specifically including:
[0105] M θ1D =FFT R (M θPRI )
[0106]
[0107] Among them, M θ1D represents the distance matrix after beamforming under the same distance condition, represents the distance matrix after static filtering under the same distance condition, Indicates M θ1D No. i DFRA The data matrix of the frame;
[0108] S7.3.3: Obtain a three-dimensional angle data matrix after beamforming under the same distance condition based on the distance matrix after static filtering under the same distance condition, specifically including:
[0109]
[0110] Among them, M θ3D Represents the three-dimensional angle data matrix after beamforming at the same distance
[0111] S7.3.4: Extract the three-dimensional angle data matrix after beamforming at the same distance to obtain the phase of the aim-th detection target at the same distance, specifically including:
[0112]
[0113] in, represents the phase of the aim-th detected target at the same distance, phase(·) represents the phase of the extraction matrix, Indicates M θ3D Middle The angle index A data matrix of distance indices;
[0114] S7.3.5: Setting i θPU =2;
[0115] S7.3.6: Calculate phase difference
[0116] S7.3.7: Based on the phase difference ΔθP(i θPU ) obtains the phase of the aim-th detection target after beamforming and phase unwrapping under the same distance, specifically including:
[0117]
[0118] in, It represents the phase of the aim-th detection target after beamforming and phase unwrapping at the same distance. express No. i θPU The phase value of each point is No. i θPU The time domain value of the life signal of each point;
[0119] S7.3.8: Determine whether i θPU ≥NDFRA , if yes, execute step S7.3.10, otherwise execute step S7.3.9;
[0120] S7.3.9: Set i′ θPU =i θPU +1, and i′ θPU Re-update θPU , return to step S7.3.6;
[0121] S7.3.10: Obtain the spectrum of vital signs of the aim-th detection target at the same distance based on the phase of the aim-th detection target after beamforming and phase unwrapping, specifically including:
[0122]
[0123] in, Indicates the spectrum of vital signs of the aim-th detection target at the same distance, FFT VS (·) represents the vital sign-dimensional Fourier transform of the matrix;
[0124] S7.4: Obtain the vital signs of the aim-th detection target under non-same-range conditions according to the virtual radar data matrix, specifically including:
[0125]
[0126] in, Indicates the vital signs of the aim-th detection target in non-same distance conditions, It means that the vital signs of the target at different distances are processed on the signal.
[0127] Preferably, the specific process of S7.4 includes:
[0128] S7.4.1: Obtaining a one-dimensional data matrix after static filtering
[0129]
[0130] in, Indicates M 1D Middle The i-th virtual radar index DFRA The data matrix of the frame;
[0131] S7.4.2: Extract the one-dimensional data matrix after static filtering to obtain the phase of the aim-th detected target in the case of non-same distance, specifically including:
[0132]
[0133] in, Indicates the phase of the aim-th detected target in non-same distance conditions, Indicated in Middle A data matrix of distance indices;
[0134] S7.4.3: Setting i PU =2;
[0135] S7.4.4: Obtaining Phase Difference
[0136] S7.4.5: Based on the phase difference ΔP(i PU ) obtains the phase of the aim-th detection target after phase unwrapping in the case of non-same distance, specifically including:
[0137]
[0138] in, It represents the phase of the aim-th detection target after phase unwrapping in the case of non-same distance. express No. i PU The phase value of each point is No. i PU The time domain value of the life signal of each point;
[0139] S7.4.6: Determine whether i PU ≥N DFRA , if yes, execute step S7.4.8, otherwise execute step S7.4.7;
[0140] S7.4.7: Setting i ′ PU =i PU +1, and i ′ PU Re-update PU , return to step S7.4.4;
[0141] S7.4.8: Merger The time domain value of the life signal of each channel in is obtained to obtain the combined time domain value of the life signal of the aim-th detection target;
[0142]
[0143] in, represents the combined life signal time domain value of the aim-th detection target, express No. i ant A data matrix with channel indices;
[0144] S7.4.9: Setting i CA =1,
[0145] S7.4.10: Calculate the correlation of the life signal channel of the aim-th detection target in non-same distance conditions:
[0146]
[0147] in, Indicates the life signal of the aim-th detection target i CA The correlation of the channels, express No. i CA A data matrix with channel indices; Expressed as and The covariance of , σ(·) is expressed as the variance of the matrix;
[0148] S7.4.11: Obtain the correlation of the life signal channel of the aim-th detection target under non-same distance conditions by accumulating the spectrum signal and will Re-update Specifically include:
[0149]
[0150] S7.4.12: Determine whether i CA >N CA , if yes, execute step S7.4.14, otherwise execute step S7.4.13;
[0151] S7.4.13: Setting i ′ CA =i CA +1, and i ′ CA Re-update CA , return to step S7.4.10;
[0152] S7.4.14: Obtain the vital signs of the aim-th detection target in non-uniform distance situations
[0153] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:
[0154] The present invention proposes a vital sign detection method based on MIMO rotating radar. First, according to the characteristics of the rotating motor and the FMCW radar, the original data of a single radar is divided into multiple virtual radar data, and the virtual radar data is processed in multiple dimensions to obtain the position information of the dynamic target and the position information of the static reference object. Then, the angle of the dynamic target is calibrated by the position information of the static reference object at a fixed position, and different processing methods are used for vital sign detection for the same distance and different distances. The present invention can realize all-round detection around the radar, reduce the problem of the small directional detection range of the radar, avoid the interference of static objects and multipaths on the positioning of dynamic targets, and can calibrate the acquired target position information, so as to accurately obtain the absolute position of the dynamic target within the omnidirectional range, enhance the strength of the vital signal, and improve the accuracy of vital signal detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0155] Figure 1 This is a schematic diagram of the flow of the vital sign detection method described in Example 1;
[0156] Figure 2 This is a flow chart of steps S1 and S2 of the vital sign detection method described in Example 2;
[0157] Figure 3 This is a flow chart of step S4 of the vital sign detection method described in Example 3;
[0158] Figure 4 This is a flow chart of steps S5 and S6 of the vital sign detection method described in Example 3. DETAILED DESCRIPTION
[0159] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;
[0160] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;
[0161] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.
[0162] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0163] Example 1
[0164] This embodiment provides a method for detecting vital signs based on MIMO rotating radar. Figure 1 As shown, including:
[0165] S1: Detect all targets in the omnidirectional range based on the MIMO rotating radar to obtain the original radar data matrix;
[0166] S2: Perform multi-dimensional processing on the original radar data matrix to obtain a distance data matrix, a Doppler dimension data matrix and an angle dimension data matrix;
[0167] S3: Obtaining the dynamic target omnidirectional matrix according to the distance data matrix, the Doppler dimension data matrix and the angle dimension data matrix;
[0168] S4: Acquire the position information of all dynamic targets according to the dynamic target omnidirectional matrix, wherein the position information includes distance and angle;
[0169] S5: Obtain a three-dimensional full-target data matrix according to the distance data matrix; obtain an all-target omnidirectional matrix according to the three-dimensional full-target data matrix; obtain an angle of a static reference object according to the all-target omnidirectional matrix;
[0170] S6: Correcting the angles of all dynamic targets using the angles of the static reference objects to obtain corrected angles of all dynamic targets;
[0171] S7: randomly selecting a detection target from the dynamic targets, and performing vital sign detection on the detection target based on the distances and correction angles of all the dynamic targets in the same distance and non-same distance situations.
[0172] In the specific implementation process, the original radar data matrix is first obtained, and the original radar data matrix is processed in multiple dimensions to obtain the distance data matrix, the Doppler dimension data matrix and the angle dimension data matrix. Then the dynamic target omnidirectional matrix is obtained, and the position information of all dynamic targets is obtained according to the dynamic target omnidirectional matrix, and the position information includes distance and angle; then the angle of the static reference object is obtained according to the distance data matrix, and the angle of all dynamic targets is corrected by using the angle of the static reference object; finally, the detection target is arbitrarily selected from the dynamic targets, and the vital signs of the detection target are detected based on the distance and correction angle of all dynamic targets at the same distance and non-same distance.
[0173] Example 2
[0174] This embodiment further defines steps S1, S2 and S3 based on embodiment 1. The flowcharts of steps S1 and S2 are as follows: Figure 2 As shown:
[0175] The specific process of step S1 is:
[0176] Set the radar frame signal transmission frequency to N SFRA , the motor speed is w per minute RPM Turn, calculate the number of virtual radars N according to the radar frame signal transmission frequency and motor speed SEG , the specific calculation method is:
[0177]
[0178] in, represents a positive integer;
[0179] In this embodiment, N is set SFRA =50,w RPM =600, we get N SEG =5;
[0180] Get the original radar data matrix M PRI :
[0181]
[0182] Among them, N ANT Indicates the total number of antenna channels, For N AFRA Nth frame ANT Each channel receives all complex sampling point data of the original signal.
[0183] The specific process of step S2 is:
[0184] S2.1: Setting k seg =1;
[0185] S2.2: The original radar data matrix M PRI The frame dimension is divided into N SEG Virtual radar data matrix kth seg Virtual radar data matrix Specifically include:
[0186]
[0187] Among them, N DFRA Indicates the total number of frames in the virtual radar, N AFRA is the total number of frames actually transmitted by the radar;
[0188] In this embodiment, N is set AFRA =3000, N ANT =8, we get N DFRA =600, at this time the kth seg Virtual radar data matrix for:
[0189]
[0190] S2.3: Yes Perform distance-dimensional Fourier transform to obtain distance-dimensional data matrix Specifically include:
[0191]
[0192] Among them, FFT R (·) represents the fast time-dimensional Fourier transform operation on the matrix, where:
[0193]
[0194] In this embodiment, N ANT =8,N DFRA =600, we get:
[0195]
[0196] S2.4: Performing a Doppler dimension Fourier transform on the distance dimension data matrix to obtain a Doppler dimension data matrix, specifically including:
[0197] Set the total number of Doppler dimension frames tfv and the Doppler dimension start frame sv, where sv≥1 and sv≤N DFRA -tfv, calculates the Doppler dimension data matrix based on the distance dimension data matrix, including:
[0198]
[0199] in, Indicates the kth seg The Doppler dimension data matrix corresponding to the virtual radar data matrix, FFT V (·) represents the Doppler Fourier transform operation, express The data matrix from row sv to row sv+tfv-1 in , where:
[0200]
[0201]
[0202] in, N ANT The transposed matrix of the column data matrix in, express The Nth frame of tfv ANT Channel data, express In the svth row, the Nth ANT Column data;
[0203] In this embodiment, N ANT =8, sv=200, tfv=16, we get:
[0204]
[0205] S2.5: Performing an angle-dimensional Fourier transform operation on the Doppler-dimensional data matrix to obtain an angle-dimensional data matrix, specifically including:
[0206]
[0207] Among them, FFT A (·) represents the angle-dimensional Fourier transform operation, express Data in row sa:
[0208]
[0209] in, express The Nth ANT Channel data, N RB Indicates the maximum distance index, N A Indicates the number of channel dimensions after zero padding, express N A Angle index and Nth RB Data of distance index;
[0210] In this embodiment, N ANT =8, sa=5, tfv=16, N RB =128, N A =180, we get:
[0211]
[0212] S2.6: Determine whether k seg ≥N SEG If yes, execute step S2.7; otherwise, set k s ′ eg =k seg +1, and k s ′ eg Re-update k seg , return to step S2.2;
[0213] S2.7: Get N SEG The distance data matrix, Doppler dimension data matrix and angle dimension data matrix corresponding to the virtual radar data matrix specifically include:
[0214]
[0215] Among them, M 1D 、M 2D and M 3D,DO Respectively represent N SEG A distance data matrix, a Doppler dimension data matrix and an angle dimension data matrix corresponding to a virtual radar data matrix;
[0216] In this embodiment, N SEG =5, we get:
[0217]
[0218] The specific process of step S3 is:
[0219] The dynamic target omnidirectional matrix is denoted as M DO,C :
[0220]
[0221] in, Indicates M DO,C k seg The dynamic target omnidirectional matrix corresponding to the virtual radar data matrix includes:
[0222]
[0223] in, Indicates acquisition From N SA1 Go to N SA2 The data matrix of the row, the dynamic target omnidirectional matrix is denoted by M DO,C Further expressed as:
[0224]
[0225] in,
[0226] In this embodiment, N SEG =5,N RB =128, N A =180, we get N SA1 =54,N SA2 =125,
[0227]
[0228] Example 3
[0229] This embodiment further defines steps S4, S5 and S6 based on embodiment 2, such as Figure 3 As shown, it is a flow chart of step S4, Figure 4 is a flow chart of steps S5 and S6.
[0230] The specific process of step S4 is as follows:
[0231] S4.1: Obtaining a dynamic target detection threshold according to the dynamic target omnidirectional matrix, specifically including:
[0232]
[0233] Among them, TH AIM represents the dynamic target detection threshold, N θB Represents the maximum angle index in the dynamic target omnidirectional matrix, N RB Represents the maximum distance index in the dynamic target omnidirectional matrix, N θB =(N SA2 -N SA1 +1)×N SEG , M DO,C (j RB ,j θB ) indicates M DO,C J θB Row j RB Column data;
[0234] In this embodiment, N SEG =5,N SA1 =54,N SA2 =125, we get N θB =360;
[0235] S4.2: Set N AIM =1;
[0236] S4.3: Get the Nth AIM The position information of a dynamic target is calculated as follows:
[0237]
[0238] in, Indicates the Nth AIM The position information of dynamic targets, argmin seg,R,θ (·) represents the position of the target in the search matrix, and the position information includes distance, angle and corresponding virtual radar index, and Respectively represent the Nth AIM The distance, angle and virtual radar index of a dynamic target, seg, R and θ represent the virtual radar index variable, distance variable and angle variable respectively;
[0239] In this embodiment,
[0240] S4.3.1: Set jR =1,j θ =1, NUM MAX =0;
[0241] S4.3.2: Get the jth omnidirectional matrix of the dynamic target R distance index, the jth θ The value of angle index NUM T =M DO,C (j R ,j θ ), where M DO,C (j R ,j θ ) indicates M DO,C J θB The jth row RB Column data;
[0242] S4.3.3: Determine whether NUM T ≥NUM MAX , if yes, go to step S4.3.4, otherwise go to step S4.3.5;
[0243] S4.3.4: Update Nth AIM The distance index and angle index of dynamic targets: and update NUM MAX =NUM T ;
[0244] S4.3.5: Determine whether j θ ≥N θB If yes, execute step S4.3.6, otherwise execute j′ θ =j θ +1, and j′ θ Re-update θ , return to step S4.3.2;
[0245] S4.3.6: Set j θ =1,j′ R =j R +1, j′ R Re-update R ;
[0246] S4.3.7: Determine whether j R ≥N RB , if yes, execute step S4.3.8, otherwise return to step S4.3.2;
[0247] S4.3.8: Get the Nth AIM Virtual radar index corresponding to dynamic targets Specifically include:
[0248]
[0249] In this embodiment, N SEG =5,N θB =360, get
[0250] S4.4: Calculate the Nth AIM A dynamic target in M DO,C The normalized magnitude of And determine whether If yes, execute step S4.6, otherwise, execute step S4.5;
[0251] In this embodiment, when N AIM =1, TH AIM =0.0608, then execute step S4.5;
[0252] S4.5: Set N A ′ IM =N AIM +1, And N A ′ IM 、M ′ DO,C Re-update N AIM 、M DO,C , return to step S4.3, where Indicates the Nth AIM A dynamic target in M DO,C submatrix of ;
[0253] In this embodiment, when N AIM =1, When
[0254]
[0255] S4.6: Get the location information of all dynamic targets
[0256] The specific process of step S5 is as follows:
[0257] S5.1: Setting j seg =1;
[0258] S5.2: Get the jth distance matrix seg A three-dimensional full target data matrix with virtual radar index in:
[0259]
[0260] Among them, si represents the number of selected angle dimension frames, si∈[1,N DFRA ], express N A Angle index Nth RB Data of distance index;
[0261] In this embodiment, si=200, N DFRA =600, N RB =128, N A =180, N ANT =8, we get
[0262]
[0263] S5.3: Based on Calculate the jth omnidirectional matrix of all targets seg The data matrix of the parts
[0264] S5.4: Determine whether j seg ≥N SEG If yes, execute step S5.6, otherwise execute step S5.5;
[0265] S5.5: Setting j s ′ eg =j seg +1, j s ′ eg Re-update seg , return to step S5.2;
[0266] S5.6: Get all-target omnidirectional matrix
[0267] S5.7: Obtain the angle θ of the static reference object based on the omnidirectional matrix of the omni-target REF , including:
[0268]
[0269] Among them, argmax θ (·) represents the angle information of the target in the search matrix, for:
[0270]
[0271] in, Indicates the Nth SEGA three-dimensional full target data matrix with virtual radar index N SA2 Angle index R REF distance index data, R REF Indicates the distance index corresponding to the static reference object.
[0272] In this embodiment, N SEG =5,N RB =128, N A =180, N SA1 =54,N SA2 =125, R REF =100, we get
[0273]
[0274]
[0275] The specific process of step S6 is as follows:
[0276]
[0277] in, represents the correction angle of all dynamic targets, Represents the angle information of all dynamic targets,
[0278] The same or similar reference numerals correspond to the same or similar components;
[0279] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;
[0280] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.
Claims
1. A vital sign detection method based on MIMO rotating radar, characterized in that: include: S1: Detect all targets in the omnidirectional range based on the MIMO rotating radar to obtain the original radar data matrix; S2: Perform multi-dimensional processing on the original radar data matrix to obtain a distance data matrix, a Doppler dimension data matrix and an angle dimension data matrix; S3: Obtaining the dynamic target omnidirectional matrix according to the distance data matrix, the Doppler dimension data matrix and the angle dimension data matrix; S4: Acquire the position information of all dynamic targets according to the dynamic target omnidirectional matrix, wherein the position information includes distance and angle; S5: Obtain a three-dimensional full-target data matrix according to the distance data matrix; obtain an all-target omnidirectional matrix according to the three-dimensional full-target data matrix; obtain an angle of a static reference object according to the all-target omnidirectional matrix; S6: Correcting the angles of all dynamic targets using the angles of the static reference objects to obtain corrected angles of all dynamic targets; S7: randomly selecting a detection target from the dynamic targets, and performing vital sign detection on the detection target based on the distances and correction angles of all the dynamic targets in the same distance and non-same distance situations.
2. The method for detecting vital signs based on MIMO rotating radar according to claim 1, characterized in that: The S1 includes: Set the radar frame signal transmission frequency to N SFRA , the motor speed is W per minute RPM Turn, calculate the number of virtual radars N according to the radar frame signal transmission frequency and motor speed SEG , the specific calculation method is: in, represents a positive integer; Get the original radar data matrix M PRI : Among them, N ANT Indicates the total number of antenna channels, For N AFRA Nth frame ANT Each channel receives all complex sampling point data of the original signal.
3. The vital sign detection method based on MIMO rotating radar according to claim 2 is characterized in that: The S2 includes: S2.1: Setting k seg =1; S2.2: The original radar data matrix M PRI The frame dimension is divided into N SEG Virtual radar data matrix kth seg Virtual radar data matrix k seg =1,2,…,N SEG , Specifically include: Among them, N DFRA Represents the total number of frames in the virtual radar, N DFRA =N AFRA / N SEG st N AFRA is the total number of frames actually transmitted by the radar; S2.3: Yes Perform distance-dimensional Fourier transform to obtain distance-dimensional data matrix Specifically include: Among them, FFT R (·) represents the fast time-dimensional Fourier transform operation on the matrix, where: S2.4: Performing a Doppler dimension Fourier transform on the distance dimension data matrix to obtain a Doppler dimension data matrix, specifically including: Set the total number of Doppler dimension frames tfv and the Doppler dimension start frame sv, where sv≥1 and sv≤N DFRA -tfv, calculates the Doppler dimension data matrix based on the distance dimension data matrix, including: in, Indicates the Doppler dimension data matrix corresponding to the ksegth virtual radar data matrix, FFT V (·) represents the Doppler Fourier transform operation, express The data matrix from row sv to row sv+tfv-1 in , where: in, N ANT The transposed matrix of the column data matrix in, express The Nth frame of tfv ANT Channel data, express In the svth row, the Nth ANT Column data; S2.5: Performing an angle-dimensional Fourier transform operation on the Doppler-dimensional data matrix to obtain an angle-dimensional data matrix, specifically including: Among them, FFT A (·) represents the angle-dimensional Fourier transform operation, sa∈[1,tfv], st express Data in row sa: in, express The Nth ANT Channel data, N RB Indicates the maximum distance index, N A Indicates the number of channel dimensions after zero padding, express N A Angle index and Nth RB Data of distance index; S2.6: Determine whether k seg ≥N SBG If yes, execute step S2.7, otherwise, set k′ seg =k seg +1, and k′ seg Re-update k seg , return to step S2.2; S2.7: Get N SEG The distance data matrix, Doppler dimension data matrix and angle dimension data matrix corresponding to the virtual radar data matrix specifically include: Among them, M 1D 、M 2D and M 3D,DO Respectively represent N SEG The distance data matrix, Doppler dimension data matrix and angle dimension data matrix corresponding to the virtual radar data matrix.
4. The vital sign detection method based on MIMO rotating radar according to claim 3 is characterized in that: The S3 includes: The dynamic target omnidirectional matrix is denoted as M DO,C : in, Indicates M DO,C k seg The dynamic target omnidirectional matrix corresponding to the virtual radar data matrix includes: in, Indicates acquisition From N SA1 Go to N SA2 The data matrix of the row, the dynamic target omnidirectional matrix is denoted by M DO,C Further expressed as: in, sN SEG ≥2.
5. The method for detecting vital signs based on MIMO rotating radar according to claim 4, characterized in that: The S4 includes: S4.1: Obtaining a dynamic target detection threshold according to the dynamic target omnidirectional matrix, specifically including: Among them, TH AIM represents the dynamic target detection threshold, N θB Represents the maximum angle index in the dynamic target omnidirectional matrix, N RB Represents the maximum distance index in the dynamic target omnidirectional matrix, N θB =(N SA2 -N SA1 +1)×N SEG , M DO,C (j RB , j θB ) indicates M DO,C J θB Row j RB Column data; S4.2: Set N AIM =1; S4.3: Get the Nth AIM The position information of a dynamic target is calculated as follows: in, Indicates the Nth AIM The position information of a dynamic target, argmtn seg,R,θ (·) represents the position of the target in the search matrix, and the position information includes distance, angle and corresponding virtual radar index, and Respectively represent the Nth AIM The distance, angle and virtual radar index of a dynamic target, seg, R and θ represent the virtual radar index variable, distance variable and angle variable respectively; S4.4: Calculate the Nth AIM A dynamic target in M DO,C The normalized magnitude of And determine whether If yes, execute step S4.6, otherwise, execute step S4.5; S4.5: Setting N′ AIM =N AIM +1, And N′ AIM , M′ DO,C Re-update N AIM 、M DO,C , return to step S4.3, where Indicates the Nth AIM A dynamic target in M DO,C submatrix of ; S4.6: Get the location information of all dynamic targets 6. The vital sign detection method based on MIMO rotating radar according to claim 5 is characterized in that: The specific process of S4.3 includes: S4.3.1: Set j R =1,j θ =1, NUM MAX =0; S4.3.2: Get the jth omnidirectional matrix of the dynamic target R distance index, the jth θ The value of angle index NUM T =M DO,C (j R , j θ ), where M DO,C (j R ,j θ ) indicates M DO,C J θB The jth row RB Column data; S4.3.3: Determine whether NUM T ≥NUM MAX , if yes, go to step S4.3.4, otherwise go to step S4.3.5; S4.3.4: Update Nth AIM The distance index and angle index of dynamic targets: and update NUM MAX =NUM T ; S4.3.5: Determine whether j θ ≥N θB If yes, execute step S4.3.6, otherwise execute j′ θ =j θ +1, and j′ θ Re-update θ , return to step S4.3.2; S4.3.6: Set j θ =1,j′ R =j R +1, j′ R Re-update R ; S4.3.7: Determine whether j R ≥N RB , if yes, execute step S4.3.8, otherwise return to step S4.3.2; S4.3.8: Get the Nth AIM Virtual radar index corresponding to dynamic targets Specifically include:
7. The vital sign detection method based on MIMO rotating radar according to claim 5, characterized in that: The S5 includes: S5.1: Setting j seg =1; S5.2: Get the jth distance matrix seg Virtual radar indexed 3D full target data matrix in: Among them, si represents the number of selected angle dimension frames, si∈[1,N DFRA ], express N A Angle index Nth RB Data of distance index; S5.3: Based on Calculate the jth omnidirectional matrix of all targets seg The data matrix of the parts S5.4: Determine whether j seg ≥N SEG If yes, execute step S5.6, otherwise execute step S5.5; S5.5: Setting j′ seg =j seg +1, j′ seg Re-update seg , return to step S5.2; S5.6: Get all-target omnidirectional matrix S5.7: Obtain the angle θ of the static reference object based on the omnidirectional matrix of the omni-target REF , including: Among them, argmax θ (·) represents the angle information of the target in the search matrix, for: in, Indicates the Nth SEG Virtual radar indexed 3D full target data matrix N SA2 Angle index R REF distance index data, R REF Indicates the distance index corresponding to the static reference object.
8. The vital sign detection method based on MIMO rotating radar according to claim 7 is characterized in that: The S6 includes: in, Indicates the correction angle of all dynamic targets, Represents the angle information of all dynamic targets, 9. The method for detecting vital signs based on MIMO rotating radar according to claim 8, characterized in that: The S7 includes: S7.1: Determine whether If yes, then execute step S7.2, otherwise execute step S7.4, where aim represents the detection target index, represents the virtual radar index of the aim-th detected target, and Oaim represents the index of other targets except aim; S7.2: Determine whether Indicates the distance index of the aim-th detected target. If yes, it is the same distance situation, execute step S7.3, otherwise, it is not the same distance situation, execute step S7.4; S7.3: Obtain the vital signs of the aim-th detected target at the same distance based on the virtual radar data matrix Specifically include: in, Indicates A virtual radar data matrix, Indicates that the signal is processed for the vital signs of the target at the same distance. Indicates the angle index of the aim-th detected target at the same distance; S7.3.1: Obtain the raw data matrix after beamforming, including: in, represents the beamforming matrix, specifically: Among them, d rx represents the distance between the receiving antennas, and λ represents the radar wavelength; S7.3.2: Obtain a distance matrix after beamforming under the same distance condition based on the original data matrix after beamforming, and obtain a distance matrix after static filtering under the same distance condition based on the distance matrix after beamforming under the same distance condition, specifically including: M θ1D =FFT R (M θPRI ) Among them, M θ1D represents the distance matrix after beamforming under the same distance condition, represents the distance matrix after static filtering under the same distance condition, Indicates M θ1D No. i DFRA The data matrix of the frame; S7.3.3: Obtain a three-dimensional angle data matrix after beamforming under the same distance condition based on the distance matrix after static filtering under the same distance condition, specifically including: Among them, M θ3D Represents the three-dimensional angle data matrix after beamforming at the same distance S7.3.4: Extract the three-dimensional angle data matrix after beamforming at the same distance to obtain the phase of the aim-th detection target at the same distance, specifically including: in, represents the phase of the aim-th detected target at the same distance, phase(·) represents the phase of the extraction matrix, Indicates M θ3D Middle The angle index A data matrix of distance indices; S7.3.5: Setting i θPU =2; S7.3.6: Calculate phase difference S7.3.7: Based on the phase difference ΔθP(i θPU ) obtains the phase of the aim-th detection target after beamforming and phase unwrapping under the same distance, specifically including: in, It represents the phase of the aim-th detection target after beamforming and phase unwrapping at the same distance. express No. i θPU The phase value of each point is No. i θPU The time domain value of the life signal at each point; S7.3.8: Determine whether i θPU ≥N DFRA If yes, go to step S7.3.10, otherwise go to step S7.3.9: S7.3.9: Set i′ θPU =i θPU +1, and i′ θPU Re-update θPU , return to step S7.3.6; S7.3.10: Obtain the spectrum of vital signs of the aim-th detection target at the same distance based on the phase of the aim-th detection target after beamforming and phase unwrapping, specifically including: in, Indicates the spectrum of vital signs of the aim-th detection target at the same distance, FFT VS (·) represents the vital sign-dimensional Fourier transform of the matrix; S7.4: Obtain the vital signs of the aim-th detection target under non-same-range conditions according to the virtual radar data matrix, specifically including: in, Indicates the vital signs of the aim-th detection target in non-same distance conditions, It means that the vital signs of the target at different distances are processed on the signal.
10. The vital sign detection method based on MIMO rotating radar according to claim 9, characterized in that: The specific process of S7.4 includes: S7.4.1: Obtaining a one-dimensional data matrix after static filtering in, Indicates M 1D Middle The i-th virtual radar index DFRA The data matrix of the frame; S7.4.2: Extract the one-dimensional data matrix after static filtering to obtain the phase of the aim-th detected target in the case of non-same distance, specifically including: in, Indicates the phase of the aim-th detected target in non-same distance conditions, Indicated in Middle A data matrix of distance indices; S7.4.3: Setting i PU =2; S7.4.4: Obtaining Phase Difference S7.4.5: Based on the phase difference ΔP(i PU ) obtains the phase of the aim-th detection target after phase unwrapping in the case of non-same distance, specifically including: in, It represents the phase of the aim-th detection target after phase unwrapping in the case of non-same distance. express No. i PU The phase value of each point is No. i PU The time domain value of the life signal of each point; S7.4.6: Determine whether i PU ≥ND FRA , if yes, execute step S7.4.8, otherwise execute step S7.4.7; S7.4.7: Set i′ PU =i PU +1, and i′ PU Re-update PU , return to step S7.4.4; S7.4.8: Merger The time domain value of the life signal of each channel in is obtained to obtain the combined time domain value of the life signal of the aim-th detection target; in, represents the combined life signal time domain value of the aim-th detection target, express No. i ant A data matrix with channel indices; S7.4.9: Setting i CA =1, S7.4.10: Calculate the correlation of the life signal channel of the aim-th detection target in non-same distance conditions: in, Indicates the life signal of the aim-th detection target i CA The correlation of the channels, express No. i CA A data matrix with channel indices; Expressed as and The covariance of , σ(·) is expressed as the variance of the matrix; S7.4.11: Obtain the correlation of the life signal channel of the aim-th detection target under non-same distance conditions by accumulating the spectrum signal and will Re-update Specifically include: S7.4.12: Determine whether i CA >N CA , if yes, execute step S7.4.14, otherwise execute step S7.4.13; S7.4.13: Set i′ CA =i CA +1, and i′ CA Re-update CA , return to step S7.4.10; S7.4.14: Obtain the vital signs of the aim-th detection target in non-uniform distance situations
