A method and system for triage and control of wounded soldiers based on multi-target radar vital sign analysis

Abstract

This invention provides a method and system for casualty classification control based on multi-target radar vital sign analysis, relating to the field of vital sign monitoring technology. The method includes: real-time acquisition of echo signals from millimeter-wave radar; background suppression, range analysis, Doppler analysis, and array angle measurement processing of the echo signals to extract observation features and construct a standardized target observation dataset; identification of effective echo points within the radar observation space for multi-target separation; establishment of target trajectory tracking using the target observation dataset to obtain target state information, extraction of chest cavity micro-motion information, and identification of respiratory and heartbeat characteristics as vital sign sequences; comprehensive analysis of the target's spatial location and vital sign sequences to assess the consistency of vital sign attribution relationships between adjacent targets; rebinding potentially overlapping vital signs; determining the casualty level using a casualty classification assessment model and generating a casualty rescue priority sequence.

Description

Technical Field

[0001] This invention relates to the field of vital sign monitoring technology, and in particular to a method and system for triage and control of wounded personnel based on multi-target radar vital sign analysis. Background Technology

[0002] In disaster relief, public safety monitoring, and life detection in complex environments, the rapid identification and location of the injured and the assessment of their vital signs are crucial for improving on-site rescue efficiency and optimizing the allocation of rescue resources. Millimeter-wave radar, due to its non-contact detection capabilities, insensitivity to lighting conditions, and ability to penetrate obstructions such as smoke and dust, has been widely used in recent years for human presence detection and vital sign monitoring. Therefore, human detection and vital sign monitoring technology based on millimeter-wave radar signal processing is gradually becoming an important technical means for life detection and status assessment in complex environments.

[0003] Currently, existing technologies mainly involve deploying millimeter-wave radar arrays to collect echo signals and performing time-frequency transformation based on multi-scale sliding windows to generate time-frequency energy distribution maps. Based on the differences in energy concentration and stability of different frequency components in the map, effective signal components corresponding to human respiration and heartbeat are extracted, and then a vital sign signal trajectory model is established to achieve high-precision monitoring of human respiration and heart rate.

[0004] However, in situations such as disaster relief where many people are densely packed, the spatial distance is small, or the breathing frequency is similar, the micro-motion spectral peaks of different people are prone to superimpose within the distance and angle resolution unit, making it difficult for the vital signs to correspond stably with the target trajectory. Summary of the Invention

[0005] To address the technical problem that existing technologies often result in the superposition of micro-motion spectral peaks of different individuals within the distance and angle resolution unit in situations such as disaster relief where multiple people are densely packed, the spatial distance is small, or the breathing frequency is similar, making it difficult to stably correspond the vital signs to the target trajectory, this invention provides a method and system for graded control of casualties based on multi-target radar vital sign analysis.

[0006] The technical solutions provided by the embodiments of the present invention are as follows: The first aspect of this invention provides a method for triage and control of wounded soldiers based on multi-target radar vital sign analysis, comprising: S1: Real-time acquisition of echo signals from millimeter-wave radar; S2: Perform background suppression, range analysis, Doppler analysis, and array angle measurement on the echo signal to extract observation features and construct a standardized target observation dataset; S3: Based on the target observation dataset, identify effective echo points in the radar observation space and perform multi-target separation based on the location of the effective echo points; S4: Establish target trajectory tracking using the target observation dataset to obtain target state information. Extract chest cavity micro-motion information of the target within the corresponding spatial gating range based on the echo signal, and identify respiratory and heartbeat characteristics as vital sign sequences based on chest cavity micro-motion information. S5: Comprehensively analyze the spatial location and vital sign sequence of the target, and conduct a consistency correlation assessment on the attribution relationship of vital signs between adjacent targets; S6: Based on the consistency association assessment results, identify the matching reliability between vital signs and targets, and rebind vital signs that may be mixed. S7: Based on the vital signs of the re-bound target, determine the casualty level through the casualty grading assessment model, and generate a casualty rescue priority sequence according to the casualty level.

[0007] A second aspect of the present invention provides a casualty triage control system based on multi-target radar vital sign analysis, comprising: processor; The memory stores computer-readable instructions, which, when executed by the processor, implement the wounded triage control method based on multi-target radar vital sign analysis as described in the first aspect.

[0008] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the wounded triage control method based on multi-target radar vital sign analysis as described in the first aspect.

[0009] The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following: In this invention, a standardized target observation dataset is constructed and multi-target separation and trajectory tracking are performed. Based on this, the micro-motion information of the target's chest cavity is extracted and the respiratory and heartbeat characteristics are identified. Furthermore, the spatial location of the target and the vital sign sequence are comprehensively analyzed, and the consistency association assessment of the vital sign attribution relationship between adjacent targets is carried out. Based on the assessment results, the vital signs that may be mixed are re-bound. This effectively solves the problem of erroneous association between vital signs and target identity in multi-target mixed scenarios, ensures the stable correspondence between vital signs and target trajectories, significantly improves the accuracy of the criticality classification of the wounded, and provides a reliable basis for priority scheduling of rescue. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating a method for classifying and controlling casualties based on multi-target radar vital sign analysis, provided in an embodiment of the present invention.

[0012] Figure 2 This is a schematic diagram of the confidence distribution of target binding in a neighborhood target feature, provided as an embodiment of the present invention.

[0013] Figure 3 This is a schematic diagram of a wounded soldier classification control system based on multi-target radar vital sign analysis, provided as an embodiment of the present invention. Detailed Implementation

[0014] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0015] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0016] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0017] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0018] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0019] Reference manual attached Figure 1 The diagram shows a flowchart of a method for classifying and controlling casualties based on multi-target radar vital sign analysis provided by an embodiment of the present invention.

[0020] This invention provides a method for casualty triage and control based on multi-target radar vital sign analysis. This method can be implemented by a casualty triage and control device based on multi-target radar vital sign analysis, which can be a terminal or a server. The processing flow of the casualty triage and control method based on multi-target radar vital sign analysis may include the following steps: S1: Real-time acquisition of echo signals from millimeter-wave radar.

[0021] Specifically, a millimeter-wave radar array continuously transmits electromagnetic waves according to a linear frequency modulation signal cycle and receives the reflected echo signals.

[0022] Among them, millimeter-wave radar refers to radar systems operating in the frequency band of 30GHz to 300GHz (corresponding to wavelengths of 1mm to 10mm).

[0023] S2: Perform background suppression, range analysis, Doppler analysis, and array angle measurement on the echo signal to extract observation features and construct a standardized target observation dataset.

[0024] In one possible implementation, S2 specifically includes sub-steps S201 to S208: S201: The echo signal is quadratured through the radar receiver to obtain the complex baseband signal sequence of each receiving channel.

[0025] Specifically, the radar receiver performs orthogonal demodulation on the echo signal to obtain the complex baseband signal sequence for each receiving channel. Preferably, the millimeter-wave radar operates in the 76GHz~81GHz frequency band, employing an FMCW radar system with a center frequency of 77GHz. The linear frequency modulation signal bandwidth is 3GHz, the modulation period is 40μs, and each modulation frame contains 128 chirp signals. The radar array adopts a 4-transmit, 8-receive MIMO array structure, forming an equivalent 32 virtual array element array through time-division multiplexing. The ADC sampling rate is 5MHz to ensure the accuracy of range-dimensional FFT analysis. The frame period is preferably 50ms to meet the time resolution required for detecting subtle movements of human respiration and heartbeat. The above parameters can be adjusted according to the specifications of the actual radar equipment, but all are within the scope of this method.

[0026] S202: Within the sliding time window, the mean of the complex baseband signal in the complex baseband signal sequence is calculated, and the mean value is subtracted from the original complex baseband signal to obtain the complex baseband signal with static background removed.

[0027] Specifically, the average value of the complex baseband signal from each receiving channel is calculated within a sliding time window, and this average value is subtracted from the original complex baseband signal to obtain the complex baseband signal with static background removed. The sliding time window is preferably 0.5s to 2s, used to statistically average the static background echo within the window, thereby eliminating stable reflection components generated by walls, the ground, and fixed equipment. The averaging calculation is performed independently for each receiving channel to avoid the gain differences between different channels affecting the background suppression effect.

[0028] S203: Perform a Fast Fourier Transform on the complex baseband signal sequence with static background removed within each linear frequency modulation signal period to obtain the range-frequency domain complex spectrum corresponding to each receiving channel.

[0029] The Fast Fourier Transform (FFT) is an efficient algorithm implementation of the Discrete Fourier Transform (DFT), which can convert time-domain signals into frequency-domain signals.

[0030] S204: Calculate the amplitude of the range frequency domain complex spectrum to obtain the target echo energy amplitude.

[0031] S205: Based on the radar frequency modulation slope and the speed of light, the range frequency in the range frequency domain complex spectrum is converted to obtain the corresponding target range.

[0032] Specifically, the formula for calculating the target distance is:

[0033] in, Indicates the distance to the target. Represents the speed of light. Represents the distance frequency in the complex spectrum of the distance frequency domain. The radar frequency modulation slope represents the rate of change of the linear frequency modulated (LFM) signal, determined by the ratio of the LFM signal bandwidth to the frequency modulation period. The calculation is performed on each range-frequency unit to obtain the corresponding range distribution information.

[0034] S206: Perform a Doppler Fast Fourier Transform along the time dimension on the complex spectrum of the range frequency domain of the continuous linear frequency modulated signal period to obtain the range Doppler spectrum, and calculate the target radial velocity based on the Doppler frequency in the range Doppler spectrum in combination with the radar wavelength.

[0035] Specifically, the formula for calculating the target radial velocity is as follows:

[0036] in, Indicates the radial velocity of the target. The radar wavelength is determined by the radar's center frequency. This represents the Doppler frequency in the distance-Doppler spectrum.

[0037] S207: The target's angle of arrival is obtained by calculating the angle of arrival by using the phase difference between the complex spectrum of the range frequency domain of each receiving channel of the radar array at the same range frequency index.

[0038] Specifically, the angle estimation is performed using the MUSIC angle measurement algorithm based on array phase difference. When a uniform linear array is used, the phase difference between adjacent array elements satisfies the array phase relationship with the target incident angle, thereby obtaining the target arrival angle through spatial spectrum search.

[0039] The MUSIC angle measurement algorithm, short for Multiple Signal Classification Algorithm, is a high-resolution direction-of-arrival estimation technique based on eigenvalue decomposition. It utilizes the orthogonality between the signal and noise subspaces to estimate the target's angle of arrival by constructing a spatial spectral function and performing spectral peak search.

[0040] S208: Record the target echo energy amplitude, target distance, target radial velocity, and target angle of arrival as a target observation dataset.

[0041] Furthermore, Z-score standardization is performed on the target observation dataset. A casualty classification control database is established to store the raw and preprocessed target observation data. Standardization is performed independently for four dimensions: target distance, target radial velocity, target angle of arrival, and target echo energy amplitude. The standardization parameters, namely the mean and standard deviation, are statistically updated in real time based on the observation samples within the sliding time window, thus forming an online standardization process. The casualty classification control database is established to store the raw and preprocessed target observation data, and continuous observation sequences are recorded using a time index for subsequent use in multi-target separation, vital sign extraction, and casualty classification assessment modules.

[0042] In this embodiment of the invention, background suppression, range resolution, Doppler resolution, and array angle measurement of millimeter-wave radar echo signals are achieved through coordinated processing. This enables the stable extraction of multi-dimensional observation features such as target range, radial velocity, angle of arrival, and echo intensity from the original echo. Furthermore, the differences in the dimensions of different features are eliminated through sliding time window statistics and Z-score standardization, thereby improving the consistency and comparability of the observation data. This results in the construction of a structured and standardized target observation dataset, providing a reliable data foundation for subsequent multi-target separation, vital sign extraction, and casualty triage assessment. This also improves the stability and feature resolution accuracy of the system in detecting multiple human targets in complex environments.

[0043] S3: Based on the target observation dataset, identify valid echo points in the radar observation space and perform multi-target separation based on the location of the valid echo points.

[0044] In one possible implementation, S3 specifically includes sub-steps S301 to S303: S301: Based on the target observation dataset, the target distance, target radial velocity and target angle of arrival of each group are used to construct a radar observation spatial position, and the corresponding target echo energy amplitude is extracted. When the target echo energy amplitude is greater than the echo threshold, the corresponding radar observation spatial position is recorded as a valid echo point.

[0045] Specifically, the radar observation space uses target range, target radial velocity, and target angle of arrival as three-dimensional spatial coordinate axes to form a three-dimensional radar observation space, which is used to uniformly describe the spatial distribution relationship of each echo point in the range, velocity, and angle dimensions. The echo threshold is preferably adaptively set according to the noise floor level of the radar system and the intensity of environmental clutter. For example, the threshold range can be determined by statistically analyzing the mean and standard deviation of the echo energy amplitude within a sliding time window to eliminate low-energy false echo points caused by environmental noise, wall reflection, or equipment interference, thereby improving the reliability of effective target detection.

[0046] It should be noted that those skilled in the art can set the echo threshold according to actual needs, and this invention does not limit it.

[0047] S302: Combine the target distance, target radial velocity, target angle of arrival, and target echo energy amplitude corresponding to the effective echo points to form a target point cloud set.

[0048] Specifically, each valid echo point is recorded in the form of a four-dimensional feature vector, namely the target distance, target radial velocity, target angle of arrival, and corresponding echo energy amplitude. The above features are combined to form a radar target point cloud set, which is used to describe the distribution of all potential target scattering points detected in the current radar observation frame.

[0049] S303: Perform density clustering on the valid echo points in the target point cloud set according to three spatial features: target distance, target radial velocity, and target angle of arrival. Then, use the DBSCAN algorithm to aggregate spatially close valid echo points into the same target cluster, resulting in a target cluster set.

[0050] DBSCAN, short for Density-Based Spatial Clustering of Applications with Noise, is a density-based spatial clustering algorithm. It groups sufficiently high-density data points into clusters and identifies noise points that do not belong to any cluster.

[0051] Specifically, in the density clustering process, target distance, target radial velocity, and target angle of arrival are used as clustering feature dimensions. Spatial proximity is determined by calculating the feature space distance between each echo point. The DBSCAN algorithm, by setting neighborhood radius and minimum neighborhood sample number parameters, performs density reachability judgment on echo points in the target point cloud set. Echo points whose distance in the feature space is less than the neighborhood radius and meet the minimum neighborhood density condition are clustered into the same target cluster. This method can automatically identify multiple spatially separated target clusters without pre-setting the number of targets and mark isolated noise points as outliers, thereby improving the robustness and accuracy of multi-target separation.

[0052] In this embodiment of the invention, by performing energy threshold screening and density clustering on echo points in the radar observation space, reliable identification and multi-target separation of effective target echoes are achieved. It can automatically aggregate spatially close echo points into corresponding target clusters, while eliminating environmental noise and outliers, thereby improving the accuracy and stability of multi-target detection in complex environments and providing clear and reliable target separation results for subsequent target trajectory tracking and vital sign analysis.

[0053] It should be noted that those skilled in the art can set the energy threshold according to actual needs, and this invention does not limit it.

[0054] S4: Establish target trajectory tracking using the target observation dataset to obtain target state information. Extract chest cavity micro-motion information of the target within the corresponding spatial gating range based on the echo signal, and identify respiratory and heartbeat characteristics as vital sign sequences based on the chest cavity micro-motion information.

[0055] In one possible implementation, S4 utilizes the target observation dataset to establish target trajectory tracking and obtain target state information, specifically including: For each target cluster set, calculate the cluster center position and use the target distance, target radial velocity, and target angle of arrival corresponding to the cluster center as the target space state vector.

[0056] Based on the target space state vector, a target state transition model and an observation model are established.

[0057] Specifically, the target space state vector describes the target's motion state in the radar observation space, and is preferably constructed as a three-dimensional state vector. The target space state vector is updated at each radar observation frame, and the sampling period for the observation frame is preferably the radar frame period, i.e., approximately 50 ms. A target state transition model and an observation model are established based on the target space state vector.

[0058] The target state transition model describes the motion changes of the target between adjacent observation times. A uniform motion model is preferred, and its state transition form is expressed as follows:

[0059] in, This represents the estimated value of the target space state vector at the next moment. This represents the estimated value of the target space state vector at the current moment. This represents the state transition matrix. In the uniform motion model, the state transition matrix is ​​constructed based on the radar sampling period and is used to describe the relationship between the target range and the target radial velocity over consecutive observation times. The process noise vector represents the random deviation between the actual motion state of the target and the target state transition model. The covariance matrix of the process noise vector can be set according to the characteristics of the target motion state change, such as by estimating it based on the statistical variance of the target distance change and the target radial velocity change in a continuous observation sequence, thereby reflecting the uncertainty of the target motion.

[0060] Furthermore, the observation model is used to describe the mapping relationship between actual radar observations and target states, and its form is expressed as:

[0061] in, This represents the target spatial state vector as observed by the radar at the current moment, which includes the observed target range, target radial velocity, and target angle of arrival information. The observation matrix is ​​used to map the target space state vector to the radar observation space. In this embodiment, since the radar observations and state variables are physically identical, the observation matrix is ​​preferably an identity matrix. This represents the estimated value of the target space state vector at the current moment. The observation noise vector is used to characterize the errors generated during radar measurement. The covariance matrix of the observation noise vector is preferably set according to the calibration data of the radar equipment to describe the uncertainty of target motion and radar measurement errors.

[0062] Based on the target state transition model and observation model, the state of each target in the continuous time series is predicted and updated using the Kalman filter algorithm to achieve target trajectory tracking. A unique target identifier is assigned to each target to obtain target state information.

[0063] Specifically, at each observation moment, the state transition model is first used to predict the target's state at the next moment, resulting in a predicted state vector. Then, the radar observation vector is matched and correlated with the predicted state vector, and the predicted state vector is updated and corrected using Kalman filtering, thereby obtaining an estimate of the target's spatial state vector at the current moment, forming a continuous and stable target trajectory. During trajectory establishment, a unique target identifier is assigned to each target to obtain target state information. The target status information includes: target identifier, target distance, target radial velocity, and target angle of arrival.

[0064] Furthermore, during continuous observation, trajectory association is performed by comparing the spatial distance differences between the current observed target and historical trajectories. When the distance difference between an observed target and an existing trajectory is less than a preset association threshold, the observed target is determined to belong to the corresponding trajectory and its original target identifier is maintained. When an observed target cannot be matched with an existing trajectory, a new target identifier is generated, thereby achieving stable target trajectory management in multi-target scenarios.

[0065] It should be noted that those skilled in the art can set the value of the association threshold according to actual needs, and this invention does not limit it.

[0066] In this embodiment of the invention, at each observation moment, the state of the target at the next moment is first predicted using a state transition model to obtain a predicted state vector. Then, the radar observation vector is matched and associated with the predicted state vector, and the predicted state vector is corrected through Kalman filtering updates, thereby obtaining an estimated value of the target spatial state vector at the current moment, forming a continuous and stable target trajectory. During trajectory establishment, a unique target identifier is assigned to each target, obtaining target state information, including: target identifier, target distance, target radial velocity, and target angle of arrival. During continuous observation, trajectory association is performed by comparing the spatial distance difference between the currently observed target and historical trajectories. When the distance difference between an observed target and an existing trajectory is less than a preset association threshold, the observed target is determined to belong to the corresponding trajectory and its original target identifier is maintained. When an observed target cannot be matched with an existing trajectory, a new target identifier is generated, thereby achieving stable target trajectory management in multi-target scenarios.

[0067] In one possible implementation, S4 involves extracting the target's thoracic cavity micro-motion information within the corresponding spatial gating range based on the echo signal, and identifying respiratory and cardiac characteristics as vital signs based on the thoracic cavity micro-motion information. Specifically, this includes: For each target with an assigned identifier, the target location is determined in the radar observation space based on the target distance and target angle of arrival in the corresponding target status information. A spatial gating range consisting of a range gate and an angle gate is set near the corresponding location. Within the spatial gating range, a complex baseband signal sequence corresponding to the target is selected. The complex baseband signal sequence is subjected to phase extraction and phase unfolding processing to obtain a phase change sequence.

[0068] It should be noted that the range gate is used to limit the extraction range of the target in the range dimension, and the angle gate is used to limit the extraction range of the target in the angular dimension. The spatial gating range is preferably set according to the radar range resolution and angular resolution to avoid interference from echoes from nearby targets on the extraction of the current target's micro-motion information. Specifically, the range gate width is preferably set to extend 1-2 range resolution units to each side of the center of the range unit containing the target's range. The angle gate width is preferably set to extend 1-2 angle resolution units to each side of the center of the angle unit containing the target's angle of arrival, thereby forming a local spatial gating region around the target's position.

[0069] Specifically, phase extraction preferably obtains the phase information of the complex signal by calculating the arctangent function of the complex baseband signal, and phase expansion processing is used to eliminate phase jump phenomena in order to restore the continuous phase change process.

[0070] Based on the principle of radar micro-motion measurement, the phase change sequence is converted into a sequence of micro-motion values ​​of human chest cavity displacement.

[0071] Specifically, the micro-displacement value of the human chest cavity and the phase change satisfy the radar phase-displacement conversion relationship. By proportionally converting the phase change using the radar operating wavelength, the minute displacement change of the human chest cavity in the time series can be obtained.

[0072] Time-frequency analysis was performed on the sequence of micro-motion values ​​of human chest cavity displacement within a sliding time window, and the positions of spectral peaks were detected within the range of human respiratory frequency and heart rate, respectively, to obtain the target's respiratory frequency and heart rate.

[0073] Specifically, the sliding time window is preferably set to 10s~30s to ensure sufficient frequency resolution for spectral analysis. The ranges for human respiratory rate and heart rate are initial calibration reference values. The human respiratory rate range is preferably set to 0.1Hz~0.5Hz, corresponding to approximately 6~30 breaths per minute for adults. The human heart rate range is preferably set to 0.8Hz~2.5Hz, corresponding to approximately 48~150 heartbeats per minute for adults. These frequency ranges are set based on statistical laws of human physiological characteristics and can be adjusted appropriately according to different application scenarios. Within the aforementioned preset respiratory and heart rate frequency bands, the spectral distribution is obtained by performing a Fast Fourier Transform on the sequence of micro-motion values ​​of the human chest cavity. The spectral peak with the largest amplitude within the corresponding frequency band is searched as a candidate vital sign frequency, thereby determining the target's respiratory rate and heart rate, respectively.

[0074] Based on respiratory rate and heart rate, a sequence of vital signs of the target is formed.

[0075] Specifically, the vital signs sequence includes time-series changes in respiratory rate and heart rate, which are linked to corresponding target markers for subsequent vital signs stability analysis and casualty grading assessment.

[0076] In this embodiment of the invention, by constructing a spatially gated range formed by a distance gate and an angle gate near the target's location, phase extraction, phase unrolling, and micro-motion displacement conversion are performed on the corresponding complex baseband signal sequence. Spectral detection is then conducted within preset respiratory and cardiac frequency bands, effectively extracting micro-motion information from the target's chest cavity and identifying respiratory and cardiac vital signs. This method suppresses interference from nearby targets and environmental echoes through spatial gating, improving the specificity and stability of micro-motion signal extraction, thereby enhancing the accuracy and reliability of vital sign identification and providing a reliable physiological characteristic data foundation for subsequent casualty status assessment and triage control.

[0077] S5: Comprehensively analyze the spatial location and vital sign sequence of the target, and conduct a consistency correlation assessment on the attribution relationship of vital signs between adjacent targets.

[0078] In one possible implementation, S5 specifically includes sub-steps S501 to S509: S501: For each target, based on the spatial distribution relationship between target distance and target angle of arrival, determine the targets adjacent to the current target in the radar observation space, and construct a neighborhood target set based on the current target and the adjacent targets.

[0079] Specifically, the neighborhood target set describes a group of candidate targets that are spatially close to the current target. Neighborhood determination is preferably based on target distance differences and target angle of arrival differences. When the distance difference between a neighboring target and the current target is less than a preset distance neighborhood threshold and the angle difference is less than a preset angle neighborhood threshold, the target is included in the neighborhood target set. The distance neighborhood threshold is preferably set based on the radar range resolution, for example, 1 to 3 range resolution units. The angle neighborhood threshold is preferably set based on the radar angle resolution, for example, 1 to 2 angle resolution units.

[0080] It should be noted that those skilled in the art can set the size of the distance neighborhood threshold and the angle neighborhood threshold according to actual needs, and this invention does not limit them.

[0081] S502: Based on the sliding time window, calculate the median of the target distance and the target angle of arrival, and subtract the corresponding median from the target distance and the target angle of arrival at the current moment to obtain the target distance deviation and the target angle of arrival deviation.

[0082] S503: Construct a target spatial consistency residual vector using target distance deviation and target angle of arrival deviation.

[0083] Specifically, only target distance and target angle of arrival are chosen to construct the spatial consistency residual vector because they directly characterize the geometric positional relationship of the target in the radar observation space. The target radial velocity, however, mainly reflects the target's motion state and may experience short-term fluctuations due to breathing movements and environmental disturbances in human body micro-motion detection scenarios. Echo intensity, on the other hand, is easily affected by changes in human posture, clothing reflection characteristics, and multipath effects, resulting in relatively low stability. Therefore, distance and angle, two stable spatial dimensions, are preferred in constructing the spatial consistency residual to improve the robustness of spatial consistency determination.

[0084] S504: Based on the target space consistency residual vector within the sliding time window, calculate the covariance matrix to obtain the residual covariance matrix.

[0085] Specifically, the length of the sliding time window is preferably determined based on the radar sampling cycle and the lowest human respiratory rate, in order to ensure that the window contains at least a number of complete respiratory cycles.

[0086] Specifically, when the lowest human respiratory rate is set to 0.1 Hz, the corresponding respiratory cycle is approximately 10 s, and the sliding time window length is preferably set to 10 s to 20 s. When the radar observation frame period is approximately 50 ms, the number of sampling points within the corresponding window is preferably 200 to 400, thereby improving the stability of covariance estimation while ensuring spectral resolution.

[0087] S505: Using the current target's respiratory rate and heart rate as the detection frequency, for each target in the neighborhood target set, extract the human chest cavity displacement micro-motion value sequence within the corresponding sliding time window, and perform single-frequency discrete Fourier transform on the human chest cavity displacement micro-motion value sequence of each target based on the detection frequency to obtain the complex correlation output at the respiratory rate and heart rate.

[0088] Specifically, respiratory rate and heart rate are derived from the dominant spectral peak frequencies of the target obtained through spectral detection in the aforementioned vital sign extraction steps. That is, the spectral peak positions obtained through spectral analysis within the preset respiratory and heart rate frequency bands are used as candidate vital sign frequencies for the current target. Single-frequency discrete Fourier transform is used to detect the response intensity at the frequency corresponding to the currently evaluated target in the micro-motion sequence of neighboring targets, to assess the consistency of neighboring targets at the current vital sign frequency. The squares of the micro-motion values ​​of the human chest cavity displacement of the corresponding target within the sliding time window are accumulated, and the square root is taken to obtain the energy normalization factor. The energy normalization factor is used to characterize the overall energy level of the micro-motion sequence. By normalizing the intensity of the periodic component, the influence of differences in the micro-motion amplitude of different targets on the vital sign detection results can be eliminated, improving the consistency of vital sign characteristic comparisons between different targets.

[0089] S506: Take the absolute value of the complex correlation output to obtain the intensity of the periodic components at the respiratory and heart rate frequencies. Accumulate the squares of the micro-motion values ​​of the human chest cavity displacement within the sliding time window corresponding to the target, and take the square root to obtain the energy normalization factor. Calculate the ratio of the periodic component intensity to the energy normalization factor to obtain the respiratory and heart rate values ​​corresponding to the current detection frequency.

[0090] Specifically, respiratory and cardiac signs are used to characterize the significance of the target's signal at the corresponding vital sign frequency. When the intensity of the periodic component is relatively high relative to the overall energy, it indicates that the target has obvious vital sign characteristics at the corresponding frequency, which can then be used for subsequent target sign consistency analysis and target binding determination.

[0091] S507: Multiply the transpose of the target space consistency residual vector by the inverse of the residual covariance matrix, and then multiply by the target space consistency residual vector itself to obtain the space consistency metric.

[0092] Specifically, the target spatial consistency residual vector is composed of the target distance deviation and the target angle of arrival deviation, used to characterize the degree of deviation of the current target from the center of the spatial location distribution of the neighboring target set. The residual covariance matrix is ​​used to describe the statistical dispersion of the spatial residual distribution in the neighboring target set, thus measuring the target spatial consistency in the form of Mahalanobis distance. When the residual covariance matrix becomes irreversible due to a small sample size or collinearity of the residual distribution, regularization is preferred. A small regularization term can be added to the diagonal of the covariance matrix to ensure the stability of the spatial consistency measurement calculation process.

[0093] S508: Divide the spatial consistency metric by two and take the opposite number to perform a natural exponent operation to obtain the spatial consistency index term.

[0094] Specifically, the exponential form corresponds to the exponential decay structure in the Gaussian probability density function, and is used to map the spatial consistency measure in Mahalanobis distance form to a spatial consistency exponential term with a value range of (0, 1], thereby transforming spatial distance differences into consistency weights in a probabilistic sense. The smaller the target spatial residual, the closer the spatial consistency exponential term is to 1; the larger the target spatial residual, the faster the spatial consistency exponential term decays.

[0095] S509: Multiply the corresponding spatial consistency index, respiratory signs, and heart rate signs to obtain the target binding score. Within the neighborhood target set, calculate the binding score for each target in the set and sum them as the denominator. Divide the target binding score by the denominator to obtain the target binding confidence value.

[0096] It should be noted that both respiratory and cardiac signs are intensity indicators obtained through energy normalization, with their values ​​preferably normalized to between 0 and 1 to maintain the same dimensional scale as the spatial consistency index. This normalization process ensures that the spatial consistency index, respiratory signs, and cardiac signs are all on a dimensionless, unified scale, guaranteeing the comparability of the binding score obtained by multiplying them. Within the neighborhood target set, a binding score is calculated for each target and accumulated as the denominator. The binding score of the target is then divided by the denominator to obtain the target binding confidence value.

[0097] Specifically, the neighborhood target set is a set of candidate targets that satisfy the neighborhood constraint in spatial location with the current target. The neighborhood range is determined based on the differences in target distance and angle. When there are no neighborhood targets, the neighborhood target set only includes the current evaluation target itself. The vital sign target binding confidence value is calculated using the above normalization method so that the sum of the vital sign target binding confidence values ​​of all targets in the neighborhood target set equals 1, thereby forming a probabilistic distribution of target vital sign attribution, used to determine the target most likely corresponding to the vital sign signal.

[0098] Specifically, the formula for the confidence value of binding vital signs to targets is as follows:

[0099] in, This represents the target binding confidence value, which is used to comprehensively assess the consistency of target spatial location and vital sign characteristics within the neighborhood target set, quantitatively evaluate the reliability of the matching between vital signs and each candidate target, and obtain the target binding confidence value of the current vital sign through normalized competition. This represents the target spatial consistency residual vector, used to describe the degree of deviation of the current target from the spatial statistical center. This represents the transpose of the target spatial consistency residual vector, used to calculate the spatial consistency metric with the inverse of the covariance matrix. This represents the inverse of the residual covariance matrix, used for statistical scaling normalization of spatial residuals. This indicates the current target's breathing rate. This indicates the current target's heart rate. It represents respiratory sign values ​​and measures the intensity of the corresponding respiratory sign response. This represents the heart rate value, which measures the intensity of the corresponding heart rate response. N represents the set of neighboring targets. This represents the target index in the neighborhood target set, used to represent each candidate target in the neighborhood target set.

[0100] For example, given the current detection frequency, the normalized denominator is 1.0472. The table details the spatial consistency index, respiratory sign values, heart rate signs, and target-sign binding confidence values ​​for five targets within the same neighborhood target set. Specifically, target 1 has a spatial consistency index of 0.82, a respiratory sign value of 0.90, a heart rate sign value of 0.85, and a target-sign binding confidence value of 0.5968. Target 2 has a spatial consistency index of 0.60, a respiratory sign value of 0.88, a heart rate sign value of 0.55, and a target-sign binding confidence value of 0.2763. Target 3 has a spatial consistency index of 0.35, a respiratory sign value of 0.40, a heart rate sign value of 0.72, and a target-sign binding confidence value of 0.0959. The spatial consistency index for objective 4 is 0.15, the respiratory signs value is 0.75, the heartbeat signs value is 0.21, and the target binding confidence value for vital signs is 0.0225. The spatial consistency index for objective 5 is 0.05, the respiratory signs value is 0.30, the heartbeat signs value is 0.60, and the target binding confidence value for vital signs is 0.0086.

[0101] Reference manual attached Figure 2This diagram illustrates a confidence distribution of target binding in a neighborhood target characteristic according to an embodiment of the present invention.

[0102] Given that the current detection target is Target 1, the respiratory rate and heart rate extracted from Target 1 are used as the detection frequency. A vital sign consistency assessment is performed on each target in the neighborhood target set, and the spatial consistency index of each target is combined to calculate the vital sign target binding confidence value. In the figure, the horizontal axis represents the target number in the neighborhood target set, the vertical axis represents the vital sign target binding confidence value of the corresponding target, the bars represent the binding confidence level of each target, and the dashed line represents the preset confidence threshold. (Refer to Table 1 and...) Figure 3 It can be seen that the target binding confidence value for the vital signs corresponding to Target 1 is 0.5968, which is significantly higher than that of other targets in the neighborhood set and significantly higher than the confidence threshold. The target binding confidence values ​​for the vital signs of Targets 2, 3, 4, and 5 are all lower than the confidence threshold and gradually decrease. This indicates that the currently detected vital signs have the highest consistency with the spatial location and micro-motion characteristics of Target 1. Therefore, it can be determined that the current vital sign signal indeed belongs to Target 1, thus completing the reliable binding between the vital signs and the target, while avoiding the aliasing problem of vital sign attribution caused by spatial proximity or spectral overlap between neighboring targets.

[0103] It should be noted that those skilled in the art can set the confidence threshold according to actual needs, and this invention does not limit it.

[0104] In this embodiment of the invention, a spatial consistency metric is constructed and jointly evaluated using respiratory and cardiac signs. This unifies and quantifies the spatial relationship between targets and vital sign characteristics. The confidence level for target binding is then calculated using normalization, enabling probabilistic determination of the attribution relationship of vital sign signals within a neighborhood target set. This method suppresses interference from targets with large spatial differences through a spatial consistency index term, while simultaneously enhancing the recognition capability of true vital sign signals using normalized vital sign intensity. This improves the accuracy and stability of binding vital signs to target identities in multi-target scenarios.

[0105] S6: Based on the consistency association assessment results, identify the matching reliability between vital signs and targets, and rebind vital signs that may be mixed.

[0106] In one possible implementation, S6 specifically includes sub-steps S601 and S602: S601: Calculate the target binding confidence value of each target in the neighborhood target set and filter out the target binding confidence value with the highest target binding confidence value.

[0107] Specifically, the confidence value of the vital sign target binding is derived from the normalized binding result calculated from the aforementioned spatial consistency index, respiratory vital sign value, and heartbeat vital sign value. Its value ranges from 0 to 1 and is used to represent the matching probability between the current vital sign signal and each candidate target.

[0108] S602: Determine whether the target binding confidence value with the highest vital sign is greater than the confidence threshold. If so, determine the target with the highest vital sign target binding confidence value as the priority matching target for the current vital sign, and bind and update the target identifier of the current vital sign with the priority matching target. Otherwise, mark it as an uncertain vital sign.

[0109] Specifically, the confidence threshold is preferably set based on historical data statistics to determine the reliability of the match between the current vital sign signal and the target. When the confidence value of the maximum vital sign target binding exceeds the confidence threshold, it indicates that the current vital sign signal has a significant belonging advantage in the neighborhood target set, thus determining that the target is the priority matching target for vital signs.

[0110] Furthermore, the current vital signs are bound and updated with the target identifiers of the priority matching targets. The currently detected respiratory and heart rates are associated with the target identifiers of the corresponding priority matching targets and recorded. The vital sign feature sequence of the target is then updated in the target status information to ensure that the vital sign data is consistent with the target trajectory. Otherwise, it is marked as an uncertain sign. When the maximum vital sign target binding confidence value does not reach the preset confidence threshold, it indicates that the current vital sign signal has aliasing or insufficient discrimination among multiple neighboring targets. In this case, the vital sign is marked as an uncertain sign, and the target binding operation is temporarily suspended. At the same time, the confidence value of the vital sign target binding can be re-evaluated using newly added data in subsequent observation frames, thereby gradually improving the reliability of the vital sign binding determination during continuous observation.

[0111] In this embodiment of the invention, by comparing the confidence levels of vital sign targets within a neighboring target set and introducing a confidence threshold determination mechanism, the reliability of the matching between vital sign signals and targets can be effectively evaluated, and the binding update of vital signs and target identifiers can be completed when the matching advantage is obvious. Simultaneously, when the attribution of a vital sign is unclear, uncertain vital signs are marked and re-evaluated in subsequent observations, thereby avoiding erroneous bindings caused by the overlap of multiple target vital signs and improving the accuracy and stability of the matching results between vital signs and targets.

[0112] S7: Based on the vital signs of the re-bound target, determine the casualty level through the casualty grading assessment model, and generate a casualty rescue priority sequence according to the casualty level.

[0113] In one possible implementation, S7 specifically includes sub-steps S701 to S704: S701: For each target whose vital signs have been updated, construct the casualty status input vector based on the corresponding respiratory rate and heart rate.

[0114] S702: Construct a casualty grading assessment model using a vital sign threshold rule base. Input the casualty status input vector into the casualty grading assessment model, assess the target's vital sign status according to the rule base, and output the corresponding casualty level.

[0115] The rule base is a predefined collection of knowledge consisting of multiple "if-then" judgment rules.

[0116] S703: When the casualty level reaches the danger level, the casualty alarm signal is triggered, and the targets are sorted in descending order of casualty level to obtain the casualty rescue priority sequence.

[0117] S704: Based on the target distance and target angle of arrival, determine the target's position in the radar observation space, display the target position, casualty level, and casualty rescue priority sequence on the monitoring terminal interface, and output a casualty alarm signal.

[0118] It should be noted that the vital sign threshold rule base is preferably constructed based on the normal physiological range of the human body and emergency medical experience, and is used to determine the range of respiratory rate and heart rate. The normal range of respiratory rate is preferably 0.15Hz~0.35Hz, approximately 9~21 breaths / minute, and the normal range of heart rate is preferably 1.0Hz~1.8Hz, approximately 60~108 beats / minute. When the target's respiratory rate or heart rate deviates from the corresponding normal range, the corresponding level of abnormality of vital signs is determined according to the degree of deviation.

[0119] It should be noted that those skilled in the art can set the size of the vital sign threshold rule library according to actual needs, and this invention does not limit it.

[0120] Specifically, the casualty status is represented by a tiered numerical rule, dividing the casualty status into three levels: Level 1 (Normal) indicates that the target's vital signs are within the normal physiological range, with a respiratory rate between 0.15Hz and 0.35Hz and a heart rate between 1.0Hz and 1.8Hz. Level 2 (Attention) indicates that the target's vital signs deviate slightly from the normal range, with a respiratory rate between 0.10Hz and 0.15Hz or between 0.35Hz and 0.50Hz, or a heart rate between 0.8Hz and 1.0Hz or between 1.8Hz and 2.2Hz. Level 3 (Danger) indicates that the target's vital signs deviate from the normal range but are still within the monitorable range. Level 3 (Danger) indicates that the target's vital signs are in danger and require priority rescue, with a respiratory rate significantly deviating from the normal range (preferably less than 0.10Hz or greater than 0.50Hz) or a heart rate significantly deviating from the normal range (preferably less than 0.8Hz or greater than 2.2Hz).

[0121] It should be noted that the various ranges of respiratory rate and heart rate are determined based on human vital sign statistics and can be appropriately adjusted according to specific application scenarios.

[0122] Specifically, the injury level values ​​are preferably encoded in ascending order of danger, for example, Level 1 = 1, Level 2 = 2, Level 3 = 3, thus ensuring that the higher the value, the higher the risk of injury. When the injury level of a target reaches the danger level, an injury alarm signal is triggered, and targets are sorted in descending order of injury level to obtain a priority sequence for injury rescue.

[0123] Specifically, sorting casualty severity levels from highest to lowest indicates that targets with higher danger levels have higher rescue priority. When multiple targets have casualty levels of the same severity, a secondary sorting is performed based on target distance, for example, prioritizing closer targets to improve rescue response efficiency. The target's position in the radar observation space is determined based on the target distance and target angle of arrival. The target position, casualty severity information, and casualty rescue priority sequence are displayed on the monitoring terminal interface, and a casualty alarm signal is output.

[0124] Specifically, the monitoring terminal interface preferably displays the target location using a two-dimensional planar radar observation map, and different injury levels can be distinguished by color coding, such as green for normal level, yellow for level of concern, and red for level of danger, so as to facilitate rescuers to quickly identify high-risk targets.

[0125] In this embodiment of the invention, a casualty grading assessment model based on a vital sign threshold rule base is constructed. For targets whose vital signs have been updated, a casualty status input vector is built based on their respiratory rate and heart rate. Casualties are then classified into normal, concern, and danger levels according to grading numerical rules, achieving automated and standardized assessment of the vital sign status of multiple casualties. When a danger level is determined, a casualty alarm signal is triggered immediately, and a rescue priority sequence is generated by sorting casualty levels from highest to lowest. Simultaneously, the target location, casualty level, and rescue priority are intuitively displayed on the monitoring terminal interface using a two-dimensional radar observation map combined with color indicators. This significantly improves the accuracy of casualty grading and the response efficiency of rescue dispatch in complex scenarios, effectively solving the problem of misjudgment caused by incorrect binding of vital signs and targets in existing technologies, and providing a clear and reliable decision-making basis for on-site rescue command.

[0126] Reference manual attached Figure 3 The diagram shows a schematic of a wounded soldier classification control system based on multi-target radar vital sign analysis provided by the present invention.

[0127] The present invention also provides a casualty classification control system 20 based on multi-target radar vital sign analysis, applied to the above-mentioned casualty classification control method based on multi-target radar vital sign analysis, comprising: Processor 201.

[0128] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201, they implement the casualty classification control method based on multi-target radar vital sign analysis as described in the method embodiment.

[0129] The casualty classification control system 20 based on multi-target radar vital sign analysis provided by the present invention can execute the above-mentioned casualty classification control method based on multi-target radar vital sign analysis and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate further.

[0130] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0131] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0132] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0133] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0134] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0135] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0136] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0137] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0138] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0139] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0140] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0141] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the casualty triage control method based on multi-target radar vital sign analysis as described in the method embodiment.

[0143] The present invention provides a computer-readable storage medium that can implement the steps and effects of the wounded triage control method based on multi-target radar vital sign analysis in the above-described method embodiments. To avoid repetition, the present invention will not elaborate further.

[0144] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0145] The following points need to be explained: (1) The accompanying drawings of the embodiments of the present invention only involve the structures involved in the embodiments of the present invention. Other structures can refer to the general design.

[0146] (2) For clarity, the thickness of layers or regions is enlarged or reduced in the drawings used to describe embodiments of the invention, i.e., these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element or there may be intermediate elements.

[0147] (3) Where there is no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0148] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for graded control of wounded soldiers based on multi-target radar vital sign analysis, characterized in that, include: S1: Real-time acquisition of echo signals from millimeter-wave radar; S2: Perform background suppression, range analysis, Doppler analysis, and array angle measurement on the echo signal to extract observation features and construct a standardized target observation dataset; S3: Based on the target observation dataset, identify effective echo points in the radar observation space, and perform multi-target separation according to the location of the effective echo points; S4: Establish target trajectory tracking using the target observation dataset to obtain target state information. Extract chest cavity micro-motion information of the target within the corresponding spatial gating range based on the echo signal. Identify respiratory and heartbeat features as vital sign sequences based on the chest cavity micro-motion information. S5: By comprehensively analyzing the spatial location of the target and the vital sign sequence, a consistency correlation assessment is conducted on the attribution relationship of vital signs between adjacent targets; S6: Based on the consistency association assessment results, identify the matching reliability between vital signs and targets, and rebind vital signs that may be mixed. S7: Based on the vital signs of the re-bound target, determine the casualty level through the casualty grading assessment model, and generate a casualty rescue priority sequence according to the casualty level.

2. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 1, characterized in that, S2 specifically includes: S201: The echo signal is quadrature demodulated through the radar receiver to obtain the complex baseband signal sequence of each receiving channel; S202: Within the sliding time window, the mean of the complex baseband signal in the complex baseband signal sequence is calculated, and the mean value is subtracted from the original complex baseband signal to obtain the complex baseband signal with static background removed; S203: Perform a fast Fourier transform on the complex baseband signal sequence with static background removed within each linear frequency modulation signal period to obtain the range-frequency domain complex spectrum corresponding to each receiving channel; S204: Calculate the amplitude of the range-frequency complex spectrum to obtain the target echo energy amplitude; S205: Based on the radar frequency modulation slope and the speed of light, the range frequency in the range frequency domain complex spectrum is converted to obtain the corresponding target range; S206: Perform a Doppler fast Fourier transform along the time dimension on the complex spectrum of the range frequency domain of the continuous linear frequency modulated signal period to obtain the range Doppler spectrum, and calculate the target radial velocity based on the Doppler frequency in the range Doppler spectrum in combination with the radar wavelength. S207: The target angle of arrival is obtained by calculating the angle of arrival by using the phase difference between the complex spectrum of the range frequency domain of each receiving channel of the radar array at the same range frequency index; S208: Record the target echo energy amplitude, the target distance, the target radial velocity, and the target angle of arrival as the target observation dataset.

3. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 2, characterized in that, S3 specifically includes: S301: Based on the target observation dataset, construct a radar observation spatial position by combining the target distance, the target radial velocity, and the target angle of arrival for each group, and extract the corresponding target echo energy amplitude. When the target echo energy amplitude is greater than the echo threshold, record the corresponding radar observation spatial position as the effective echo point. S302: Combine the target distance, target radial velocity, target angle of arrival, and target echo energy amplitude corresponding to the effective echo points to form a target point cloud set; S303: Perform density clustering on the effective echo points in the target point cloud set according to the three spatial features of the target distance, the target radial velocity, and the target angle of arrival; and use the DBSCAN algorithm to aggregate the effective echo points that are close to each other in space into the same target cluster to obtain a target cluster set.

4. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 3, characterized in that, In step S4, establishing target trajectory tracking using the target observation dataset and obtaining target state information specifically includes: For each target cluster set, the cluster center position is calculated, and the target distance, target radial velocity, and target angle of arrival corresponding to the cluster center are used as the target space state vector; Based on the target space state vector, a target state transition model and an observation model are established; Based on the target state transition model and the observation model, the state of each target in a continuous time series is predicted and updated using the Kalman filter algorithm to achieve target trajectory tracking, and a unique target identifier is assigned to each target to obtain the target state information.

5. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 1, characterized in that, In step S4, the extraction of chest cavity micro-motion information of the target within the corresponding spatial gating range based on the echo signal, and the identification of respiratory and cardiac characteristics as vital signs based on the chest cavity micro-motion information, specifically includes: For each target with an assigned identifier, the target location is determined in the radar observation space based on the target distance and target angle of arrival in the target status information. A spatial gating range consisting of a range gate and an angle gate is set near the corresponding location. Within the spatial gating range, a complex baseband signal sequence corresponding to the target is selected. The complex baseband signal sequence is subjected to phase extraction and phase unfolding processing to obtain a phase change sequence. Based on the principle of radar micro-motion measurement, the phase change sequence is converted into a sequence of micro-motion values ​​of human chest cavity displacement; Time-frequency analysis was performed on the sequence of micro-motion values ​​of the human chest cavity displacement within the sliding time window, and the positions of spectral peaks were detected within the range of human respiratory frequency and heart rate, respectively, to obtain the target's respiratory frequency and heart rate. Based on the respiratory rate and the heart rate, a sequence of vital signs of the target is formed.

6. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 1, characterized in that, S5 specifically includes: S501: For each target, based on the spatial distribution relationship between target distance and target angle of arrival, determine the targets adjacent to the current target in the radar observation space, and construct a neighborhood target set based on the current target and the adjacent targets; S502: Based on the sliding time window, calculate the median of the target distance and the target angle of arrival respectively, and subtract the corresponding median from the target distance and the target angle of arrival at the current moment to obtain the target distance deviation and the target angle of arrival deviation; S503: Construct a target spatial consistency residual vector using the target distance deviation and the target angle of arrival deviation; S504: Based on the target space consistency residual vector within the sliding time window, calculate the covariance matrix to obtain the residual covariance matrix; S505: Using the respiratory rate and heart rate of the current target as the detection frequency, for each target in the neighborhood target set, extract the human chest cavity displacement micro-motion value sequence within the corresponding sliding time window, and perform single-frequency discrete Fourier transform on the human chest cavity displacement micro-motion value sequence of each target based on the detection frequency to obtain the complex correlation output at the respiratory rate and heart rate. S506: Take the absolute value of the complex correlation output to obtain the intensity of the periodic components at the respiratory frequency and the heart rate; accumulate the squares of the human chest cavity displacement micro-motion value sequence of the corresponding target within the sliding time window, and take the square root to obtain the energy normalization factor; calculate the ratio of the periodic component intensity to the energy normalization factor to obtain the respiratory sign value and heart rate sign value corresponding to the current detection frequency. S507: Multiply the transpose of the target spatial consistency residual vector by the inverse of the residual covariance matrix, and then multiply by the target spatial consistency residual vector itself to obtain the spatial consistency metric. S508: Divide the spatial consistency metric by two and take the opposite number to perform a natural exponent calculation to obtain the spatial consistency index term; S509: Multiply the corresponding spatial consistency index, respiratory signs, and heartbeat signs to obtain the target binding score; within the neighborhood target set, calculate the binding score for each target in the set and sum them as the denominator; divide the target's binding score by the denominator to obtain the target binding confidence value.

7. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 6, characterized in that, S6 specifically includes: S601: Calculate the target binding confidence value of the vital signs for each target in the neighborhood target set, and select the target binding confidence value with the highest vital signs. S602: Determine whether the target binding confidence value of the maximum vital sign is greater than the confidence threshold; if so, determine the target with the target binding confidence value of the maximum vital sign as the priority matching target of the current vital sign, and bind and update the target identifier of the current vital sign with the priority matching target; otherwise, mark it as an uncertain vital sign.

8. The method for graded control of wounded soldiers based on multi-target radar vital sign analysis according to claim 1, characterized in that, Specifically, S7 includes: S701: For each target whose vital signs binding update has been completed, construct the casualty status input vector based on the corresponding respiratory rate and heart rate; S702: Construct the wounded soldier classification assessment model through the vital sign threshold rule base, input the wounded soldier status input vector into the wounded soldier classification assessment model, assess the vital sign status of the target according to the rule base, and output the corresponding wounded soldier level; S703: When the level of the injured person reaches the danger level, trigger the injured person alarm signal, and sort the targets according to the injured person level from large to small to obtain the injured person rescue priority sequence; S704: Based on the target distance and target angle of arrival, determine the target's position in the radar observation space, display the target position, the casualty level, and the casualty rescue priority sequence on the monitoring terminal interface, and output the casualty alarm signal.

9. A wounded soldier triage control system based on multi-target radar vital sign analysis, characterized in that, include: processor; A memory storing computer-readable instructions, which, when executed by the processor, implement the wounded triage control method based on multi-target radar vital sign analysis as described in any one of claims 1 to 8.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the wounded triage control method based on multi-target radar vital sign analysis as described in any one of claims 1 to 8.

Images