Multi-target life detection method and system based on radar signals

Abstract

The invention relates to the technical field of life detection, and discloses a radar signal-based multi-target life detection method, which comprises the following steps of 1, collecting a multi-target mixed signal received by a radar, and preprocessing the signal; step 2, performing phase space reconstruction on the preprocessed signal to obtain a high-dimensional phase space trajectory; 3, constructing a self-adaptive dictionary, and carrying out sparse representation on the signals; and step 4, separating initial vital sign signals of each target from the mixed signals through a sparse optimization algorithm. According to the invention, the technical scheme of multi-band radar signal and deep learning fusion is adopted, and the technical effect of high-precision separation of multi-target vital sign signals in a complex electromagnetic environment is achieved. Compared with a technical scheme depending on single frequency band or fixed threshold classification in the prior art, the method solves the defects of serious signal aliasing and low separation precision in a dense obstacle or multi-target mixed scene.

Description

Technical Field

[0001] The present invention relates to the technical field of life detection, and in particular to a multi-target life detection method and system based on radar signals. Background Art

[0002] Radar signal processing technology has important application value in the field of vital signs monitoring, especially in natural disaster relief, medical health monitoring and military reconnaissance, and can provide non-contact and efficient detection methods. By analyzing the vital signs signals (such as breathing rate and heart rate) in the radar reflection echo, the target can be remotely detected and located. However, with the complexity of application scenarios, especially in multi-target dense areas and complex electromagnetic environments, existing technologies face problems such as severe signal aliasing and low positioning accuracy, which makes it difficult to meet actual needs.

[0003] At present, the mainstream radar life detection technology is mainly based on the principles of synthetic aperture radar (SAR) and Doppler radar. These technologies extract vital signs signals by analyzing reflected echoes and show good performance in single targets or simple environments. For example, Doppler radar can effectively extract breathing and heartbeat signals by detecting the frequency shift caused by target movement; SAR technology improves resolution through synthetic aperture processing to detect stationary targets. In addition, existing technologies usually use Fourier transform or fixed threshold classifiers to process signals to achieve target separation and positioning.

[0004] Existing technologies are particularly prone to signal aliasing in complex electromagnetic environments. In areas with densely populated targets, radar signals are susceptible to multipath effects and noise interference, resulting in the aliasing of vital signs signals from multiple targets. Traditional methods (such as Fourier transform) have limited processing capabilities for nonlinear and non-stationary signals and cannot effectively separate signals with overlapping spectra; while fixed threshold classifiers have a high misjudgment rate in time-varying environments and are difficult to accurately distinguish vital signs signals from different individuals. These problems severely limit the application effect of existing technologies in complex scenarios, and a new technical solution that can effectively solve the problem of signal aliasing is urgently needed. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a multi-target life detection method and system based on radar signals. The prior art has the problem that multi-target vital sign signals are severely aliased and difficult to accurately separate and locate in a complex electromagnetic environment.

[0006] To achieve the above objectives, the present invention is implemented by the following technical solutions: A multi-target life detection method based on radar signals, comprising the following steps: Step 1: Collect multi-target mixed signals received by the radar and pre-process the signals; Step 2: Reconstruct the phase space of the preprocessed signal to obtain a high-dimensional phase space trajectory; Step 3: construct an adaptive dictionary and perform sparse representation on the signal; Step 4: Separate the preliminary vital sign signals of each target from the mixed signal through a sparse optimization algorithm; Step 5: Use the nonlinear dynamic model to correct the preliminary separation signal, eliminate residual aliasing, and obtain the final separation signal; Step 6: Based on the final separated signal, calculate the target position and extract vital sign information; Step 7: Output target location information and vital signs signal.

[0007] Preferably, the preprocessing in step 1 includes: filtering the signal to remove high-frequency noise and baseline drift, and normalizing the signal.

[0008] Preferably, the adaptive dictionary in step three is constructed by a K-SVD algorithm, which specifically includes: optimizing dictionary atoms and sparse coefficients by minimizing the reconstruction error of a linear combination of a signal and dictionary atoms.

[0009] Preferably, the sparse optimization algorithm in step 4 is an orthogonal matching pursuit algorithm, which specifically includes: gradually approaching the optimal sparse representation of the signal by iteratively selecting dictionary atoms.

[0010] Preferably, the nonlinear dynamic model in step five is used to describe the coupling relationship between signals, specifically including: eliminating residual aliasing in the preliminary separated signal by a correction function.

[0011] Preferably, the target position calculation in step six is ​​implemented by a time-reversal positioning algorithm, specifically including: calculating the spatial position of the target by constructing a time-reversal channel response function.

[0012] A multi-target life detection system based on radar signals, comprising: A signal acquisition module is used to collect mixed signals of multiple targets received by the radar, and its output end is connected to the signal preprocessing module; A signal preprocessing module is used to filter and normalize the signal, and its output end is connected to the phase space reconstruction module; A phase space reconstruction module is used to reconstruct the phase space of the preprocessed signal, and its output end is connected to the dictionary construction module; A dictionary building module is used to build an adaptive dictionary and perform sparse representation on the signal, and its output end is connected to the signal separation module; A signal separation module, used to separate the preliminary vital sign signals of each target through a sparse optimization algorithm, and its output end is connected to the signal correction module; A signal correction module, used for eliminating residual aliasing by using a nonlinear dynamics model, and an output end of which is connected to a target positioning module; A target positioning module is used to calculate the target position and extract vital sign information, and its output end is connected to the result output module; Result output module, used to output target location information and vital signs signals Preferably, the signal separation module includes an orthogonal matching pursuit algorithm unit, which is used to gradually approximate the optimal sparse representation of the signal by iteratively selecting dictionary atoms.

[0013] Preferably, the target positioning module includes a time-reversal positioning algorithm unit, which is used to calculate the spatial position of the target by constructing a time-reversal channel response function.

[0014] The present invention provides a multi-target life detection method and system based on radar signals, which has the following beneficial effects: 1. The present invention adopts the technical solution of multi-band radar signal fusion and deep learning, achieving the technical effect of high-precision separation of multi-target vital sign signals in complex electromagnetic environments. Compared with the technical solution relying on single frequency band or fixed threshold classification in the prior art, it solves the shortcomings of severe signal aliasing and low separation accuracy in dense obstacle or multi-target mixed scenes.

[0015] 2. The present invention adopts the technical solution of time-reversal positioning algorithm and multi-sensor fusion to achieve the technical effect of sub-meter positioning accuracy at long distances. Compared with the technical solution of traditional radar positioning with large error in the prior art, it solves the problem of insufficient positioning accuracy in densely blocked environments.

[0016] 3. The present invention adopts the technical solution of dynamic adaptive filtering and noise suppression model to achieve the technical effect of stable operation of the system under strong noise background. Compared with the technical solution of fixed filtering or single noise processing in the prior art, it solves the shortcomings of high false alarm rate and poor system stability in high noise environment.

[0017] 4. The present invention adopts the technical solution of layered processing framework and GPU accelerated computing, achieving the technical effect of an average processing delay of less than 2 seconds. Compared with the technical solution with complex processing flow and long response time in the prior art, it solves the problem that it cannot meet the demand in scenarios with high real-time requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the method flow of the present invention; Figure 2 It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Please refer to the attached Figure 1-2 The embodiment of the present invention provides a multi-target life detection method and system based on radar signals, including: Signal acquisition and preprocessing are the basic steps of the entire technical solution. They need to ensure the quality and consistency of the input signal and provide a reliable data foundation for subsequent key processing such as phase space reconstruction and sparse representation.

[0021] In this embodiment, the specific implementation of signal acquisition and preprocessing is as follows: A multi-band composite radar system is used for signal acquisition, in which the X-band (8-12GHz) is used for high-resolution surface target detection, and the L-band (1-2GHz) is used for obstacle penetration detection. The radar transmitter uses multiple-input multiple-output (MIMO) technology to transmit time-frequency-space three-dimensional encoded pulse signals, with an adjustable pulse width range of 10ns to 1μs and a pulse repetition frequency (PRF) adaptive range of 1kHz to 10MHz to meet the detection needs in different scenarios. The receiving end is equipped with a 16-channel MIMO array with a noise figure of less than 2dB and a dynamic range of more than 80dB, ensuring that weak vital signs signals can still be captured in a strong interference environment.

[0022] Signal preprocessing includes filtering and normalization. For high-frequency noise suppression, a bandpass filter with a passband range of 0.1Hz to 5Hz is used to retain the effective frequency band of breathing and heartbeat signals. The elimination of baseline drift is achieved through wavelet transform, for example, the db4 wavelet basis function is used for 5-layer decomposition to remove the low-frequency drift component in the signal. Normalization is achieved through the following formula: ; in, is the original signal; is the signal mean; is the signal standard deviation, This is the normalized signal. This process makes the signal amplitudes of different channels and frequency bands uniform to the same level, avoiding the amplitude deviation problem in subsequent processing.

[0023] To adapt to complex electromagnetic environments, adaptive noise cancellation technology can be further introduced. For example, the environmental noise signal is collected through the reference channel, and the filter coefficients are updated in real time using the least mean square algorithm to subtract the noise component from the main signal. The update formula is: ; in, is the filter weight vector, is the step size factor, is the error signal, is the reference noise signal. Through this method, the signal-to-noise ratio can be significantly improved, providing a purer input signal for subsequent processing.

[0024] The radar system parameters can be dynamically adjusted according to the actual scenario. For example, the L-band is used first when penetrating concrete walls, and the X-band is switched when high-resolution positioning is required. In addition, the signal sampling rate is set to 2GSPS (gigasamples per second) to ensure the complete capture of high-frequency signals. The pre-processed signal is transmitted to the processing unit through a high-speed optical fiber interface to avoid signal attenuation and interference during transmission.

[0025] In some embodiments, in response to the special needs of multi-target scenarios, preliminary screening of blind source separation technology can be introduced in the preprocessing stage. For example, the mixed signal is roughly separated by independent component analysis to screen out potential vital sign signal components, which are then handed over to subsequent modules for refined processing. Such extended technology can further improve the robustness of the system in multi-target dense scenarios.

[0026] After signal acquisition and preprocessing, the present invention further transforms the one-dimensional time series signal into a trajectory in a high-dimensional space through phase space reconstruction technology, providing richer structural information for subsequent signal separation and feature extraction. Phase space reconstruction is a key step in solving the problem of nonlinear signal aliasing. Its core lies in mapping the dynamic characteristics of the signal into a high-dimensional space through the embedding theorem, thereby revealing the inherent laws of the signal.

[0027] In this embodiment, the specific implementation method of phase space reconstruction is as follows: First, the delay time is determined by the autocorrelation function method . Autocorrelation function The calculation formula is: ; in, is the preprocessed signal, is the signal length, is the delay time. Select the first zero crossing point of the autocorrelation function As the optimal delay time, it ensures that the trajectory of the reconstructed phase space can be fully developed.

[0028] The embedding dimension m is determined by the false nearest neighbor method. The specific steps include: gradually increasing the embedding dimension, calculating the false nearest neighbor ratio, and stopping the increase when the ratio is less than a preset threshold (for example, 5%). This method can effectively avoid excessive folding or redundancy of phase space trajectories.

[0029] In some embodiments, according to Takens embedding theorem, the preprocessed signal Reconstruct the phase space and obtain the high-dimensional phase space trajectory : ; in, is the embedding dimension, is the delay time. Through phase space reconstruction, the dynamic characteristics of the signal are mapped into a high-dimensional space, providing a more intuitive geometric structure for subsequent signal separation.

[0030] To further improve the reconstruction effect, the local projection method can be introduced to smooth the phase space trajectory. For example, the high-dimensional trajectory is projected into the low-dimensional space through the local linear embedding (LLE) algorithm, retaining the main dynamic characteristics of the signal while removing the trajectory jitter caused by noise. Its objective function is: ; in, is the projection point in the low-dimensional space; is the local weight coefficient; is the number of neighbor points.

[0031] In some embodiments, for multi-target scenarios in complex electromagnetic environments, multi-scale phase space reconstruction technology can be further introduced. The signal is decomposed into multiple scale components through wavelet transform, and phase space reconstruction is performed separately, and finally the reconstruction results of each scale are fused. This method can effectively capture the multi-scale characteristics of the signal and improve the accuracy of subsequent signal separation.

[0032] In some embodiments, to verify the effect of phase space reconstruction, the chaotic characteristics of the signal can be evaluated by calculating the maximum Lyapunov exponent. The Lyapunov exponent is calculated using the Wolf algorithm: ; in, is the distance between two points on the phase space trajectory, is the initial distance. If the Lyapunov exponent is positive, it indicates that the signal has chaotic characteristics and is suitable for processing using nonlinear dynamics methods.

[0033] The results of phase space reconstruction can be displayed through visualization technology. Projecting high-dimensional trajectories into three-dimensional space, observing their geometric structure and dynamic evolution laws. Such visualization methods not only help to understand signal characteristics, but also provide an intuitive basis for the optimization of subsequent algorithm parameters.

[0034] Through the above-mentioned phase space reconstruction technology, the present invention can transform the one-dimensional time series signal into a trajectory in a high-dimensional space, laying a solid foundation for subsequent signal separation and feature extraction.

[0035] After completing the phase space reconstruction, the present invention transforms the high-dimensional phase space trajectory into a sparse coefficient combination by constructing an adaptive dictionary and sparsely representing the signal, providing a mathematical basis for subsequent signal separation. This step aims to utilize the sparsity characteristics of the signal to transform the complex aliasing problem into a solvable sparse optimization problem, thereby effectively separating multi-target vital sign signals.

[0036] In some embodiments, the K-SVD algorithm is used to construct an adaptive dictionary The algorithm minimizes the signal reconstruction error by alternately optimizing dictionary atoms and sparse coefficients. Its objective function is: ; in, is the high-dimensional trajectory after phase space reconstruction, is the dictionary matrix to be optimized, is a sparse coefficient vector, is the sparsity constraint, controlling the maximum number of non-zero elements in the coefficients, is the total number of signal samples.

[0037] In some embodiments, the dictionary can be initialized using a random matrix or a predefined basis function (such as a Gabor basis or a wavelet basis). The atoms are generated by random Gaussian distribution and gradually optimized through iterative updates. Each iteration includes two stages: sparse coding and dictionary update: Sparse coding stage: fixed dictionary , using the orthogonal matching pursuit (OMP) algorithm to solve the sparse coefficients ; Dictionary update phase: fixed sparse coefficients , update dictionary atoms column by column , the combination of atoms and their corresponding coefficients is optimized by singular value decomposition (SVD).

[0038] In some embodiments, the dictionary atom update formula is: ; in, Indicates to remove the current atom The residual matrix after is the Frobenius norm, is the row vector of the corresponding sparse coefficients. Decomposed by SVD , the updated atom Pick The first column of , the sparse coefficient is updated to .

[0039] In some embodiments, in order to improve the adaptability of the dictionary to vital sign signals, physiological feature prior knowledge can be introduced, and typical waveforms of breathing and heartbeat signals (such as sine waves and Morlet wavelets) can be used as initial dictionary atoms, and then fine-tuned through the K-SVD algorithm. Such methods can accelerate algorithm convergence and improve the interpretability of sparse representation.

[0040] In some embodiments, a block dictionary construction strategy can be adopted to address the complexity of multi-objective scenarios. For example, the phase space trajectory is divided into multiple sub-blocks, and sub-dictionaries are constructed respectively, and global optimization is achieved through joint sparse constraints. Its objective function is expanded to: ; in, for Sub-block signals, is the corresponding sub-dictionary, is the global sparsity constraint.

[0041] In some embodiments, the result of the sparse representation can be evaluated by residual analysis. With the original signal The mean square error is: ; When the MSE is lower than a preset threshold (e.g., 10 −4 ), the sparse representation is judged to have reached the convergence criterion.

[0042] In some embodiments, to further enhance robustness, a noise-aware sparse representation model may be introduced. For example, a norm constraint of the noise term is added to the objective function: in, is the regularization parameter to balance the reconstruction error and sparsity. Such improvements can effectively suppress the influence of noise on the sparse coefficient and improve the stability of signal separation.

[0043] Through the above-mentioned dictionary construction and sparse representation technology, the present invention can convert high-dimensional phase space trajectories into sparse coefficient combinations, provide a mathematical basis for subsequent signal separation, and adapt to the needs of multi-target detection in complex electromagnetic environments.

[0044] Nonlinear dynamic correction: After completing the sparse signal separation, the present invention corrects the preliminary separated signal through the nonlinear dynamic model to eliminate residual aliasing and obtain the final separated signal. This step aims to utilize the nonlinear dynamic characteristics to further optimize the signal separation results and improve the extraction accuracy of multi-target vital sign signals.

[0045] In this embodiment, the specific implementation method of nonlinear dynamics correction is as follows: In some embodiments, a nonlinear dynamic model is first constructed to describe the coupling relationship between signals. For example, a nonlinear coupling function is used to describe the interaction between target signals, and its form includes linear terms, quadratic terms, and cubic terms, and the coupling coefficient is determined by least squares fitting.

[0046] The nonlinear dynamic model is used to correct the initial separation signal. The specific steps are as follows: Calculate coupling signals: Calculate the coupling effect of each target signal on other signals based on the nonlinear coupling function; Eliminate residual aliasing: Subtract the coupled signal from the initial separation signal to obtain the corrected signal. This method can effectively eliminate residual aliasing between signals and improve separation accuracy.

[0047] To improve the correction accuracy, an adaptive coupling coefficient update mechanism can be introduced. The coupling coefficient is dynamically adjusted through the gradient descent method to minimize the correction error. This method can optimize the correction effect in real time according to the signal characteristics.

[0048] In view of the complexity of multi-target scenarios, a phased correction strategy can be adopted. The signal is divided into multiple frequency bands, and nonlinear dynamic correction is performed separately. Finally, the correction results of each frequency band are fused. This method can effectively deal with the nonlinear aliasing problem of the signal and improve the separation accuracy. To further enhance the robustness, a correction model for noise suppression can be introduced. In the correction process, the norm constraint of the noise term is added to balance the correction error and noise suppression. This method can effectively suppress the influence of noise on the correction results and improve the stability of the system. To verify the effect of nonlinear dynamic correction, the signal separation accuracy can be evaluated by spectral analysis. For example, the spectral difference of the signal before and after correction is calculated to ensure that the frequency components of the main vital signs are retained, while eliminating the pseudo-frequency components introduced by aliasing.

[0049] Through the above-mentioned nonlinear dynamic correction, the present invention can further optimize the signal separation result, eliminate residual aliasing, improve the extraction accuracy of multi-target vital sign signals, and provide reliable input for subsequent target positioning and vital sign extraction.

[0050] Target positioning and vital signs extraction: After completing the nonlinear dynamics correction, the present invention calculates the target position through the target positioning algorithm and extracts the vital signs signal. This step aims to use the corrected separation signal to achieve high-precision target positioning and vital signs information extraction, providing key data for the final result output.

[0051] In this embodiment, the specific implementation methods of target positioning and vital sign extraction are as follows: In some embodiments, a time-reversal positioning algorithm (TIMLE) is used to calculate the target position. The algorithm accurately calculates the spatial coordinates of the target by constructing a time-reversal channel response function and combining the time-reversal characteristics of the multipath signal. The channel impulse response is estimated by the correlation between the received signal and the reference signal, and time-reversed processing is performed on it to finally determine the target position.

[0052] In order to improve positioning accuracy, multi-sensor fusion technology is introduced. The positioning results of multiple radar nodes are weighted and fused, and the weight coefficient is dynamically adjusted according to the signal-to-noise ratio or distance to obtain the optimal target position. This method can effectively reduce the positioning error of a single node and improve the robustness of the system.

[0053] The extraction of vital signs signals is achieved through spectrum analysis. Fast Fourier transform is performed on the corrected separated signal to detect the peak frequency in the spectrum and determine the respiratory and heart rates. By setting the frequency band threshold (such as 0.1Hz-0.5Hz for respiratory frequency band and 0.8Hz-2Hz for heart rate band), noise interference is eliminated to ensure the validity of the extraction results.

[0054] Adaptive filtering technology is introduced to smooth the spectrum. For example, the Kalman filter is used to dynamically update the state estimation, remove the pseudo-frequency components caused by noise, and retain the real vital sign frequency. This method can adjust the filtering parameters in real time according to the signal characteristics and adapt to the changes in the complex electromagnetic environment. In view of the complexity of the multi-target scene, a frequency band extraction strategy is adopted. The signal is divided into different frequency bands, and the breathing and heartbeat features are extracted respectively, and the final vital sign information is determined through joint optimization. This method can avoid cross-interference between frequency bands and improve the extraction accuracy in multi-target scenes.

[0055] Through the above-mentioned target positioning and vital sign extraction technology, the present invention can achieve high-precision target positioning and vital sign information extraction, and provide reliable data support for the final result output.

[0056] Result output and visualization: After completing target positioning and vital sign extraction, the present invention presents the target location information and vital sign signals to the user in an intuitive form through the result output and visualization module. This step aims to convert complex technical processing results into easy-to-understand visual information to provide support for rescue or monitoring decisions. Through efficient visualization design, users can quickly obtain key information, improving the practicality of the system and user experience.

[0057] In this embodiment, the specific implementation method of result output and visualization is as follows: The target location information is visualized through a two-dimensional or three-dimensional map. The target location is mapped to a geographic coordinate system to generate a heat map or a point distribution map. The heat map calculates the spatial distribution of the target signal strength and intuitively displays the target dense area. The calculation is based on the target signal strength and location coordinates The Gaussian kernel function is used to generate a smooth hotspot distribution. Users can quickly identify target-dense areas by color depth, making it easier to quickly locate key areas at the rescue site.

[0058] Vital sign signals are visualized through time series graphs or spectrograms. Plot the respiratory signal and heartbeat signal as time series graphs respectively, and mark the main frequency components. The time series graph can intuitively show the trend of the target's vital signs, making it easier for users to analyze the target's life status. The spectrogram is generated through fast Fourier transform, highlighting the respiratory and heartbeat frequencies, helping users to quickly identify the target's physiological characteristics. For example, the respiratory frequency is usually between 0.1Hz and 0.5Hz, while the heartbeat frequency is between 0.8Hz and 2Hz. By setting the frequency band threshold, noise interference can be eliminated to ensure the effectiveness of the extraction results.

[0059] A dynamic update mechanism is introduced to ensure the real-time visualization results. The target position and vital signs are updated at fixed intervals (e.g., 1 second) to reflect the latest status of the target. This method can meet the needs of real-time monitoring and is particularly suitable for scenarios with high timeliness requirements such as disaster relief or military reconnaissance. The dynamic update mechanism is implemented through asynchronous communication between background data processing and front-end display, ensuring a smooth and responsive user interface.

[0060] In view of the complexity of multi-target scenarios, a hierarchical visualization strategy is adopted. Targets are divided into multiple levels according to signal strength or distance, and displayed in different colors or marks. By setting the signal strength threshold, high-priority targets and low-priority targets are distinguished, which makes it easier for users to quickly locate key targets. Targets with signal strength higher than the threshold are marked in red, indicating that they need priority rescue; targets with medium signal strength are marked in yellow, indicating that further confirmation is required; targets with low signal strength are marked in green, indicating that they have a lower priority. This hierarchical design can help users make decisions quickly in complex scenarios.

[0061] Introducing interactive visualization functions to enhance the flexibility of user operations. Users can view detailed information (such as target distance, vital sign frequency, etc.) by clicking on the target marker, or view historical data by sliding the timeline. This method can meet the diverse needs of users and improve the practicality of the system. The interactive function is implemented through a front-end framework (such as React or Vue.js) to ensure that the user interface is responsive and the operation is smooth. In addition, users can adjust the map display range by zooming, dragging, and other operations to further optimize the user experience.

[0062] For multi-target scenarios in complex electromagnetic environments, multimodal fusion visualization technology is introduced. The radar detection results are fused with other sensor data (such as infrared images and video streams) to provide more comprehensive target information. This method can make up for the limitations of a single sensor and improve the accuracy of target recognition. In disaster relief scenarios, radar detection results can be superimposed with infrared images taken by drones to help rescuers locate trapped people more accurately. Multimodal fusion is achieved through data synchronization and coordinate alignment technology to ensure the consistency of data from different sensors in time and space.

[0063] To further improve the visualization effect, data compression and optimization technology can be introduced. Principal component analysis (PCA) can be used to reduce data dimensions and reduce the amount of calculation for front-end rendering to ensure smooth display in large-scale data scenarios. In addition, WebGL and other graphics rendering technologies can be used to improve the rendering efficiency of maps and charts, providing users with a smoother interactive experience.

[0064] Through the above-mentioned result output and visualization technology, the present invention can transform complex technical processing results into intuitive visual information, providing users with clear and real-time target location and vital sign data support. This efficient visualization design not only improves the practicality of the system, but also provides strong support for users' decision-making.

[0065] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-target life detection method based on radar signals, characterized in that: The following steps are involved: Step 1: Collect multi-target mixed signals received by the radar and pre-process the signals; Step 2: reconstruct the phase space of the preprocessed signal to obtain a high-dimensional phase space trajectory; Step 3: construct an adaptive dictionary and perform sparse representation on the signal; Step 4: Separate the preliminary vital sign signals of each target from the mixed signal through a sparse optimization algorithm; Step 5: Use the nonlinear dynamic model to correct the preliminary separation signal, eliminate residual aliasing, and obtain the final separation signal; Step 6: Based on the final separated signal, calculate the target position and extract vital sign information; Step 7: Output target location information and vital signs signal.

2. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The preprocessing in step 1 includes: filtering the signal to remove high-frequency noise and baseline drift, and normalizing the signal.

3. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The phase space reconstruction described in step 2 is implemented by Takens embedding theorem, which specifically includes: determining the delay time according to the autocorrelation function method, and determining the embedding dimension by the false nearest neighbor method.

4. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The adaptive dictionary described in step 3 is constructed by the K-SVD algorithm, which specifically includes: optimizing the dictionary atoms and sparse coefficients by minimizing the reconstruction error of the linear combination of the signal and the dictionary atoms.

5. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The sparse optimization algorithm described in step 4 is an orthogonal matching pursuit algorithm, which specifically includes: gradually approaching the optimal sparse representation of the signal by iteratively selecting dictionary atoms.

6. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The nonlinear dynamic model described in step five is used to describe the coupling relationship between signals, specifically including: eliminating residual aliasing in the preliminary separated signal through a correction function.

7. The multi-target life detection method based on radar signals according to claim 1, characterized in that: The target position calculation described in step six is ​​implemented by a time-reversal positioning algorithm, which specifically includes: calculating the spatial position of the target by constructing a time-reversal channel response function.

8. A multi-target life detection system based on radar signals, according to a multi-target life detection method based on radar signals according to any one of claims 1 to 7, characterized in that: include: The signal acquisition module is used to collect the multi-target mixed signal received by the radar, and its output end is connected to the signal preprocessing module; A signal preprocessing module is used to filter and normalize the signal, and its output end is connected to the phase space reconstruction module; A phase space reconstruction module is used to reconstruct the phase space of the preprocessed signal, and its output end is connected to the dictionary construction module; A dictionary building module is used to build an adaptive dictionary and perform sparse representation on the signal, and its output end is connected to the signal separation module; A signal separation module is used to separate the preliminary vital sign signals of each target through a sparse optimization algorithm, and its output end is connected to the signal correction module; A signal correction module, used for eliminating residual aliasing by using a nonlinear dynamics model, and an output end of which is connected to a target positioning module; A target positioning module is used to calculate the target position and extract vital sign information, and its output end is connected to the result output module; The result output module is used to output target position information and vital sign signals.

9. The multi-target life detection system based on radar signals according to claim 8, characterized in that: The signal separation module includes an orthogonal matching pursuit algorithm unit, which is used to gradually approach the optimal sparse representation of the signal by iteratively selecting dictionary atoms.

10. The multi-target life detection system based on radar signals according to claim 8, characterized in that: The target positioning module includes a time-reversal positioning algorithm unit, which is used to calculate the spatial position of the target by constructing a time-reversal channel response function.

Images