A non-line-of-sight target detection method based on intelligent metasurfaces on curved roads

Abstract

This invention relates to a non-line-of-sight target detection method based on a smart metasurface on curved roads, comprising: setting up a smart metasurface device for relaying the detection of non-line-of-sight targets within a non-line-of-sight range; illuminating the smart metasurface device with a detection radar; and modulating the beam emitted by the detection radar based on the smart metasurface device to change the beam's propagation path and illuminate the non-line-of-sight target; the detection radar receiving the echo signal returned by the smart metasurface device and performing non-line-of-sight target detection and early warning based on the echo signal. This invention can accurately obtain the distance, speed, and angle information of non-line-of-sight (weak) targets, and has advantages such as all-weather and 24-hour operation, high sensing accuracy, wide sensing range, and low construction cost.

Description

Technical Field

[0001] This invention relates to the field of traffic safety, and more particularly to a non-line-of-sight target detection method based on intelligent metasurfaces on curved roads. Background Technology

[0002] Road traffic safety has become a global concern.

[0003] Curved intersections, acting as bottlenecks in road traffic systems, possess complex traffic characteristics and are often accident hotspots. Currently, blind spots at curved intersections are primarily addressed by using convex mirrors to expand the driver's field of vision. However, this method is susceptible to light and weather conditions and cannot meet the urgent needs of future smart cities, such as those requiring autonomous driving. Intelligent metasurfaces, a current research hotspot, combined with existing vehicle-mounted radar sensors, can utilize electromagnetic waves to achieve all-weather, all-day, non-line-of-sight detection, expanding the field of information perception. Simultaneously, this information can be converted into digital data to meet the requirements of future smart cities, including autonomous driving.

[0004] Smart metasurfaces possess powerful beam manipulation capabilities, enabling control over multi-dimensional information such as time, space, and spectrum of electromagnetic waves. As a low-cost wireless relay, they have been extensively studied in the field of wireless communication. For example, Chinese patent application number 202210151425.7 discloses a design method for a communication system based on an omnidirectional smart metasurface. However, a communication system only requires one metasurface reflection for both transmitting and receiving signals. In contrast, radar systems typically have their transmit and receive ports located at the same position, requiring two metasurface reflections and one target reflection for both signals. This means that non-line-of-sight target detection based on smart metasurfaces will face severe challenges due to low signal-to-noise ratios and clutter. Therefore, research on its application in radar relay transmission is relatively limited. Summary of the Invention

[0005] The purpose of this invention is to provide a non-line-of-sight target detection method based on intelligent metasurfaces on curved roads.

[0006] To achieve the above-mentioned objectives, this invention provides a non-line-of-sight target detection method based on intelligent metasurfaces on curved roads, comprising the following steps:

[0007] S1. An intelligent metasurface device for relay detection of non-line-of-sight targets is set up for non-line-of-sight range.

[0008] S2. Illuminating the intelligent metasurface device with a detection radar, and modulating the beam emitted by the detection radar based on the intelligent metasurface device to change the propagation path of the beam and illuminate a non-line-of-sight target; wherein the intelligent metasurface device modulates the beam with the strongest echo signal as the target.

[0009] S3. The detection radar receives the echo signal transmitted back by the intelligent metasurface device, and performs non-line-of-sight target detection and early warning based on the echo signal.

[0010] According to one aspect of the present invention, in step S3, the step of the detection radar receiving the echo signal transmitted back by the intelligent metasurface device and performing non-line-of-sight target detection and early warning based on the echo signal includes:

[0011] S31. The detection radar receives the echo signal transmitted back by the intelligent metasurface device, preprocesses the echo signal, and obtains multidimensional information of the non-line-of-sight target; wherein, the multidimensional information includes: distance information, velocity information, and angle information;

[0012] S32. Generate a range Doppler image based on the preprocessed echo signal;

[0013] S33. Input the distance Doppler image into a pre-trained recognition model, and perform non-line-of-sight target detection and early warning based on the recognition model.

[0014] According to one aspect of the present invention, in step S2, in the step of the intelligent metasurface device controlling the beam with the goal of the strongest echo signal, the strength of the echo signal is determined based on the constructed optimization problem model.

[0015] The optimization problem model includes: the unit incident electric field strength of the intelligent metasurface device and the radar receiving power of the detection radar.

[0016] According to one aspect of the invention, the elemental incident electric field intensity is constructed using a narrow-band assumption, expressed as follows:

[0017]

[0018] Among them, E in Indicates the incident electric field strength of the unit cell;

[0019] j represents the imaginary unit;

[0020] Z0 represents the characteristic impedance of the medium;

[0021] φ1 represents the phase term related to the propagation path between the detection radar and the smart metasurface device, as well as the phase response of the radar transmitting antenna;

[0022] The spatial power density of plane waves at the intelligent metasurface device is expressed as:

[0023]

[0024] Among them, PT G represents the peak power of the detection radar. T F represents the gain of the radar transmitting antenna. R (θ R ,φ R ) indicates the direction of radar observation (θ) R ,φ R The normalized radar radiated power pattern, r1=||S R -S RIS || represents the distance between the detection radar and the smart metasurface device, S R S indicates the location of the detection radar. RIS Indicates the location of the intelligent metasurface device;

[0025] S1 represents the steering matrix of the intelligent metasurface device related to the direction of incoming waves, expressed as:

[0026]

[0027] in, , Let u represent the manifold vectors of the vertical and horizontal intelligent metasurface devices, where u RIS =sin(θ) RIS cos(φ) RIS ), υ RIS =sin(θ) RIS sin(φ) RIS ) represent the direction cosines, λ0 represents the radar operating wavelength, and (θ) represent the direction cosines. RIS ,φ RIS ) indicates the orientation of the intelligent metasurface device. Indicates the relative orientation of the detection radar and the intelligent metasurface device;

[0028] The radar receiving power of the detection radar is constructed based on reciprocity conditions, where, assuming the intelligent metasurface device has zero loss and maximum beam pointing gain, the radar receiving power under ideal conditions is expressed as:

[0029]

[0030] Among them, P rx This indicates the radar receiving power of the detection radar;

[0031] G represents antenna power gain;

[0032] F tot To merge items of the same type, it is represented as:

[0033]

[0034] in, Indicates the relative orientation of the detection radar and the smart metasurface device. The normalized unit radiation power pattern of the intelligent metasurface device. Indicates the relative direction between the non-line-of-sight target and the intelligent metasurface device. The unit radiation power pattern of the normalized intelligent metasurface device;

[0035] η RIS Indicates unit efficiency;

[0036] r2 represents the distance between the radar and the non-line-of-sight target, expressed as: r2=||s T -s RIS ||, where S T Indicates the position of the phase center of a non-line-of-sight target;

[0037] σ represents the reflectivity of the target power along the line-of-sight direction of the intelligent metasurface device.

[0038] According to one aspect of the present invention, in step S31, the step of the detection radar receiving the echo signal transmitted back by the intelligent metasurface device and preprocessing the echo signal, wherein the preprocessing is used to suppress clutter in the echo signal, and includes:

[0039] S311. Perform radio frequency interference suppression on the received echo signal and generate a first processing signal;

[0040] S312. Perform moving target detection on the first processed signal, suppress static clutter in the first processed signal, and generate a second processed signal;

[0041] S313. Process the second processing signal to obtain the non-line-of-sight target.

[0042] According to one aspect of the present invention, step S313, the step of processing the second processing signal to obtain the non-line-of-sight target, includes:

[0043] S3131. Perform parallel processing on the second processed signal, wherein the second processed signal is processed by the first branch to obtain the first target information, and the clutter map is established by the second branch and the second target information is obtained by using dynamic clutter suppression.

[0044] S3132. Summarize the first target information and the second target information to describe the non-line-of-sight target.

[0045] According to one aspect of the present invention, step S3131, the step of processing the second processing signal using a first branch to obtain the first target information, includes:

[0046] The second processed signal is incoherently accumulated;

[0047] The second processed signal after incoherent accumulation is detected using CFAR to obtain the first target information.

[0048] According to one aspect of the present invention, in step S3131, in the step of establishing a clutter map using a second branch and obtaining second target information using a dynamic clutter suppression method, the second target information includes external clutter information and internal clutter information, then it includes:

[0049] S3131a. Dynamic clutter suppression is performed on the clutter outside the clutter map to obtain the clutter outside information;

[0050] S3131b. Compare the extraneous information with a preset threshold. If the extraneous information meets the preset threshold, determine that the extraneous information is target information and merge the extraneous information with the first target information.

[0051] S3131c. Target information is filtered for the corresponding unit amplitude of the clutter in the clutter map using a guard window to obtain the clutter information; wherein the target information included in the clutter information is target information that is slow and small in velocity and scale and / or easily confused with clutter, which is smaller than the first target information.

[0052] According to one aspect of the present invention, in step S3131b, in the step of comparing the clutter information with a preset threshold, if the clutter information does not meet the preset threshold, then the clutter information is determined to be non-target information and the clutter information is processed together with the clutter information in step S3131c to filter out the target information that is slow, small and / or easily confused with clutter.

[0053] According to one aspect of the present invention, in step S31, in the step of obtaining multidimensional information of the non-line-of-sight target, the distance information and the velocity information are obtained based on the preprocessed echo signal, and the angle information is obtained based on the intelligent metasurface device.

[0054] According to one aspect of the present invention, the present invention can accurately obtain the distance, speed, and angle information of non-line-of-sight (weak) targets, and has advantages such as all-weather and all-day operation, high sensing accuracy, wide sensing range, and low construction cost.

[0055] According to one aspect of the present invention, the present invention solves the target detection problem in low signal-to-clutter environments by preprocessing the radar echo signal. This effectively improves the echo gain at the desired target location, suppresses clutter interference from undesired directions, and improves the system's detection performance.

[0056] According to one aspect of the present invention, the present invention can enable vehicle-mounted radar to detect and warn of non-line-of-sight targets at curved road intersections, and has a wider range of application prospects. Attached Figure Description

[0057] Figure 1 This is a flowchart illustrating the steps of a non-line-of-sight target detection method according to an embodiment of the present invention;

[0058] Figure 2 This is an application scenario diagram of a non-line-of-sight target detection method according to an embodiment of the present invention;

[0059] Figure 3 This is a flowchart of a non-line-of-sight target detection method according to an embodiment of the present invention;

[0060] Figure 4 This is a flowchart of echo signal preprocessing according to an embodiment of the present invention;

[0061] Figure 5 This is a diagram showing the radar sensing results when a non-line-of-sight target moves without a metasurface.

[0062] Figure 6 This is a diagram showing the radar sensing results when a non-line-of-sight target moves with a metasurface. Detailed Implementation

[0063] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The embodiments cannot be described in detail here, but the embodiments of the present invention are not limited to the following embodiments.

[0064] Combination Figure 1 , Figure 2 and Figure 3 As shown, according to one embodiment of the present invention, a non-line-of-sight target detection method based on a smart metasurface on a curved road includes the following steps:

[0065] S1. An intelligent metasurface device for relay detection of non-line-of-sight targets is set up for non-line-of-sight range.

[0066] S2. Illuminating the intelligent metasurface device with a detection radar, and modulating the beam emitted by the detection radar based on the intelligent metasurface device to change the beam's propagation path and illuminate a non-line-of-sight target; wherein the intelligent metasurface device modulates the beam with the strongest echo signal as the target.

[0067] S3. The detection radar receives the echo signal transmitted back by the intelligent metasurface device and performs non-line-of-sight target detection and early warning based on the echo signal.

[0068] Combination Figure 1 , Figure 2 and Figure 3As shown, according to one embodiment of the present invention, in step S1, the intelligent metasurface device for relay detection of non-line-of-sight targets is set up for the non-line-of-sight range. The intelligent metasurface device is controlled by a matching FPGA module. The FPGA module has a data memory and a fast modulation rate (up to 100 ns) and stable voltage control (0.8 V). It can implement flexible encoding sequence switching in the encoding sequence database, meeting the requirement for efficient and adjustable units in the intelligent metasurface device. In this embodiment, the intelligent metasurface device has the characteristics of large bandwidth (up to 1 GHz) and wide angle (range -60° to +60°). In this embodiment, the non-line-of-sight range can be the curve of a road.

[0069] Combination Figure 1 , Figure 2 and Figure 3 As shown, according to one embodiment of the present invention, in step S2, the intelligent metasurface device is illuminated by a detection radar, and the beam emitted by the detection radar is controlled based on the intelligent metasurface device. The detection radar can be installed in a fixed location or on a mobile vehicle (e.g., a vehicle). The beam emitted by the detection radar is used to achieve detection and early warning of targets outside the line-of-sight range. In this embodiment, by installing the intelligent metasurface device at a road bend, blind spots are filled. Since the transmitting and receiving antennas of the detection radar are located on the same side, the emitted beam can be reflected by the intelligent metasurface device to reach the target in the blind spot.

[0070] In this embodiment, the intelligent metasurface device uses a supporting FPGA module to adjust the position of its constituent units, thereby achieving precise control over the radar beam propagation direction and target illumination intensity. In step S2 of this embodiment, where the intelligent metasurface device controls the beam with the strongest echo signal as the target, the strength of the echo signal is determined based on a constructed optimization problem model.

[0071] The optimization problem model includes the unit incident electric field strength of the intelligent metasurface device and the radar receiving power of the detection radar.

[0072] In this embodiment, the incident electric field strength of the unit cell is constructed based on the narrow-band assumption, which is expressed as:

[0073]

[0074] Among them, E in Indicates the incident electric field strength of the unit cell;

[0075] j represents the imaginary unit, i.e.

[0076] Z0 represents the characteristic impedance of the medium;

[0077] φ1 represents the phase term related to the propagation path between the detection radar and the smart metasurface device, as well as the phase response of the radar transmitting antenna;

[0078] The spatial power density of plane waves at the intelligent metasurface device is expressed as:

[0079]

[0080] Among them, P T G represents the peak power of the detection radar. T F represents the gain of the radar transmitting antenna. R (θ R ,φ R ) indicates the direction of radar observation (θ) R ,φ R The normalized radar radiated power pattern, r1=||s R -s RIS || represents the distance between the detection radar and the intelligent metasurface device, s R Indicates the location of the detection radar, s RIS Indicates the location of the intelligent metasurface device;

[0081] S1 represents the steering matrix of the intelligent metasurface device related to the direction of incoming waves, expressed as:

[0082]

[0083] in, , Let v represent the manifold vectors of the vertical and horizontal intelligent metasurface devices, where v RIS =sin(θ) RIS cos(φ) RIS ), υ RIS =sin(θ) RIS sin(φ) RIS ) represent the direction cosines, λ0 represents the radar operating wavelength, and (θ) represent the direction cosines. RIS ,φ RIS ) indicates the orientation of the intelligent metasurface device. This indicates the relative direction between the detection radar and the intelligent metasurface device; M represents the row number of the intelligent metasurface device; N represents the column number of the intelligent metasurface device.

[0084] The radar received power of the detection radar is constructed based on the reciprocity condition (i.e., assuming that the intelligent metasurface device does not change its programmable reflection coefficient on the reverse path, and that the radar antenna patterns are the same during the transmission and reception phases, i.e., reciprocity holds). To simplify the formula and eliminate the explicit dependence on angle, the radar received power is expressed under ideal conditions as follows, assuming that the intelligent metasurface device has zero loss and maximum beam pointing gain:

[0085]

[0086] Among them, P rx This indicates the radar receiving power of the detection radar;

[0087] G represents antenna power gain;

[0088] F tot To merge items of the same type, it is represented as:

[0089]

[0090] Where F(θ, φ) represents the normalized unit radiation power pattern of the intelligent metasurface device in the gaze direction (θ, φ), specifically, Indicates the relative orientation of the detection radar and the smart metasurface device. The normalized unit radiation power pattern of the intelligent metasurface device. Indicates the relative direction between the non-line-of-sight target and the intelligent metasurface device. The unit radiation power pattern of the normalized intelligent metasurface device;

[0091] η RIS Indicates unit efficiency;

[0092] r2 represents the distance between the radar and the non-line-of-sight target, expressed as: r2=||s T -s RIS ||, where S T Indicates the position of the phase center of a non-line-of-sight target;

[0093] σ represents the monostable RCS of the intelligent metasurface device along the LOS direction, i.e., the reflectivity of the target power along the line-of-sight direction.

[0094] Combination Figure 1 , Figure 2 and Figure 3As shown, according to one embodiment of the present invention, during the radar detection process assisted by the intelligent metasurface device, three reflections occur (wherein, the first signal reflection: the radar is reflected from the metasurface to the target; the second signal reflection: the metasurface is reflected from the target back to the metasurface; and the third signal reflection: the target is reflected from the metasurface back to the radar). This means that radar detection based on the intelligent metasurface device will inevitably face the severe challenge posed by the low signal-to-noise ratio. Furthermore, in step S3, the step of the detection radar receiving the echo signal returned by the intelligent metasurface device and performing non-line-of-sight target detection and early warning based on the echo signal includes:

[0095] S31. The detection radar receives the echo signal transmitted back by the intelligent metasurface device, preprocesses the echo signal, and acquires multi-dimensional information of non-line-of-sight targets; among which, the multi-dimensional information includes: range information, velocity information, and angle information;

[0096] S32. Generate a range Doppler image based on the preprocessed echo signal;

[0097] S33. Input the distance Doppler image into the pre-trained recognition model, and perform non-line-of-sight target detection and early warning based on the recognition model.

[0098] Combination Figure 1 , Figure 2 , Figure 3 , Figure 4 As shown, according to one embodiment of the present invention, in step S31, the step of receiving the echo signal from the smart metasurface device by the detection radar and preprocessing the echo signal, wherein the preprocessing is used to suppress clutter in the echo signal, includes:

[0099] S311. Perform radio frequency interference suppression on the received echo signal and generate a first processing signal;

[0100] S312. Perform moving target detection on the first processed signal, suppress static clutter in the first processed signal, and generate a second processed signal;

[0101] S313. Process the second processing signal to obtain a non-line-of-sight target.

[0102] Combination Figure 1 , Figure 2 , Figure 3 , Figure 4 As shown, according to one embodiment of the present invention, step S313, the step of processing the second processing signal to obtain a non-line-of-sight target, includes:

[0103] S3131. Perform parallel processing on the second processed signal, wherein the second processed signal is processed by the first branch to obtain the first target information, and the clutter map is established by the second branch and the second target information is obtained by using dynamic clutter suppression.

[0104] S3132. Summarize the information on the first target and the second target to describe the non-line-of-sight target.

[0105] The above settings effectively improve the real-time operating speed and efficiency of this solution, achieving the beneficial effect of rapid target tracking.

[0106] Combination Figure 1 , Figure 2 , Figure 3 , Figure 4 As shown, according to one embodiment of the present invention, step S3131, the step of processing the second processing signal using the first branch to obtain the first target information, includes:

[0107] Incoherent accumulation is performed on the second processed signal;

[0108] The second-processed signal, after incoherent accumulation, is detected using CFAR to obtain first target information; where the first target information can be represented as a vehicle driving in the blind zone. CFAR stands for Constant False Alarm Rate, a radar signal detection method that automatically adjusts the radar's sensitivity to maintain a constant false alarm probability when external interference intensity changes.

[0109] like Figure 4 As shown, according to one embodiment of the present invention, in step S3131, in the step of establishing a clutter map using the second branch and obtaining the second target information using a dynamic clutter suppression method, the second target information includes external clutter information and internal clutter information, which includes:

[0110] S3131a. Dynamic clutter suppression is performed using the clutter map method to obtain clutter information in the clutter map;

[0111] S3131b. Compare the extraneous information with a preset threshold. If the extraneous information meets the preset threshold, determine that the extraneous information is the target information and merge the extraneous information with the first target information.

[0112] S3131c. Target information is filtered based on the corresponding unit amplitude of the guard window within the clutter map to obtain clutter information; wherein, the target information included in the clutter information is slow and small target information with a velocity and scale smaller than the first target information and / or easily confused with clutter. In this embodiment, slow and small target information and / or easily confused with clutter can be pedestrians, etc., in blind spots.

[0113] like Figure 4As shown, according to one embodiment of the present invention, in step S3131b, in the step of comparing the clutter information with a preset threshold, if the clutter information does not meet the preset threshold, the clutter information is determined to be non-target information and the clutter information is processed together with the clutter information in step 3131c to filter out target information that is slow, small and / or easily confused with clutter.

[0114] In this embodiment, in step S313, the step of processing the second processing signal to obtain non-line-of-sight targets can obtain various target information (i.e., first target information about the driving vehicle, and target information that is slow and small and / or easily confused with clutter) based on the above steps. By summarizing and classifying the various target information, the final detection result of non-line-of-sight targets in the blind spot can be formed.

[0115] like Figure 3 As shown, according to one embodiment of the present invention, in step S31, the step of acquiring multidimensional information of a non-line-of-sight target, the distance information and velocity information are obtained based on the preprocessed echo signal, and the angle information is obtained based on the intelligent metasurface device. In this embodiment, as mentioned above, the intelligent metasurface device modulates the beam with the target of the strongest echo signal. Therefore, when the echo signal is modulated to its strongest state, the angle information of the non-line-of-sight target in the blind zone can be obtained based on the angle of the unit on the intelligent metasurface device.

[0116] To further illustrate the technical effects of the present invention, examples are provided.

[0117] First, the experimental scenario is pre-defined, specifying the detection angle range and road range. Then, the recognition model is pre-trained to enable it to detect and warn of targets when strong target points appear in the distance Doppler image. In this embodiment, the scenario is as follows: Figure 2 As shown, the road is curved and obstructed by the wall, creating a blind spot and making it impossible to perceive the target's movement trajectory.

[0118] Secondly, based on the established scenario, a set of control experiments without metasurfaces were conducted, and the experimental results are as follows: Figure 5 As shown, the experimental results indicate that the detection radar used can only achieve dynamic perception and detection of targets at line of sight. The measured data in this scenario verify that the detection radar cannot achieve continuous and uninterrupted perception and tracking of targets at non-line of sight.

[0119] Finally, a set of non-line-of-sight target perception experiments with intelligent metasurface devices were conducted based on the established scenario to verify the non-line-of-sight target detection performance of the proposed method. (Example: Corner scene) Figure 2As shown, due to the obstruction of the wall, the detection radar and the target's movement range form a non-line-of-sight range. Intelligent metasurface beam manipulation was used to detect non-line-of-sight targets, thus filling the blind spots in the non-line-of-sight area. Experimental results are as follows... Figure 6 As shown in the figure. Experimental results demonstrate that the non-line-of-sight target detection method proposed in this invention can achieve dynamic perception and detection of non-line-of-sight targets. The measured data in this scenario verify that continuous and uninterrupted perception and tracking of non-line-of-sight targets can be achieved by introducing a smart metasurface. The results clearly show the position and velocity of a single vehicle target, which is consistent with the actual situation.

[0120] The above description is merely an example of a specific solution of the present invention. For any devices and structures not described in detail herein, it should be understood that they are implemented using common devices and methods already available in the art.

[0121] The above description is merely one embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting non-line-of-sight targets based on intelligent metasurface on curved road, characterized in that, Includes the following steps: S1. An intelligent metasurface device for relay detection of non-line-of-sight targets is set up for non-line-of-sight range. S2. Illuminating the intelligent metasurface device with a detection radar, and modulating the beam emitted by the detection radar based on the intelligent metasurface device to change the propagation path of the beam and illuminate a non-line-of-sight target; wherein the intelligent metasurface device modulates the beam with the strongest echo signal as the target; wherein the strength of the echo signal is determined based on a constructed optimization problem model. The optimization problem model includes: the unit incident electric field strength of the intelligent metasurface device and the radar receiving power of the detection radar; The incident electric field intensity of the unit cell is constructed based on the narrow-band assumption and is expressed as follows: wherein represents the unit incident electric field strength; j Represents the imaginary unit; Z 0 represents the characteristic impedance of the medium; This represents the phase term related to the propagation path between the detection radar and the smart metasurface device, as well as the phase response of the radar transmitting antenna; The spatial power density of plane waves at the intelligent metasurface device is expressed as: in, Indicates the peak power of the detection radar. Indicates the gain of the radar transmitting antenna. Indicates the direction of radar observation. Normalized radar radiated power pattern, This indicates the distance between the detection radar and the intelligent metasurface device. Indicates the location of the detection radar. Indicates the location of the intelligent metasurface device; S 1 represents the steering matrix of the intelligent metasurface device related to the direction of incoming waves, expressed as: in, , Denotes the manifold vectors of the vertical and horizontal intelligent metasurface devices, where, , They represent the direction cosines, λ 0 indicates the radar operating wavelength. Indicates the orientation of the intelligent metasurface device. This indicates the relative orientation of the detection radar and the intelligent metasurface device. Indicates the row number of the intelligent metasurface device. Indicates the column number of the intelligent metasurface device; The radar receiving power of the detection radar is constructed based on reciprocity conditions, where, assuming the intelligent metasurface device has zero loss and maximum beam pointing gain, the radar receiving power under ideal conditions is expressed as: in, P rx This indicates the radar receiving power of the detection radar; G Indicates antenna power gain; F tot To merge items of the same type, it is represented as: in, Indicates the relative orientation of the detection radar and the smart metasurface device. The normalized unit radiation power pattern of the intelligent metasurface device. Indicates the relative direction between the non-line-of-sight target and the intelligent metasurface device. The unit radiation power pattern of the normalized intelligent metasurface device; η RIS Indicates unit efficiency; r 2 represents the distance between the radar and a non-line-of-sight target, expressed as: ,in, S T Indicates the position of the phase center of a non-line-of-sight target; σ Reflectivity of the target power along the line-of-sight direction for the intelligent metasurface device; S3. The detection radar receives the echo signal transmitted back by the intelligent metasurface device, and performs non-line-of-sight target detection and early warning based on the echo signal.

2. The non-line-of-sight target detection method according to claim 1, characterized in that, In step S3, the step of the detection radar receiving the echo signal transmitted back by the intelligent metasurface device and performing non-line-of-sight target detection and early warning based on the echo signal includes: S31. The detection radar receives the echo signal transmitted back by the intelligent metasurface device, preprocesses the echo signal, and obtains multidimensional information of the non-line-of-sight target; wherein, the multidimensional information includes: distance information, velocity information, and angle information; S32. Generate a range Doppler image based on the preprocessed echo signal; S33. Input the distance Doppler image into a pre-trained recognition model, and perform non-line-of-sight target detection and early warning based on the recognition model.

3. The non-line-of-sight target detection method according to claim 2, characterized in that, In step S31, the step of the detection radar receiving the echo signal transmitted back by the intelligent metasurface device and preprocessing the echo signal includes clutter suppression of the echo signal, which comprises: S311. Perform radio frequency interference suppression on the received echo signal and generate a first processing signal; S312. Perform moving target detection on the first processed signal, suppress static clutter in the first processed signal, and generate a second processed signal; S313. Process the second processing signal to obtain the non-line-of-sight target.

4. The non-line-of-sight target detection method according to claim 3, characterized in that, Step S313, the step of processing the second processing signal to obtain the non-line-of-sight target, includes: S3131. Perform parallel processing on the second processed signal, wherein the second processed signal is processed by the first branch to obtain the first target information, and the clutter map is established by the second branch and the second target information is obtained by using dynamic clutter suppression. S3132. Summarize the first target information and the second target information to describe the non-line-of-sight target.

5. The non-line-of-sight target detection method according to claim 4, characterized in that, In step S3131, the step of processing the second processing signal using the first branch to obtain the first target information includes: The second processed signal is incoherently accumulated; The second processed signal after incoherent accumulation is detected using CFAR to obtain the first target information.

6. The non-line-of-sight target detection method according to claim 5, characterized in that, In step S3131, the step of establishing a clutter map using the second branch and obtaining the second target information using dynamic clutter suppression, wherein the second target information includes external clutter information and internal clutter information, includes: S3131a. Dynamic clutter suppression is performed on the clutter outside the clutter map to obtain the clutter outside information; S3131b. Compare the extraneous information with a preset threshold. If the extraneous information meets the preset threshold, determine that the extraneous information is target information and merge the extraneous information with the first target information. S3131c. Target information is filtered for the corresponding unit amplitude of the clutter in the clutter map using a guard window to obtain the clutter information; wherein the target information included in the clutter information is target information that is slow and small in velocity and scale and / or easily confused with clutter, which is smaller than the first target information.

7. The non-line-of-sight target detection method according to claim 6, characterized in that, In step S3131b, in the step of comparing the clutter information with a preset threshold, if the clutter information does not meet the preset threshold, then the clutter information is determined to be non-target information and the clutter information is processed together with the clutter information in step S3131c to filter out the target information that is slow, small and / or easily confused with clutter.

8. The non-line-of-sight target detection method according to claim 7, characterized in that, In step S31, the step of obtaining multidimensional information of the non-line-of-sight target, the distance information and the velocity information are obtained based on the preprocessed echo signal, and the angle information is obtained based on the intelligent metasurface device.

Images