Underwater sound source target detection, identification and positioning method and device based on millimeter wave radar

Abstract

The invention discloses an underwater sound source target detection, identification and positioning method and device based on a millimeter wave radar, and belongs to the technical field of underwater target detection, identification and positioning. The method comprises the following steps: acquiring a water surface echo signal by using a millimeter wave radar, and performing Fourier transform to obtain a distance-time spectrum; extracting a slow time phase sequence of each range gate in a radar beam coverage range, and obtaining a sound-induced water surface vibration signal after unwinding and filtering; performing correlation calculation on the vibration signal and a preset target feature library, or realizing target recognition through a deep learning model; a multi-radar cooperation or single-radar digital beam forming technology is adopted for target positioning, and large-range water area searching is achieved. According to the invention, the millimeter-wave radar is used for detecting micron-sized water surface vibration with high phase sensitivity, the defects of short distance, complex arrangement and the like of traditional acoustic and optical detection are overcome, and the system has large-range, real-time and non-contact underwater target detection and positioning capability.

Description

A Method and Device for Underwater Sound Source Target Detection, Identification and Localization Based on Millimeter-Wave Radar Technical Field

[0001] This invention belongs to the field of underwater target detection, identification and positioning technology, specifically relating to a method and device for underwater sound source target detection, identification and positioning based on millimeter-wave radar. Background Technology

[0002] Underwater target detection has important applications in fields such as marine environmental perception and resource exploration. Current mainstream detection technologies mainly include optical detection, acoustic detection, laser detection, and the emerging millimeter-wave radar detection.

[0003] Optical detection relies on underwater imaging for target identification, but light waves attenuate drastically in water, resulting in extremely limited effective detection range. Acoustic detection primarily utilizes the excellent propagation characteristics of sound waves in water, and can be divided into active sonar and passive sonar. Active sonar emits sound waves and receives echoes, but its equipment is typically bulky and consumes a lot of power; passive sonar only receives target radiated noise without actively emitting signals, offering better concealment, but usually requires pre-deployed fixed arrays, has weak low-frequency sensing capabilities, limited detection range, and insufficient mobility. Laser detection technology uses high-precision lasers to detect water surface fluctuations caused by sound sources to invert underwater targets, offering high accuracy, but its effective range is small, making it difficult to meet the needs of large-scale searches. In recent years, millimeter-wave radar detection technology has developed rapidly. Its principle is to indirectly sense underwater targets by detecting micro-vibrations on the water surface caused by underwater sound sources. Existing technologies, such as Chinese patent (application number CN202111295294.1), disclose an underwater target detection method based on microwave sensing of water surface vibrations. This method uses a platform equipped with a microwave transceiver to collect signals and eliminate interference from waves, thereby locating the target water area. However, this technology mainly focuses on the detection of designated locations or cross-media communication research, and lacks a systematic solution for large-area rapid search, positioning and identification in the context of vast waters.

[0004] In summary, existing technologies all have limitations in underwater target detection: they may have short detection ranges, inflexible deployment, or small coverage areas, making it difficult to effectively meet the urgent need for large-scale, real-time, and mobile search and identification of underwater acoustic targets (such as submarines and marine life). Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and apparatus for detecting, identifying, and locating underwater sound sources based on millimeter-wave radar. It utilizes the high phase sensitivity of millimeter-wave radar to detect micron-level water surface vibrations, overcoming the shortcomings of traditional acoustic and optical detection methods, such as short detection range and complex deployment. It possesses the capability for wide-range, real-time, and non-contact underwater target detection and location.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for detecting, identifying, and locating underwater acoustic sources based on millimeter-wave radar, the method comprising:

[0008] Step 1: Use millimeter-wave radar to detect acoustic surface ripples caused by underwater sound source targets and obtain the radar echo signals reflected from them.

[0009] Step 2: Perform a Fourier transform on the radar echo signal to obtain the range gate information corresponding to the water surface area covered by the radar beam;

[0010] Step 3: Extract the slow-time phase sequence of multiple consecutive radar echo signals with the same range gate, and obtain the characteristics of the acoustically induced water surface wave signal after phase dewinding and filtering.

[0011] Step 4: Compare the features with a preset target feature library to identify the type of underwater sound source target;

[0012] Step 5: Based on the aforementioned characteristics, determine the location of the underwater sound source target using multi-radar cooperative positioning or single-radar digital beamforming technology.

[0013] Furthermore, step 2 specifically includes:

[0014] Perform a fast time-dimensional Fourier transform on the radar echo signal to generate a time-range spectrum;

[0015] Based on the radar platform's altitude and beam elevation angle, the actual range of distances where the radar beam illuminates the water surface is calculated, thereby determining one or more corresponding range gates.

[0016] Furthermore, step 3 specifically includes:

[0017] The extracted slow-time phase sequence is processed using a one-dimensional phase unwinding algorithm to eliminate phase jumps caused by the water surface vibration amplitude exceeding the principal phase value range.

[0018] The unwound phase sequence is bandpass filtered to remove low-frequency and DC components introduced by natural water surface ripples and platform motion, as well as high-frequency components caused by system noise.

[0019] Furthermore, step 4 specifically includes:

[0020] Time-frequency analysis was performed on the acoustically induced water surface wave signal to extract its time-domain and frequency-domain features;

[0021] The extracted feature vectors are input into a pre-trained target recognition model, which is a classifier trained by deep learning based on a target feature library containing various known underwater sound source target features.

[0022] The target recognition model outputs the type identification result of the underwater sound source target.

[0023] Furthermore, in step 5, the multi-radar cooperative positioning is achieved in the following way:

[0024] At least three spatially non-collinear millimeter-wave radars are deployed around the same water area to be measured, and time synchronization is achieved through satellite timing.

[0025] By employing the time difference of arrival or time of arrival method, based on the time information of the same vibration signal arriving at different radars, the two-dimensional position coordinates of the underwater sound source target on the water surface are determined through geometric calculation.

[0026] Furthermore, in step 5, the single-radar digital beamforming technology is implemented in the following way:

[0027] Employ millimeter-wave radar with a wide beam to cover vast water areas;

[0028] At the receiving end, digital beamforming technology is used to simultaneously generate multiple narrow receiving beams pointing to different directions;

[0029] When a target signal is detected at a certain range gate, the azimuth of the target is determined by comparing the phase or amplitude information of the signal in different receiving beams, using the phase method for angle measurement or the amplitude comparison method, and then combining the range information to complete the positioning.

[0030] Furthermore, the method also includes:

[0031] Step 6: Feed back the identification and positioning results to the fine-tuning detection equipment to guide further exploration of the target area; and add the target information confirmed by the fine-tuning detection and its corresponding acoustic water surface wave signal features as new samples to the target feature library for updating and optimizing the target identification model.

[0032] On the other hand, the present invention provides an underwater acoustic source target detection, identification, and positioning device based on millimeter-wave radar, comprising:

[0033] The echo acquisition module is used to detect acoustic water surface fluctuations caused by underwater sound source targets using millimeter-wave radar and acquire the radar echo signals reflected by them.

[0034] The distance calculation module is used to perform Fourier transform on the radar echo signal to obtain the distance gate information corresponding to the water surface area covered by the radar beam;

[0035] The feature acquisition module is used to extract the slow-time phase sequence of multiple consecutive radar echo signals with the same range gate, and after phase dewinding and filtering, obtain the features of the acoustically induced water surface wave signal.

[0036] The type recognition module is used to compare the features with a preset target feature library to identify the type of underwater sound source target;

[0037] The location identification module is used to determine the location of the underwater sound source target based on the features, through multi-radar cooperative positioning or single-radar digital beamforming technology.

[0038] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar.

[0039] Fourthly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned method for detecting, identifying, and locating underwater sound sources based on millimeter-wave radar.

[0040] The beneficial effects of this invention are as follows:

[0041] This invention enables wide-area, real-time, and mobile underwater target detection and monitoring. Utilizing a large-beam-angle millimeter-wave radar combined with mobile platforms such as airborne systems, it can cover vast water areas. Through real-time echo phase processing via a sliding window, it achieves continuous, real-time monitoring and rapid target extraction over a large area of ​​water.

[0042] This improves the intelligence and efficiency of target identification and localization. By establishing and utilizing an underwater sound source target feature database and constructing a deep learning recognition model based on it, the extracted vibration signal features can be automatically compared and identified, simplifying the identification process and improving identification efficiency and accuracy. Simultaneously, the use of multi-radar cooperative localization (such as TDOA / TOA) or single-radar digital beamforming (DBF) technology enables precise localization of the detected target.

[0043] The system is flexible in deployment and highly adaptable. The radar system can be flexibly deployed in various ways, such as fixed installation and airborne, and multiple radars can be synchronized through satellite time synchronization. The solution has good adaptability and can meet the detection needs in different scenarios, overcoming the shortcomings of traditional sonar deployment, such as complexity and poor mobility.

[0044] The detection method is non-contact and highly covert. Utilizing millimeter-wave radar to remotely detect micro-vibrations on the water surface is a non-contact detection method, making it difficult for underwater targets to detect and providing excellent concealment. It is suitable for military and sensitive water area monitoring. Attached Figure Description

[0045] Figure 1 is a flowchart of the underwater acoustic source target detection, identification and localization method based on millimeter-wave radar according to the present invention.

[0046] Figure 2 is a schematic diagram of single radar signal processing and target extraction;

[0047] Figure 3 is a schematic diagram of multi-radar cooperative detection and geometric positioning;

[0048] Figure 4 shows the phase distribution of water surface vibration inside the gate at different distances under simulation conditions;

[0049] Figure 5 shows the phase-time curve and spectrum of water surface vibration caused by a 100Hz sound source within a certain distance gate. Detailed Implementation

[0050] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0051] As shown in Figure 1, this invention provides a method for detecting, identifying, and locating underwater sound sources based on millimeter-wave radar. Its core lies in utilizing the high phase sensitivity of millimeter-wave radar to detect micrometer-level water surface fluctuations caused by underwater sound sources. Then, through signal processing, feature matching, and geometric localization, a large-scale search, identification, and location of underwater targets can be achieved. The method includes:

[0052] Step 1: Use millimeter-wave radar to detect acoustic surface ripples caused by underwater sound source targets and obtain the radar echo signals reflected from them.

[0053] Step 2: Perform a Fourier transform on the radar echo signal to obtain the range gate information corresponding to the water surface area covered by the radar beam;

[0054] Step 3: Extract the slow-time phase sequence of multiple consecutive radar echo signals with the same range gate, and obtain the characteristics of the acoustically induced water surface wave signal after phase dewinding and filtering.

[0055] Step 4: Compare the features with a preset target feature library to identify the type of underwater sound source target;

[0056] Step 5: Based on the aforementioned characteristics, determine the location of the underwater sound source target using multi-radar cooperative positioning or single-radar digital beamforming technology.

[0057] In step 1:

[0058] Sound waves are essentially mechanical waves. The periodic pushing and pulling of an underwater sound source forces the water surface above the source to vibrate in accordance with the frequency of the sound waves. The amplitude of the water surface vibration depends on the sound intensity and the frequency of the sound waves. The amplitude of the water surface vibration caused by an underwater sound source is relatively small, typically on the order of micrometers, and traditional radar ranging methods cannot detect such weak vibration signals.

[0059] When radar detects a target, changes in the distance from the target to the radar will cause corresponding changes in the echo phase. Depending on the requirements of the detection mission, one of two main deployment modes can be selected:

[0060] Single-radar mobile platform mode: A millimeter-wave radar is mounted on a mobile platform such as an unmanned aerial vehicle (UAV). The radar antenna employs a large beam angle design (e.g., both azimuth and elevation beamwidths are greater than 60°) to ensure sufficient coverage of the water area below. The radar operates in frequency-modulated continuous wave mode.

[0061] Multi-radar network collaborative mode: As shown in Figure 3, at least three millimeter-wave radars are deployed around or above the same water area to be measured. The three radars are not collinear in space and achieve high-precision time synchronization (synchronization accuracy needs to reach the nanosecond to microsecond level) through their built-in satellite (such as GPS / BeiDou) timing modules. The radars can be fixed on the shore, buoys, or carried by airborne platforms.

[0062] By analyzing the echo phase, the distance from the target to the radar can be deduced; this is called phase ranging. Phase ranging has high accuracy and can effectively extract the subtle movement characteristics of the target, making it useful for extracting surface vibrations caused by underwater sound sources. The following equation illustrates the principle of phase ranging:

[0063] ,

[0064] The radar echo phase is denoted by d, and d is the distance from the water surface to the radar. The wavelength is the center frequency of the radar radio frequency signal.

[0065] When the distance changes slightly, the phase changes significantly, and the magnitude of this change is related to the signal wavelength. At excessively high radar frequencies, even small water surface ripples can cause excessive phase changes, leading to phase wrapping. Acoustic water surface ripples are on the micrometer scale, and the radar used in this invention is limited to millimeter-wave radar. (For example, when the target amplitude is 40µm and the radar RF signal center frequency wavelength is 2mm, the phase change amplitude is 14.4°.)

[0066] Millimeter-wave radar using a frequency-modulated continuous wave (FMCV) system detects targets by transmitting a signal of a specific waveform. Upon encountering a target, the radar reflects an echo signal with the same waveform. To reduce sampling complexity, the received signal undergoes dechirping. The echo signal is then mixed with a delayed copy of the transmitted signal to obtain an intermediate frequency (IF) signal whose frequency is proportional to the target distance. This IF signal is then subjected to analog-to-digital (AD) sampling to obtain digitized IF sampled data. This data forms the original data matrix in both the fast time dimension (sampling sequence of a single pulse) and the slow time dimension (successive pulse cycles).

[0067] In step 2:

[0068] The total echo data contains echo data from multiple periods. A Fourier transform is performed on the original data matrix along the fast time dimension to form a time-range map. As shown in the upper half of Figure 2, each column of data represents a set of echo data. A Fourier transform is performed on each set of echo data to display the intermediate frequency (IF) signal in the spectrum, distinguishing the echo energy at different distances, corresponding to each column of data in the lower half of Figure 2. Based on this, a coordinate transformation can be performed using the IF frequency and the actual distance, with each frequency point corresponding to a range gate. The fast time-domain FFT results of the n echo data are then superimposed in the time domain to obtain the time-range spectrum. Based on the known altitude and beam direction of the radar platform, the actual range of distances at which the radar beam illuminates the water surface can be calculated. Within this range, there are one or more range gates; the signals within these range gates contain information about all reflection points on the water surface, and potential targets are hidden within them.

[0069] In step 3:

[0070] Find each range gate corresponding to the water surface covered by the radar beam, extract the signal phase information of the same range gate corresponding to the n adjacent sets of echo data, and process the continuous echo in real time using the n sets of echo signals as a sliding window.

[0071] The phase of a target is generally limited to between -π and π. When the phase change caused by water surface ripples exceeds this range, the true phase change will be added to or subtracted by an integer multiple of 2π, which means that phase entanglement occurs. The target phase is unwound to restore the continuous, non-jumping true phase change curve.

[0072] Acoustic water surface ripples are often superimposed on natural water ripples. Bandpass filtering is applied to the extracted target distance-gated time-domain phase to remove DC components, low-frequency natural ripple frequencies, and high-frequency components introduced by system noise. Detection is then performed on the filtered phase spectrum, identifying echoes with specific frequency information as potential target signals. Corresponding phase time-domain and frequency characteristics are extracted, such as signal dominant frequency, bandwidth, harmonic structure, time-domain envelope shape, and formants.

[0073] In step 4:

[0074] Underwater acoustic targets, such as submarine communications, propellers, torpedo movements, and marine life, emit sounds with varying pitches, timbres, amplitudes, envelopes, durations, phases, bandwidths, and formants. A preliminary target feature database can be established based on existing data. This database can be updated after actual verification of target detection results. A feature recognition model can then be built using deep learning based on this database.

[0075] The extracted phase time-domain features and frequency features are compared with the target feature library. For example, if the feature information is input into a trained feature recognition model, the target classification and recognition results (e.g., "AIP submarine low-speed navigation noise") and the corresponding confidence level are automatically output.

[0076] In step 5, the corresponding positioning method is adopted according to the system deployment mode:

[0077] Method 1: Employing a wide-range radar. This radar can be airborne, transmitting a wide beam to cover a broad water area. Digital beamforming (DBF) technology is used for reception, simultaneously generating multiple beams to divide the water area into a grid. For areas where a target is detected, phase-based angle measurement is used for positioning. For example, when a target is detected at a certain range threshold, by comparing the phase or amplitude of the target signal in different receiving beams, and using interferometry principles or amplitude comparison, the precise azimuth angle of the target on that range loop can be estimated, achieving two-dimensional range-azimuth positioning.

[0078] Method 2: Millimeter-wave radar antennas employ a large beam angle to cover a wide area of ​​water. At least three radars are used to detect the same water area. The three radars cannot be aligned in a straight line, and they are synchronized via satellite time synchronization. The radars can be fixedly mounted or deployed via airborne methods. One radar shares the transmission and reception, while the other two radars only receive signals of the same frequency (in this case, the range gate signal corresponding to the direct arrival wave of the radar must be filtered out). The Time Difference of Arrival (TDOA) algorithm is used, where each radar accurately measures the arrival time of the same vibration event signal and calculates the time difference between the slave and master stations. Each time difference corresponds to a hyperbola with two radars as foci. The intersection of the hyperbolas between any two of the three radars is the target location.

[0079] Method 3: Millimeter-wave radar antennas employ a large beam angle to cover a wide area of ​​water. At least three radars are used to detect the same water area. The three radars cannot be aligned in a straight line, and they are synchronized via satellite time synchronization. The radars can be fixedly mounted or deployed via airborne means. The three radars operate at different frequencies and all have transceiver capabilities. The Time of Arrival (TOA) method is used, which involves calculating the absolute slant range from the target to each radar. A sphere is constructed with each radar as its center and the corresponding slant range as its radius. The intersection of this sphere with the horizontal plane is a circle, and the intersection of the three circles represents the target's position on the water surface. Solving this geometric problem allows for accurate positioning.

[0080] Furthermore, the identification and positioning results are fused to form complete situational information including target type, location, and timestamp. Detailed exploration is then conducted using high-precision detection equipment such as sonar or lasers, and the detection results are fed back to the target feature database to update the database, enabling the system to continuously learn and improve its performance.

[0081] Figure 4 shows the processing results of the acoustically induced water surface wave phase signal within different distance gates under simulation conditions. The horizontal axis represents slow time (i.e., continuous pulse period), the vertical axis represents the distance gate number, and the chromaticity represents the phase change amplitude. By comparing the phase change patterns of the target distance gate and the non-target distance gate, the characteristic vibration region caused by the underwater sound source can be intuitively identified, verifying the effectiveness of extracting micro-vibration signals using phase information.

[0082] Figure 5 shows the range-gate phase extraction result corresponding to 100Hz acoustic surface ripples. It presents a range-gate phase sequence extracted from radar echoes and its spectral analysis results for simulated water surface vibrations caused by a 100Hz sound source. The upper figure shows the curve of phase change over time after unwinding and filtering, exhibiting periodic fluctuations consistent with the sound source frequency. The lower figure shows the corresponding spectrum, with a significant peak at 100Hz, further confirming that this method can effectively separate and identify underwater acoustic vibration signals of specific frequencies, providing a reliable time-frequency basis for feature matching and target recognition.

[0083] On the other hand, the present invention provides an underwater acoustic source target detection, identification, and positioning device based on millimeter-wave radar, the various modules of which can implement the various steps of the aforementioned method, specifically including:

[0084] The echo acquisition module is used to detect acoustic water surface fluctuations caused by underwater sound source targets using millimeter-wave radar and acquire the radar echo signals reflected by them.

[0085] The distance calculation module is used to perform Fourier transform on the radar echo signal to obtain the distance gate information corresponding to the water surface area covered by the radar beam;

[0086] The feature acquisition module is used to extract the slow-time phase sequence of multiple consecutive radar echo signals with the same range gate, and after phase dewinding and filtering, obtain the features of the acoustically induced water surface wave signal.

[0087] The type recognition module is used to compare the features with a preset target feature library to identify the type of underwater sound source target;

[0088] The location identification module is used to determine the location of the underwater sound source target based on the features, through multi-radar cooperative positioning or single-radar digital beamforming technology.

[0089] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar.

[0090] Fourthly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned method for detecting, identifying, and locating underwater sound sources based on millimeter-wave radar.

[0091] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting, identifying, and locating underwater acoustic sources based on millimeter-wave radar, characterized in that, The method includes: Step 1, using millimeter-wave radar to detect acoustic surface fluctuations caused by underwater sound source targets and acquiring the reflected radar echo signals; Step 2, performing Fourier transform on the radar echo signals to obtain range gate information corresponding to the water surface area covered by the radar beam; Step 3, extracting the slow-time phase sequence of multiple consecutive sets of radar echo signals with the same range gate, and obtaining the characteristics of the acoustic surface fluctuation signal after phase decoupling and filtering; Step 4, comparing the characteristics with a preset target feature library to identify the type of underwater sound source target; Step 5, determining the location of the underwater sound source target based on the characteristics using multi-radar cooperative positioning or single-radar digital beamforming technology.

2. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, Step 2 specifically includes: performing a fast time-dimensional Fourier transform on the radar echo signal to generate a time-range spectrum; calculating the actual range of the radar beam illuminating the water surface based on the radar platform's height and beam elevation angle, thereby determining one or more corresponding range gates.

3. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, Step 3 specifically includes: processing the extracted slow-time phase sequence using a one-dimensional phase unwinding algorithm to eliminate phase jumps caused by the water surface vibration amplitude exceeding the principal phase value range; and performing bandpass filtering on the unwinded phase sequence to filter out low-frequency and DC components introduced by natural water surface ripples and platform motion, as well as high-frequency components caused by system noise.

4. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, Step 4 specifically includes: performing time-frequency analysis on the acoustically induced water surface wave signal to extract its time-domain and frequency-domain features; inputting the extracted feature vector into a pre-trained target recognition model, which is a classifier trained through deep learning based on a target feature library containing various known underwater sound source target features; and outputting the type recognition result of the underwater sound source target from the target recognition model.

5. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, In step 5, the multi-radar cooperative positioning is achieved in the following way: at least three spatially non-collinear millimeter-wave radars are deployed around the same water area to be measured, and time synchronization is achieved through satellite time synchronization; the two-dimensional position coordinates of the underwater sound source target on the water surface are determined by geometric calculation based on the time information of the same vibration signal arriving at different radars using the time difference of arrival or time of arrival method.

6. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, In step 5, the single-radar digital beamforming technology is implemented in the following way: a millimeter-wave radar with a wide beam is used to cover a wide area of ​​water; at the receiving end, digital beamforming technology is used to generate multiple narrow receiving beams pointing to different directions simultaneously; when a target signal is detected at a certain range threshold, the azimuth angle of the target is determined by comparing the phase or amplitude information of the signal in different receiving beams, using the phase method for angle measurement or the amplitude comparison method, and the positioning is completed by combining the range information.

7. The underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar according to claim 1, characterized in that, The method further includes: step 6, feeding back the identification and positioning results to the fine-tuning detection device to guide further exploration of the target area; and adding the target information confirmed by the fine-tuning detection and its corresponding acoustic water surface wave signal features as new samples to the target feature library for updating and optimizing the target identification model.

8. An underwater acoustic source target detection, identification, and positioning device based on millimeter-wave radar, characterized in that, include: The echo acquisition module is used to detect acoustic water surface fluctuations caused by underwater sound source targets using millimeter-wave radar and acquire the radar echo signals reflected by them. The distance calculation module is used to perform Fourier transform on the radar echo signal to obtain the distance gate information corresponding to the water surface area covered by the radar beam; The feature acquisition module is used to extract the slow-time phase sequence of multiple consecutive radar echo signals at the same range gate, and after phase dewinding and filtering, obtain the features of the acoustically induced water surface wave signal. The type recognition module is used to compare the features with a preset target feature library to identify the type of underwater sound source target; The location identification module is used to determine the location of the underwater sound source target based on the features, through multi-radar cooperative positioning or single-radar digital beamforming technology.

9. An electronic device, characterized in that, include: One or more processors; A memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the underwater acoustic source target detection, identification and localization method based on millimeter-wave radar as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, enable the processor to implement the underwater acoustic source target detection, identification, and positioning method based on millimeter-wave radar as described in any one of claims 1-7.

Images