Sea area underwater UXO positioning method and system and underwater UXO treatment process

Abstract

The invention relates to the technical field of ocean engineering investigation, and discloses a sea area underwater UXO positioning method and system and an underwater UXO treatment process, and the method comprises the following steps: firstly, collecting underwater magnetic field intensity change data by using an ocean magnetic detector; acquiring a seabed image for verifying a magnetic field abnormal region; then an earthquake reflection technology is adopted; carrying out integration through a self-adaptive data fusion algorithm; utilizing a machine learning model to train historical UXO case data; and finally, dynamically adjusting a detection strategy and a path through a reinforcement learning technology by combining virtual reality and augmented reality technologies, and finally completing accurate positioning of the UXO. By adopting a multi-sensor information fusion technology, multi-modal data are fused and processed in real time, the position, shape and burial depth information of a UXO target can be accurately identified, efficient detection and accurate positioning in a complex seabed environment are realized, and stable performance of the system in the complex environment is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of marine engineering investigation, and particularly to a method for positioning underwater UXO in sea areas, a system and an underwater UXO treatment process. Background Art

[0002] Marine engineering investigation, especially the sea area survey before the construction of offshore wind farms, usually requires the use of marine geological survey technologies and equipment to detect and mark unexploded ordnance (UXO) in the sea area to ensure construction safety. Currently, marine magnetic detectors play an important role in detecting underground metal objects and have become an indispensable part of sea area investigation.

[0003] Existing detection methods are limited by equipment performance and data analysis accuracy, and often cannot accurately determine the precise position and burial depth of UXO. Especially in the case of deep sediment coverage on the seabed, it is difficult to effectively identify, resulting in low detection and removal efficiency and high risks. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a method for positioning underwater UXO in sea areas, a system and an underwater UXO treatment process, which solve the problems in the prior art that when a single sensor system detects UXO in a complex seabed environment, it is easily interfered by factors such as terrain undulation, sediment coverage and environmental noise, resulting in inaccurate target recognition, low positioning accuracy, inability to effectively obtain the complete spatial information of UXO, and restricting the safety and efficiency of explosive disposal operations.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for positioning underwater UXO in sea areas, comprising the following steps: First, use a marine magnetic detector to collect data on changes in underwater magnetic field intensity, and based on this magnetic field data, construct a magnetic anomaly vector field, and identify the abnormal central region through a magnetic field reconstruction algorithm; Then, use a sidescan sonar to perform imaging scans on the seabed to obtain seabed images for verifying the magnetic anomaly region and assisting in confirming whether it is a potential target of UXO. Combine deep learning image segmentation technology to analyze the shape, texture and edges of potential UXO targets in the sonar images to enhance the verification effect of the magnetic anomaly region; Then, adopt seismic reflection technology to obtain information on rock interfaces at different depths of the seabed through acoustic wave reflection, and based on the echo delay and interlayer structure inversion algorithm, estimate the vertical burial depth and cover thickness of UXO in the sediment layer; Integrate marine magnetometers, side-scan sonars, and seismic reflection data through an adaptive data fusion algorithm to construct a multi-source perception-driven UXO probability information field, and use this information field as a basis to guide subsequent detection strategies. The integration is used to generate a multi-dimensional seabed detection map and optimize the estimation of the potential location and depth of UXOs. Use a machine learning model to train historical UXO case data, predict the probability of the existence of UXOs and their burial depths, and classify the confidence of abnormal targets through a model uncertainty estimation mechanism to improve the false alarm filtering ability. On the basis of constructing the information field, introduce an information entropy weight distribution, and automatically select the detection priority area through the maximum information gain criterion for detection efficiency and path adjustment. Combine virtual reality and augmented reality technologies to perform underwater three-dimensional space reconstruction, display the position and depth of UXOs in real time, and feedback the detection decision results and the confidence evaluation of the model output through a visual interaction interface. Finally, through reinforcement learning technology, dynamically adjust the detection strategy and path, and automatically make dynamic adjustments according to real-time detection data and environmental changes to finally complete the precise positioning of UXOs.

[0006] Preferably, the magnetic field data collected by the marine magnetometer is preprocessed, including noise removal, filtering, and abnormal feature extraction, to remove environmental noise interference and enhance the magnetic field anomaly signal generated by UXOs. During the side-scan sonar imaging process, multi-band sonar data is used to simultaneously obtain seabed images with different resolutions to form multi-scale image information, thereby improving the recognition accuracy of UXOs. The magnetic anomaly vector field is constructed from magnetic field data by using a magnetic field reconstruction algorithm. The magnetic anomaly vector field generated by the magnetic field reconstruction algorithm constructs a dynamic UXO probability information field according to different abnormal intensity distributions for determining the target position and depth estimation, and automatically generates a detection path according to this information field. The generation of the magnetic anomaly vector field is carried out through the following formula: ; where is the total measured magnetic field, is the background magnetic field, is the magnetic anomaly vector field; The dynamic UXO probability information field is constructed through the following formula according to the relationship between the magnetic field anomaly intensity and the position: ; where is the probability of the existence of a UXO at position , is a constant that controls the relationship between the magnetic field anomaly intensity and the UXO probability.

[0007] Preferably, the seismic reflection technology uses a multi-channel seismic data acquisition device to generate seismic reflection profiles at different depths by controlling the acoustic wave frequency and incident angle, so as to accurately estimate the burial depth of UXO. The data fusion uses a multi-modal data fusion algorithm based on Bayesian inference or Kalman filtering to jointly process the data of marine magnetic detectors, side-scan sonars and seismic reflections to form a three-dimensional spatial distribution model; When the deep learning image segmentation technology classifies and identifies sonar images, it uses a combination of convolutional neural networks and generative adversarial networks to optimize image recognition from multiple angles and continuously improve the accuracy of target location judgment through model iteration.

[0008] Preferably, the three-dimensional space reconstruction uses a point cloud data reconstruction algorithm or a voxel reconstruction algorithm to generate a three-dimensional model of the seabed environment in combination with sonar images and seismic profile data; It further includes anomaly detection. By using an anomaly detection algorithm to monitor real-time data, abnormal signals are identified and quickly resampled to improve the accuracy of UXO detection; The seismic reflection information further corrects the burial depth of UXO by combining the time delay between reflection layers and the wave velocity model with the seabed sediment type, and provides target interlayer density contrast data to optimize the estimation of UXO depth. The reflection depth is estimated by the following formula: ; where v is the wave velocity, Δt is the time delay of acoustic wave reflection, and d is the depth of the reflection point; the rock interface information is used to help estimate the burial depth of UXO and supplement the insufficient depth in the data. The three-dimensional reconstruction uses a point cloud data reconstruction algorithm or a voxel reconstruction algorithm to generate a three-dimensional model of the seabed environment in combination with sonar images and seismic profile data.

[0009] A marine underwater UXO positioning system includes: A data acquisition module for obtaining the original data of the seabed environment and UXO targets from marine magnetic detectors, side-scan sonars and seismic reflection technologies, including magnetic field data, sonar images, and seismic reflection data; A data preprocessing module, connected to the data acquisition module, for cleaning, filtering and enhancing the acquired data to make the signal clearer; A data fusion module, connected to the data preprocessing module, for integrating multi-modal data collected by different sensors to form a complete seabed environment information; A machine learning analysis module, connected to the data fusion module, for identifying and classifying UXO targets based on historical UXO data and currently collected data through a machine learning model; 3D reconstruction and visualization module, used for 3D reconstruction and visual presentation of the spatial position of UXO and the seabed environment; Intelligent decision-making and optimization module, used for dynamically adjusting the detection strategy, optimizing the detection path and sensor parameters, and improving the positioning efficiency of UXO; System feedback and adjustment module, used for dynamically adjusting system parameters; Data storage and management module, connected to the system feedback and adjustment module, used for storing and managing all raw data, processed data and analysis results.

[0010] An underwater UXO treatment process in sea areas includes the following steps: S1. Prepare the shaped charge, determine the position and burial depth data of UXO according to detection, and clean part of the overburden; S2. Position the positioning ship. According to the position of UXO determined by detection, use wide area differential positioning to move the reef blasting ship or a self-made square barge with a guiding frame to the charge placement position, use 4 anchors or 6 anchors for positioning, and process and assemble the shaped charge on the positioning ship; After the charge placement is completed, move the ship to a safe distance, charge and detonate. Under the high-temperature jet generated by the explosion of the shaped charge, the shell of UXO is penetrated, causing the internal historical explosive to deflagrate, the pressure in the projectile body increases, the shell ruptures and decomposes, so that UXO undergoes low-level treatment.

[0011] Preferably, in S1, when the burial depth of UXO < 1m, direct charge placement and blasting treatment can be carried out. When the burial depth of UXO > 1m, use a large-scale sediment suction device to clear the surface sediment in an area with a diameter of 6m to 8m centered on the determined plane coordinate position. Measure the seabed elevation through a water weight, sounding rod, and multibeam sonar. After confirmation, make the thickness of the sediment overburden above the unexploded bomb < 1m, and blow and suck part of the surface overburden.

[0012] Preferably, in S1, the preparation of the shaped charge includes the following sub-steps: There are two types of shaped charges. One is a cylindrical shaped charge, and the other is a strip-shaped shaped charge. The shell of the shaped charge is welded with thin steel plates, the internal shaped charge liner is made of pressed steel plates, and the charge is filled with pressure-resistant and waterproof emulsion explosive. This explosive uses glass microspheres as the sensitizer of the emulsion explosive, and the sensitizer is used to improve the pressure resistance, detonation velocity and explosion power of the emulsion explosive; Considering that the density of the explosive is close to the density of seawater, there is a sealed air cavity at the bottom of the shaped charge, and the sealed air cavity is used to prevent the projectile from tipping and floating, and a counterweight is installed at the bottom of the shaped charge; The single charge amount of the cylindrical shaped charge is 28kg, and the charge amount of the strip-shaped shaped charge is 75.5kg. The main charge in the charge package uses pressure-resistant emulsion explosive, and the initiating detonator uses an enhanced digital electronic detonator.

[0013] Preferably, in S2, the assembled shaped charge is a cylindrical shaped charge and a strip-shaped shaped charge, which are determined according to the type and shape parameters of the unexploded ordnance obtained from the detection data of the marine magnetic scanner. A small-spacing grid should be used. When a strip-shaped shaped charge is used in the central area, the grid spacing can be appropriately increased, but it should not be greater than 1 m. In the area of 2-3 m centered on the positioning point, charge the small-spacing grid. According to the shape of the UXO, adjust the grid to confirm that the shell of the bomb body can be penetrated by the shaped charge jet.

[0014] Preferably, after step S3: Since the historical explosive combustion in the UXO bomb body damages the shell, changing the magnetic characteristics of the whole unexploded ordnance, re-detect it with a marine magnetic detector, compare it with the database to confirm the explosive disposal result. Under the condition that visibility permits, use an underwater camera robot to take pictures, analyze the explosive disposal result according to the image data. According to the result, it can be recognized that this process can be repeated, and the secondary blasting positioning grid is misaligned by half a position.

[0015] The present invention provides a method and system for positioning underwater UXO in sea areas and an underwater UXO treatment process. It has the following beneficial effects: 1. By adopting the multi-sensor information fusion technology, combining the marine magnetic detector, side-scan sonar and seismic reflection data, and real-time fusing and processing multi-modal data, the present invention can accurately identify the position, shape and burial depth information of UXO targets, overcome the limitations of traditional single-sensor technologies, achieve efficient detection and accurate positioning in complex seabed environments, greatly improve the detection accuracy of UXO, and ensure the stable performance of the system in complex environments.

[0016] 2. By applying an automated data analysis platform and intelligent recognition algorithms, quickly processing and analyzing multi-sensor data, automatically identifying and classifying potential UXO targets, the present invention reduces the need for manual intervention, realizes the processing of a large amount of data and target recognition in a short time, significantly reduces the UXO recognition time and human resource input, and improves the overall operation efficiency.

[0017] 3. By adopting an enhanced three-dimensional positioning method and combining multi-sensor information fusion technology, accurately positioning the UXO in space, updating the target position in real time, ensuring the accuracy of target data, realizing the accurate positioning of UXO in complex seabed environments, and greatly improving the safety and feasibility of subsequent processing and removal of UXO.

[0018] 4. The core technology of the present invention is to arrange shaped charges in small grid areas. The charging method is simple. It changes from the conventional underwater diving charging, which needs to be fixed on the shell of the bomb body, to the water surface dropping by using a guiding rod on a positioning ship. Through the form of charging in small grid areas, multiple charges are used to cover a certain area to offset the positioning error at sea, ensuring that the shell of the unexploded ordnance single body is penetrated by the shaped charge jet. Brief Description of the Drawings

[0019] Figure 1 It is a flowchart of a method for positioning underwater UXO in a sea area according to the present invention; Figure 2 It is a schematic diagram of the system architecture of a system for positioning underwater UXO in a sea area according to the present invention; Figure 3 It is a flowchart of a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 4 It is a schematic diagram of sucking and removing a partial covering layer on the surface of UXO in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 5 It is a partial plane schematic diagram of sucking and removing a partial covering layer on the surface of UXO in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 6 It is a schematic diagram of a cylindrical shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 7 It is a sectional schematic diagram of a strip shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 8 It is a schematic diagram of a strip shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 9 It is a positioning schematic diagram of a positioning ship in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 10 It is a cloth medicine grid schematic diagram of a cylindrical shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 11 It is a cloth medicine grid schematic diagram of a strip shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention; Figure 12 It is an operation schematic diagram of a positioning ship dropping a shaped charge cartridge in a processing technology for positioning underwater UXO in a sea area according to the present invention. Detailed Embodiments

[0020] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Please refer to the attached Figure 1 , an embodiment of the present invention provides a method for positioning underwater UXO in a sea area, which is characterized by including the following steps: First, use a marine magnetic detector to collect underwater magnetic field intensity change data, construct a magnetic anomaly vector field based on this magnetic field data, and identify the abnormal central area through a magnetic field reconstruction algorithm; Then, use a sidescan sonar to image and scan the seabed to obtain a seabed image, which is used to verify the magnetic anomaly area and assist in confirming whether it is a potential target of UXO. Combine deep learning image segmentation technology to analyze the shape, texture, and edges of potential UXO targets in the sonar image to enhance the verification effect of the magnetic anomaly area; Then, adopt seismic reflection technology to obtain rock interface information at different depths of the seabed through acoustic wave reflection, and estimate the vertical burial depth and cover thickness of UXO in the sediment layer based on the echo delay and interlayer structure inversion algorithm; Integrate the marine magnetic detector, sidescan sonar, and seismic reflection data through an adaptive data fusion algorithm to construct a multi-source perception-driven UXO probability information field, and use this information field as the basis to guide subsequent detection strategies. Integrate to generate a multi-dimensional seabed detection map and optimize the estimation of the potential location and depth of UXO; Use a machine learning model to train historical UXO case data, predict the presence probability and burial depth of UXO, and classify the confidence of abnormal targets through a model uncertainty estimation mechanism to improve the false alarm filtering ability; On the basis of constructing the information field, introduce an information entropy weight distribution, and automatically select the detection priority area through the maximum information gain criterion for detection efficiency and path adjustment; Combine virtual reality and augmented reality technologies to perform underwater three-dimensional space reconstruction, display the position and depth of UXO in real time, and feedback the detection decision results and confidence evaluation of the model output through a visual interaction interface; Finally, through reinforcement learning technology, dynamically adjust the detection strategy and path, and automatically make dynamic adjustments according to real-time detection data and environmental changes, and finally complete the precise positioning of UXO.

[0022] The magnetic field data collected by the marine magnetic detector is preprocessed, including noise removal, filtering, and abnormal feature extraction, to remove environmental noise interference and enhance the magnetic field anomaly signal generated by UXO. During the sidescan sonar imaging process, multi-band sonar data is used to obtain seabed images with different resolutions at the same time to form multi-scale image information, thereby improving the recognition accuracy of UXO; The magnetic anomaly vector field is constructed from magnetic field data by using a magnetic field reconstruction algorithm to generate a magnetic anomaly vector field. According to different abnormal intensity distributions, a dynamic UXO probability information field is constructed to determine the target position and depth estimation, and a detection path is automatically generated according to this information field; The generation of the magnetic anomaly vector field is carried out through the following formula: ; wherein, is the total measured magnetic field, is the background magnetic field, is the magnetic anomaly vector field; The dynamic UXO probability information field is constructed according to the relationship between the magnetic field anomaly intensity and position through the following formula: ; wherein, is the probability of the presence of UXO at position , is a constant that controls the relationship between the magnetic field anomaly intensity and the UXO probability.

[0023] Specifically, first, magnetic field data in the sea area are collected by a marine magnetic detector. These data include signals from various interference sources in the environment. To ensure that the UXO magnetic field anomaly signal can be accurately extracted, the collected magnetic field data need to be preprocessed. First, noise reduction processing is performed. By applying various noise filtering algorithms, environmental magnetic field interference is removed, and the real magnetic field signal is retained. To further improve the signal quality, in this embodiment, various filtering processing techniques are adopted, including low-pass filtering, band-pass filtering, etc. These methods can effectively remove high-frequency noise and low-frequency background interference, ensuring that the target signal is clearer; After the preprocessing of the magnetic field data is completed, an anomaly feature extraction technique is adopted to further enhance the magnetic field anomaly signal generated by the UXO. This step uses certain mathematical models and algorithms to extract the anomaly points in the data and calibrate the features of these anomaly points, providing a basis for subsequent UXO target recognition. This process can not only identify possible UXO targets but also provide preliminary information for subsequent position and depth estimation. The detection of magnetic field anomalies usually adopts methods of gradient analysis and anomaly feature extraction. Assume that the magnetic detector collects magnetic field data at the spatial point (x, y, z) , then the local magnetic field gradient can be calculated by the following formula: ; Gradient: represents the change rate of the magnetic field in space and reflects the change of the magnetic field intensity with the spatial coordinates The change rate of the magnetic field in the direction; The change rate of the magnetic field in the direction; The change rate of the magnetic field in the direction, so it is used in use to detect areas with large magnetic field changes and help locate magnetic field anomaly points.

[0024] Magnetic field anomaly detection criterion The intensity M of a magnetic field anomaly can be expressed as the gradient norm: ; where: Magnetic field anomaly intensity, that is, the magnitude of the change rate of the magnetic field in three-dimensional space.

[0025] Square root of the sum of squares: Calculate the Euclidean norm of the magnetic field gradient, which represents the magnitude of the gradient vector. According to magnetic anomaly detection theory, if M exceeds a certain threshold θ, it is determined as an anomaly point. Therefore, when M > θ, this point is determined as a magnetic anomaly point; In addition, the application of side-scan sonar imaging technology in UXO detection is also one of the key technologies of the present invention. During the side-scan sonar imaging process, multi-band sonar data is used. By collecting acoustic wave reflection information at different frequencies, multi-scale seabed images are generated. These images provide an all-round understanding of the seabed situation from different resolution perspectives, which can better help analyze and identify UXO targets. The multi-scale processing of sonar images can accurately capture target features of different sizes and depths, thereby improving the accuracy of UXO identification; In the process of data processing and fusion, in this embodiment, a dynamic UXO probability information field is constructed through the joint analysis of magnetic field data and side-scan sonar image data. This information field is based on the combination of the magnetic anomaly vector field and sonar image features. The magnetic anomaly vector field is generated through a magnetic field reconstruction algorithm, and further combined with image information to dynamically update the UXO probability information field. According to different anomaly intensity distributions, this information field can not only determine the location of the UXO target, but also estimate the depth of the target. This dynamically updated process enables the UXO detection path to be adjusted according to the real-time changes in magnetic field and sonar data, thereby improving the detection efficiency and accuracy; The magnetic field reconstruction algorithm is an innovative technology of the present invention. It reconstructs the magnetic field by using marine magnetic field data and combining the unique magnetic field anomaly characteristics of UXOs. Specifically, the magnetic field reconstruction algorithm models and processes the original magnetic field data to reconstruct a more accurate magnetic anomaly vector field, and generates a probability distribution field of UXO targets based on this. Using this probability information field, the position and depth of the target can be estimated, and the detection path can be automatically planned according to real-time information. In this process, the accuracy and real-time performance of the reconstruction algorithm are the keys to improving UXO detection performance As a further expansion, the magnetic field reconstruction algorithm described in the present invention can further improve the detection accuracy of UXOs by combining different magnetic field data processing technologies, such as gradient analysis and anomaly recognition algorithms. By combining different magnetic field data sources and using multi-algorithm fusion technology, potential UXO targets can be more accurately identified in complex marine environments.

[0026] Through the comprehensive application of the above technologies, the present invention can achieve efficient and accurate UXO detection in complex marine environments. Compared with traditional methods, the present invention can not only improve the detection accuracy and reduce false alarms, but also significantly improve the operation efficiency through automated detection path planning. Especially during the process of dynamically updating the UXO probability information field, it can adjust the detection strategy in real time to adapt to different marine conditions, thereby achieving more flexible and accurate target recognition and positioning.

[0027] The seismic reflection technology uses a multi-channel seismic data acquisition device. By controlling the acoustic wave frequency and incident angle, it generates seismic reflection profiles at different depths to accurately estimate the burial depth of UXO. The data fusion uses a multi-modal data fusion algorithm based on Bayesian inference or Kalman filtering to jointly process the data from marine magnetic detectors, side-scan sonars, and seismic reflections to form a three-dimensional spatial distribution model. When using the deep learning image segmentation technology to classify and identify sonar images, it combines a convolutional neural network and a generative adversarial network to optimize image recognition from multiple angles and continuously improve the accuracy of target location judgment through model iteration.

[0028] Specifically, first, the multi-channel seismic reflection technology is used to estimate the UXO depth. Generally, a preliminary estimate of the UXO burial depth can be obtained through the seismic reflection profile. The core of this step lies in reasonably controlling the acoustic wave frequency and incident angle to generate reflection images at different depths. During data acquisition, a multi-channel seismic data acquisition device is used to obtain reflection profiles at multiple depth levels by using different frequency and angle combinations. Through the acoustic wave propagation equation: ; Where: : Sound pressure The Laplace operator in space, representing the degree of curvature (or diffusion) of the sound pressure in space The second-order partial derivative of the sound pressure with respect to time, representing the acceleration change of the sound pressure.

[0029] : Sound speed, which depends on the physical properties of the medium (such as seawater or sediment).

[0030] 0: Represents no external force source (homogeneous medium) Furthermore, by calculating the reflection coefficient for multiple groups of data: ; Where: R: Reflection coefficient, representing the proportion of the reflection intensity of the acoustic wave at the medium interface; ρ1, ρ2: Densities of the upper and lower media (such as sediment and seawater); v1, v2: Sound speeds in the upper and lower media; Molecular part: The sound pressure difference caused by the medium discontinuity; Denominator part: The sum of the total energy flux density, representing the sum of the incident and transmitted wave energies. By analyzing the change in reflection intensity and its distribution in the depth direction, the burial depth of the UXO can be preliminarily located.

[0031] To fuse the data, the data of the magnetic detector and the side-scan sonar are combined to further improve the recognition accuracy of the UXO. The magnetic detector is used to obtain the magnetic anomaly signal of the UXO, and the side-scan sonar can generate an underwater three-dimensional image. To enhance the recognition of the shape and burial position of the UXO, Bayesian inference is used for multi-modal data fusion in the present invention, using the Bayesian formula: ; Where: : The posterior probability of the existence of the UXO under the condition of the observed data Condition.

[0032] : The probability (likelihood) of the observed data on the premise of the existence of the UXO.

[0033] : The prior probability (empirical or historical probability) of the existence of the UXO.

[0034] : The marginal probability of the observed data in all possible cases. The data can also be updated using the Kalman filter. By observing the real-time data, the estimation of the UXO position is dynamically adjusted to enhance the tracking ability of moving targets.

[0035] It further includes anomaly detection. By using an anomaly detection algorithm to monitor the real-time data, abnormal signals are identified and a quick rescan is performed to improve the accuracy of UXO detection; The seismic reflection information further corrects the burial depth of the UXO by combining the time delay between reflection layers and the wave velocity model with the type of seabed sediment, and provides density contrast data between target layers to optimize the estimation of the UXO depth. The reflection depth is estimated by the following formula: ; Where v is the wave velocity, Δt is the time delay of the acoustic wave reflection, and d is the depth of the reflection point; the rock interface information is used to help estimate the burial depth of the UXO and supplement the deficiencies in the data.

[0036] Specifically, for 3D space reconstruction, a point cloud data reconstruction algorithm or a voxel reconstruction algorithm is adopted. First, seabed surface data is obtained from sonar images, and depth information is further supplemented using seismic profile data. The point cloud data reconstruction algorithm utilizes the spatial coordinates of each seabed sampling point extracted from the sonar image. By fitting these data points, the 3D surface of the seabed is reconstructed. In 3D space, the coordinates of each point are represented as , where are the 2D coordinates of the sonar image and is the depth information of this point. These data are processed through computational geometry algorithms, such as the least squares method, to obtain a function of the seabed surface; The voxel reconstruction algorithm adopts a block method, dividing the seabed space into multiple small units (voxels). Each voxel generates a more accurate 3D model by calculating its internal properties (such as depth, density). Voxel reconstruction can better handle the complex structure of seabed sediments. The basic idea of voxel reconstruction is to divide the space into small units, with the size of each unit being , and updating its value in 3D space using the properties of each voxel. This 3D modeling method can make full use of the complementarity of sonar images and seismic data, providing an accurate seabed environment model for UXO detection. Whether through point cloud reconstruction or voxel reconstruction, the generated 3D model can accurately reflect the seabed morphology and its impact on UXO detection; , its mean value is , and the standard deviation is , then the standardized value of this point can be calculated by the following formula: ; If , then this data is determined to be an abnormal signal, and the system will trigger a rescan. Abnormal detection not only improves the detection accuracy but also ensures efficient resource allocation; When combining seismic reflection information with sonar images can provide more comprehensive information about the seabed environment. By combining the interlayer time delay of seismic reflections with the wave velocity model and optimizing the comparison of formation reflection intensity and sediment types, we can more accurately estimate the burial depth of UXO. In actual operation, the reflection time delay of sound waves is closely related to the wave velocity. We can use the reflection time delay and the wave velocity to estimate the depth of UXO

[0037] ; In this way, we can accurately map the reflection data at different depths to the spatial position of the UXO and accurately estimate its depth.

[0038] The rock interface information provides additional support for UXO depth estimation. Due to the different characteristics of the sediment layer and the rock layer, there are differences in the propagation speed and reflection intensity of the reflected waves. By analyzing the seismic data and combining the density ratio of the sediment layer and the rock layer, the buried depth of the UXO can be further corrected. When the density of the seabed sediment layer is combined with the wave velocity data, the actual buried depth of the UXO can be more accurately calculated, thus improving the accuracy formula of the explosive disposal operation as follows: Wave velocity of the rock layer and the wave velocity of the sediment layer There are differences between them. Therefore, by calculating the reflected wave velocity at the rock interface, the depth of the UXO can be accurately determined. The reflection time delay of the reflection at the rock interface is estimated by the following formula: ; This formula can be inverted through the actual acoustic wave reflection data to obtain the depth of the UXO. When the sediment layer is deeper, the rock interface information can further correct and supplement the deficiencies in the data and enhance the detection accuracy.

[0039] Please refer to Appendix Figure 2 , a submarine UXO positioning system in the sea area, including: A data acquisition module, used to obtain the original data of the seabed environment and UXO targets from marine magnetometers, side-scan sonars, and seismic reflection technologies, including magnetic field data, sonar images, and seismic reflection data; A data preprocessing module, connected to the data acquisition module, used to clean, filter, and enhance the acquired data to make the signal clearer; A data fusion module, connected to the data preprocessing module, used to integrate multi-modal data collected by different sensors to form a complete seabed environment information; A machine learning analysis module, connected to the data fusion module, used to identify and classify UXO targets based on historical UXO data and current acquired data through a machine learning model; A three-dimensional reconstruction and visualization module, used to perform three-dimensional reconstruction and visualization of the spatial position of the UXO and the seabed environment; An intelligent decision-making and optimization module, used to dynamically adjust the detection strategy, optimize the detection path and sensor parameters, and improve the UXO positioning efficiency; A system feedback and adjustment module for dynamically adjusting system parameters; A data storage and management module, connected to the system feedback and adjustment module, for storing and managing all raw data, processed data, and analysis results.

[0040] Specifically, the data acquisition module is responsible for obtaining raw data of the seabed environment and UXO targets from marine magnetic detectors, side-scan sonars, and seismic reflection technologies, including magnetic field data, sonar images, and seismic reflection data. The data preprocessing module processes the raw data using techniques such as denoising, filtering, and abnormal feature extraction to remove environmental noise and equipment interference and highlight signals related to UXOs. The machine learning analysis module, based on historical UXO data and real-time acquisition data, identifies and classifies potential UXO targets through a machine learning model, automatically recognizes and classifies UXO targets, reduces false alarms, and improves the accuracy of target recognition. By training the machine learning model, analyzing newly acquired data, and predicting the probability of UXO presence, it automatically screens potential targets; A three-dimensional reconstruction and visualization module that performs three-dimensional reconstruction and visualization of the spatial position of UXOs and the seabed environment. It provides an intuitive three-dimensional view to help operators accurately understand the position of UXOs and their surrounding environment. By using point cloud data reconstruction algorithms or voxel reconstruction algorithms, it fuses the acquired sonar images and seismic reflection data to generate an accurate three-dimensional model. The intelligent decision-making and optimization module dynamically adjusts the detection strategy, optimizes the detection path and sensor parameters, and improves the detection efficiency. It optimizes the detection process according to different environmental conditions and real-time data to ensure efficient and accurate UXO positioning. Through optimization algorithms such as reinforcement learning, it analyzes historical data and real-time detection data to adjust the detection path and sensor settings to adapt to changing environments and task requirements. The system feedback and adjustment module dynamically adjusts system parameters based on real-time data feedback to ensure the optimal working state of the system, improve the flexibility and adaptability of the system, and ensure the efficient completion of detection tasks. By interacting with other modules, it monitors the performance of the system in real time and automatically performs parameter adjustment and optimization. The data storage and management module is responsible for storing and managing all raw data, processed data, and analysis results to ensure the security, integrity, and traceability of the data, facilitating subsequent analysis and historical data review. It stores all data in a database and uses a database management system to ensure the secure storage and rapid retrieval of data.

[0041] Please refer to the appendix Figure 3 - Appendix Figure 12 , a process for treating underwater UXOs in a sea area, including the following steps: S1. Prepare a shaped charge, determine the position and burial depth data of the UXO based on detection, and clean part of the overburden; S2. Position the positioning vessel. Based on the position of the UXO determined by detection, use wide-area differential positioning to move the reef blasting vessel or a self-made barge with a guiding frame to the medicine placement position. Use 4-anchor or 6-anchor positioning, and process and assemble the shaped charge on the positioning vessel. S3. After the medicine placement is completed, move the vessel to a safe distance, charge and detonate. Under the high-temperature jet generated by the explosion of the shaped charge, the shell of the UXO is penetrated, causing the internal historical explosive to deflagrate. The pressure inside the projectile increases, and the shell ruptures and decomposes, enabling low-level treatment of the UXO.

[0042] In S1, when the burial depth of the UXO is <1 m, direct medicine placement and blasting treatment can be carried out. When the burial depth of the UXO is >1 m, use large-scale sediment suction equipment to clear the surface covering sediment in an area with a diameter of 6 m to 8 m centered on the determined planar coordinate position. Measure the seabed elevation using a water weight, sounding rod, and multibeam sonar. After confirmation of suction, make the thickness of the sediment covering layer above the unexploded ordnance <1 m, and blow and suck part of the surface covering layer.

[0043] Specifically, when the burial depth of the UXO is less than 1 meter, since its top is close to the seabed surface layer, large-scale cleaning operations are not required, and direct medicine placement and blasting treatment can be carried out, saving operation time and resources. Such treatment can quickly complete the directional destruction task on the premise of ensuring the stability around the UXO; When the burial depth of the UXO is greater than 1 meter, in order to ensure the transmission of the blasting effect to the target, local excavation operations must be carried out first. At this time, use large-scale sediment suction equipment (such as a cutter suction dredging equipment or a high-pressure water jet nozzle) centered on the planar coordinates of the UXO to blow and suck and remove the surface covering sediment within a range with a diameter of 6 meters to 8 meters. During the blowing and sucking process, the equipment effectively removes the sediment layer by combining hydraulic disturbance and negative pressure suction, avoiding disturbing or damaging the UXO; During the operation process, tools such as a water weight (measuring the relative position change of the UXO), a sounding rod (manually or remotely measuring the seabed height), and a multibeam sonar (precisely mapping the terrain change of the cleared area) need to be used in combination to continuously monitor the seabed elevation. Through these devices, accurately measure the remaining thickness of the sediment after suction to ensure that the covering layer above the unexploded ordnance is finally thinned to less than 1 meter, creating suitable conditions for subsequent medicine placement, ensuring the blasting effect while reducing the operation risk, precisely controlling the cleaning depth and range, neither overly disturbing the surrounding environment of the target nor fully exposing the top structure of the UXO to ensure the effective penetration and detonation of the shaped charge.

[0044] In S1, the preparation of the shaped charge includes the following sub-steps: There are two types of shaped charges. One is the cylindrical shaped charge, and the other is the strip-shaped shaped charge. The housing of the shaped charge is welded with thin steel plates, the internal shaped charge liner is made of pressed steel plates, and the charge uses pressure-resistant and waterproof emulsion explosive. This explosive uses glass microspheres as the sensitizer of the emulsion explosive, and the sensitizer is used to improve the pressure resistance, detonation velocity and explosion power of the emulsion explosive; Considering that the density of the explosive is close to that of seawater, there is a sealed air cavity at the bottom of the shaped charge. The sealed air cavity is used to prevent the projectile from tipping over and floating, and a counterweight is installed at the bottom of the shaped charge; The single charge of the cylindrical shaped charge is 28 kg, and the charge of the strip-shaped shaped charge is 75.5 kg. The main charge in the charge uses pressure-resistant emulsion explosive, and the initiating detonator uses an enhanced digital electronic detonator.

[0045] Specifically, there are two forms of shaped charges: cylindrical and strip-shaped. The cylindrical shaped charge is usually used for smaller blasting requirements, while the strip-shaped shaped charge is suitable for occasions that require greater explosion power. The design of each charge is to ensure the maximization of the blasting effect while meeting the requirements of different water areas and UXO types; The outer shell of the shaped charge is welded by thin steel plates, which not only ensures the structural strength of the charge but also avoids being too heavy to affect buoyancy. The internal shaped charge liner is made of pressed steel plates and is used to focus the explosion energy and improve the explosion efficiency. The design of the shaped charge liner makes the energy generated by the explosion more concentrated, ensuring that it can effectively penetrate the outer shell of the UXO; The shaped charge is filled with pressure-resistant and waterproof emulsion explosive. This emulsion explosive uses glass microspheres as the sensitizer, and the function of the sensitizer is to enhance the pressure resistance, detonation velocity and explosion power of the emulsion explosive. The emulsion explosive itself has strong underwater adaptability, can withstand greater water pressure and produce a powerful explosion, and the addition of glass microspheres improves its stability and explosion effect in a high-pressure environment; Considering that the density of the explosive is close to that of seawater, a sealed air cavity is designed at the bottom of the shaped charge. This sealed air cavity is used to prevent the charge from tilting or floating during underwater operation, thus ensuring the stability of the charge during the blasting process. In addition, a counterweight is installed at the bottom of the shaped charge to further enhance stability and prevent position deviation under the influence of water flow.

[0046] In S2, the assembled shaped charges use cylindrical shaped charges and strip-shaped shaped charges, which are determined according to the type and shape parameters of the unexploded ordnance obtained from the detection data of the marine magnetic sweeper. A small-spacing grid should be used. When a strip-shaped shaped charge is used in the central area, the grid spacing can be appropriately increased, but it should not be greater than 1 m. In the area of 2-3 m centered on the positioning point, small-spacing grid charging is carried out, and the grid is adjusted according to the shape of the UXO to confirm that the outer shell of the ordnance can be penetrated by the shaped charge jet.

[0047] Specifically, for the selection of shaped charges and grid charge placement, shaped charges are divided into two types: cylindrical and strip-shaped. The selection of the charge is determined based on the type, shape, and size of the UXO obtained from the detection data. When the location and size of the UXO are known: for smaller or shallower UXOs, cylindrical shaped charges are usually selected. This type of charge is suitable for small-scale blasting tasks and can provide sufficient destructive power. For larger or deeper-buried UXOs, strip-shaped shaped charges are selected. This type of charge can provide greater explosive power and is suitable for more complex situations; During the layout process of shaped charges, the grid charge spacing should be adjusted according to the type and shape of the UXO. The size of the grid spacing directly affects the blasting effect: For smaller and irregularly shaped UXOs, a small-spacing grid (such as 30 cm to 50 cm) should be used to ensure that each charge can cover the outer shell of the UXO and enhance the penetration power of the shaped jet. In the central area of the UXO, especially when the UXO is larger and relatively regular in shape, when using strip-shaped shaped charges, the grid spacing can be appropriately increased. However, the grid spacing shall not exceed 1 meter. This is to reduce the number of charges while maintaining sufficient blasting effect; Positioning and adjusting the grid layout, the layout of shaped charges should be adjusted around the positioning point of the UXO. Usually, within a range of 2 to 3 meters from the UXO, a small-spacing grid charge placement is used. Through precise positioning, ensure that the charges are evenly distributed around the target, so that the shaped jet can penetrate the outer shell of the UXO from multiple directions, ensuring sufficient explosive power of the charges. According to the shape and burial depth of the UXO, adjust the grid layout so that the charges can cover the weakest area of the UXO. Through reasonable charge placement, ensure that the shaped jet of the charges can directly penetrate the outer shell of the projectile and produce an effective explosion effect. Through reasonable design of the grid layout, ensure that the shaped charges can generate sufficient pressure on the outer shell of the UXO. The high-temperature and high-pressure effect generated by the shaped jet at the moment of explosion will quickly penetrate the outer shell of the UXO, causing the internal explosive to detonate. According to the type, shape, and burial depth of the UXO, the density and spacing of the charge layout need to be dynamically adjusted to ensure the best explosion effect.

[0048] After step S3: Since the historical explosive combustion in the UXO projectile has damaged the outer shell, changing the overall magnetic characteristics of the unexploded bomb, re-detect with a marine magnetic detector, compare with the database, and confirm the demining result. Under the condition that visibility permits, use an underwater camera robot to take pictures. Analyze the demining treatment result based on the image data. According to the result, it can be recognized that this process can be repeated, and the secondary blasting positioning grid is misaligned by half a position.

[0049] Specifically, after the explosion of the shaped charge, the destruction of the UXO shell causes changes in magnetic characteristics. Through re-detection with a marine magnetic detector and comparison with the database data, it is confirmed whether the UXO has been completely destroyed. By comparing the re-detection results with the database data, the explosive disposal result is confirmed. If they match, it indicates that the UXO has been completely destroyed; if there are abnormalities, follow-up processing is required. Under conditions where visibility permits, an underwater camera robot is used to take images of the blasting area, and the explosive disposal effect is confirmed through the image data.

[0050] According to the image analysis results, if any unprocessed parts are found, secondary blasting can be carried out. During secondary blasting, the grid is misaligned by half a position to ensure coverage of the unprocessed area, enhance the blasting effect, confirm the integrity of UXO processing, and perform necessary secondary blasting.

[0051] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for positioning UXO underwater in a sea area, characterized in that, It includes the following steps: First, use a marine magnetic detector to collect underwater magnetic field intensity change data, construct a magnetic anomaly vector field based on this magnetic field data, and identify the abnormal central region through a magnetic field reconstruction algorithm; Then, use a side-scan sonar to image and scan the seabed, obtain seabed images for verifying the magnetic anomaly region and assisting in confirming whether it is a potential target of UXO. Combine deep learning image segmentation technology to analyze the shape, texture, and edges of potential UXO targets in the sonar images to enhance the verification effect of the magnetic anomaly region; Then, adopt seismic reflection technology to obtain rock interface information at different depths of the seabed through acoustic wave reflection, and estimate the vertical burial depth and cover thickness of UXO in the sediment layer based on the echo delay and interlayer structure inversion algorithm; Integrate the marine magnetic detector, side-scan sonar, and seismic reflection data through an adaptive data fusion algorithm to construct a multi-source perception-driven UXO probability information field, and use this information field as the basis to guide subsequent detection strategies. The integration is used to generate a multi-dimensional seabed detection map and optimize the estimation of the potential position and depth of UXO; Use a machine learning model to train historical UXO case data, predict the existence probability and burial depth of UXO, and classify the confidence of abnormal targets through a model uncertainty estimation mechanism to improve the false alarm filtering ability; On the basis of constructing the information field, introduce an information entropy weight distribution, and automatically select the detection priority area through the maximum information gain criterion for detection efficiency and path adjustment; Combine virtual reality and augmented reality technologies to perform underwater three-dimensional space reconstruction, display the position and depth of UXO in real time, and feedback the detection decision results and confidence evaluation of the model output through a visual interaction interface; Finally, through reinforcement learning technology, dynamically adjust the detection strategy and path, and automatically perform dynamic adjustment according to real-time detection data and environmental changes, and finally complete the precise positioning of UXO.

2. The method for positioning UXO underwater in a sea area according to claim 1, wherein: The magnetic field data collected by the marine magnetic detector undergoes preprocessing, including noise removal, filtering, and abnormal feature extraction, to remove environmental noise interference and enhance the magnetic field anomaly signal generated by UXO. During the side-scan sonar imaging process, multi-band sonar data is used to simultaneously obtain seabed images with different resolutions to form multi-scale image information, thereby improving the recognition accuracy of UXO; The magnetic anomaly vector field is constructed from the magnetic field data by using a magnetic field reconstruction algorithm to generate a magnetic anomaly vector field. According to different abnormal intensity distributions, a dynamic UXO probability information field is constructed to determine the target position and depth estimation, and a detection path is automatically generated according to this information field; The generation of the magnetic anomaly vector field is carried out through the following formula: ； Among them, is the total magnetic field measured, is the background magnetic field, is the magnetic anomaly vector field; The dynamic UXO probability information field is constructed according to the relationship between the magnetic field anomaly intensity and position through the following formula: ； wherein, is the probability of the presence of UXO at the location , is a constant that controls the relationship between the magnetic field anomaly intensity and the UXO probability.

3. The method for positioning UXO underwater in a sea area according to claim 1, wherein: The seismic reflection technology uses a multi-channel seismic data acquisition device. By controlling the acoustic wave frequency and incident angle, it generates seismic reflection profiles at different depths to accurately estimate the burial depth of UXO. The data fusion uses a multi-modal data fusion algorithm based on Bayesian inference or Kalman filtering to jointly process the data from marine magnetic detectors, side-scan sonars, and seismic reflections to form a three-dimensional spatial distribution model; When the deep learning image segmentation technology classifies and identifies sonar images, it uses a combination of convolutional neural networks and generative adversarial networks to optimize image recognition from multiple angles and continuously improve the accuracy of target location judgment through model iteration.

4. A method for positioning UXO underwater in a sea area according to claim 1, characterized in that: The three-dimensional space reconstruction uses a point cloud data reconstruction algorithm or a voxel reconstruction algorithm to generate a three-dimensional model of the seabed environment by combining sonar images and seismic profile data. It further includes anomaly detection. By using an anomaly detection algorithm to monitor real-time data, it identifies anomaly signals and performs a quick rescan to improve the accuracy of UXO detection. The seismic reflection information further corrects the burial depth of UXO by combining the inter-layer time delay of reflection and wave velocity model with the seabed sediment type, and provides the target inter-layer density contrast data to optimize the estimation of UXO depth. The reflection depth is estimated by the following formula: ； where v is the wave velocity, Δt is the time delay of acoustic wave reflection, and d is the reflection point depth. The rock interface information is used to help estimate the burial depth of UXO and supplement the insufficient depth in the data. The three-dimensional reconstruction uses a point cloud data reconstruction algorithm or a voxel reconstruction algorithm to generate a three-dimensional model of the seabed environment by combining sonar images and seismic profile data. It further includes anomaly detection. By using an anomaly detection algorithm to monitor real-time data, it identifies anomaly signals and performs a quick rescan to improve the accuracy of UXO detection. The seismic reflection information further corrects the burial depth of UXO by combining the inter-layer time delay of reflection and wave velocity model with the seabed sediment type, and provides the target inter-layer density contrast data to optimize the estimation of UXO depth. The reflection depth is estimated by the following formula: ； where v is the wave velocity, Δt is the time delay of acoustic wave reflection, and d is the reflection point depth. The rock interface information is used to help estimate the burial depth of UXO and supplement the deficiencies in the data.

5. An underwater UXO positioning system in a sea area, characterized in that, Applied to a method for positioning underwater UXO in a sea area according to any one of claims 1-4, it includes: A data acquisition module for obtaining the original data of the seabed environment and UXO targets from a marine magnetic detector, a side-scan sonar, and seismic reflection technology, including magnetic field data, sonar images, and seismic reflection data. A data preprocessing module connected to the data acquisition module for cleaning, filtering, and enhancing the acquired data to make the signal clearer. A data fusion module connected to the data preprocessing module for integrating multi-modal data collected by different sensors to form a complete seabed environment information. A machine learning analysis module connected to the data fusion module for identifying and classifying UXO targets based on historical UXO data and currently acquired data through a machine learning model. 3D reconstruction and visualization module, used for 3D reconstruction and visual presentation of the spatial position of UXO and the seabed environment; Intelligent decision-making and optimization module, used for dynamically adjusting the detection strategy, optimizing the detection path and sensor parameters, and improving the positioning efficiency of UXO; System feedback and adjustment module, used for dynamically adjusting system parameters; Data storage and management module, connected to the system feedback and adjustment module, used for storing and managing all raw data, processed data and analysis results.

6. A process for underwater UXO treatment in the sea area, which is applied to an underwater UXO positioning method in the sea area according to any one of claims 1-4, characterized in that, Including the following steps: S1. Prepare the shaped charge, determine the position and burial depth data of UXO according to detection, and clean part of the overburden; S2. Position the positioning ship. According to the position of UXO determined by detection, use wide area differential positioning to move the reef blasting ship or a self-made square barge with a guiding frame to the charge placement position, use 4 anchors or 6 anchors for positioning, and process and assemble the shaped charge on the positioning ship; S3. After the charge placement is completed, move the ship to a safe distance, charge and detonate. Under the high-temperature jet generated by the explosion of the shaped charge, the shell of UXO is penetrated, causing the internal historical explosive to deflagrate, the pressure in the projectile body increases, the shell ruptures and decomposes, and UXO is subjected to low-level treatment.

7. A process for treating UXO underwater in a sea area according to claim 6, characterized in that: In S1, when the burial depth of UXO < 1m, the charge placement and blasting treatment can be directly carried out. When the burial depth of UXO > 1m, use large-scale sediment suction equipment to clear the surface sediment in an area with a diameter of 6m to 8m centered on the determined plane coordinate position. Measure the seabed elevation through a water drogue, sounding rod, and multibeam sonar. After confirmation by suction, make the thickness of the sediment overburden above the unexploded bomb < 1m, and blow and suck part of the surface overburden.

8. A process for treating underwater UXO in a sea area according to claim 6, characterized in that: In S1, the preparation of the shaped charge includes the following sub-steps: There are two types of shaped charges. One is a cylindrical shaped charge, and the other is a strip-shaped shaped charge. The shell of the shaped charge is welded with thin steel plates, the internal shaped charge liner is made of pressed steel plates, and the charge is filled with pressure-resistant and waterproof emulsion explosive. This explosive uses glass microspheres as the sensitizer of the emulsion explosive, and the sensitizer is used to improve the pressure resistance, detonation velocity and explosion power of the emulsion explosive; Considering that the density of the explosive is close to the density of seawater, there is a sealed air cavity at the bottom of the shaped charge, and the sealed air cavity is used to prevent the projectile body from tipping over and floating, and a counterweight is installed at the bottom of the shaped charge; The single charge amount of the cylindrical shaped charge is 28kg, and the charge amount of the strip-shaped shaped charge is 75.5kg. The main charge in the charge package uses pressure-resistant emulsion explosive, and the initiating detonator uses an enhanced digital electronic detonator.

9. A seabed underwater UXO treatment process according to claim 6, characterized in that: In S2, the assembled shaped charges use cylindrical shaped charges and strip-shaped shaped charges, which are determined according to the type and shape parameters of the unexploded bomb obtained from the detection data of the marine magnetic scanner. A small-spacing grid should be used. When a strip-shaped shaped charge is used in the central area, the grid spacing can be appropriately increased, but it should not be greater than 1m. In the area of 2 to 3m centered on the positioning point, charge the small-spacing grid, and adjust the grid according to the shape of UXO to confirm that the shell of the projectile body can be penetrated by the shaped charge jet.

10. A method for processing underwater UXO in a sea area according to claim 6, characterized in that: After Step S3: Since the historical explosive in the UXO warhead burns explosively and damages the outer shell, changing the magnetic characteristics of the whole unexploded bomb, re-detect it with a marine magnetic detector, compare it with the database to confirm the explosive disposal result. Under the condition that visibility permits, use an underwater camera robot to take pictures, analyze the explosive disposal result based on the image data. According to the result, it can be determined that this process can be repeated, and the secondary blasting positioning grid is misaligned by half a position.

Images