A hydrological glider searching method for surrounding waters of a typical island reef in the South China Sea

Abstract

The application relates to a South China Sea typical island reef surrounding hydrological submersible searching method, which comprises the following steps: generating an elliptical searching area consistent with the sea current direction based on the GPS coordinates of the submersible laying point and marine environment data, and performing grid subdivision; evaluating the stability and generating a navigation path through the ship motion parameters and fluid simulation optimized trajectory; using the coordinate transformation and speed constraint, and adopting an A* algorithm to plan a searching path parallel to the long axis of the ellipse; dynamically updating the navigation state through signal strength detection and frequency adjustment optimization communication; comprehensively considering the communication success rate and time index, and generating a task optimization strategy through weighted decision; adjusting the power and trajectory according to the path deviation, and attempting acoustic communication in the fixed-point grid; finally, verifying the task completion state through GPS positioning and path planning, and outputting a task data set, so that the method realizes high-precision and self-adaptive submersible searching and recovery control.

Description

Technical Field

[0001] This application relates to the field of electronic digital data processing technology, specifically to a method for searching for hydrographic moorings around typical islands and reefs in the South China Sea. Background Technology

[0002] When conducting underwater mooring recovery operations in deep-sea areas, if effective communication cannot be established between the acoustic release device and the shipboard deck unit, it is usually necessary to rely on human experience to judge the area where the mooring may drift and conduct blind searches. Existing technologies lack systematic search strategies and often rely on the experience of operators, resulting in problems such as low search efficiency, low success rate, and great influence from the marine environment.

[0003] Especially when the buoy is displaced, has insufficient power, weak signal, or is interfered with by organisms or fishing nets, existing methods are unable to quickly and accurately locate the buoy, making buoy retrieval difficult and even causing equipment loss. Summary of the Invention

[0004] In order to solve the problems existing in the prior art, the purpose of this application is to provide a method for searching for hydrographic moorings around typical islands and reefs in the South China Sea.

[0005] The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea as described in this application includes the following steps:

[0006] S101. An elliptical search area is set with the location of the mooring buoy as the center, based on marine environmental data, with its major semi-axis direction consistent with the historical ocean current direction.

[0007] S102. Based on the sea condition and marine topography data of the search area, construct a simulation environment and import the search vessel model into it;

[0008] S103. In the simulation model, with the constraints of keeping the probe vertical and the speed below a preset threshold, the ship's speed, heading, and a series of initial search paths parallel to the major axis of the ellipse are output, and the paths cover the entire search area.

[0009] S104. Control the search vessel to drift or propel itself along the initial search path, and attempt to communicate with the buoy releaser via the acoustic deck unit during the voyage.

[0010] S105. Calculate a comprehensive coefficient based on communication success rate and search time, and dynamically optimize the search strategy;

[0011] S106. If the actual navigation path deviates from the initial path, adjust it back through power. If the probe is not vertical, suspend communication and set multiple grid points on the path to attempt fixed-point communication.

[0012] S107. If the buoy becomes dislodged, a surface search will be conducted using the GPS positioning system.

[0013] Preferably, in step S101, starting from the GPS coordinates of the buoy deployment point, the latitude and longitude of the center point are first obtained;

[0014] Subsequently, based on marine environmental data such as temperature, salinity, and current velocity, preprocessing was performed and key information was extracted. If current velocity data exists, the historical ocean current direction angle was calculated through vector analysis.

[0015] By integrating the coordinates of the center point and the ocean current angle, the shape parameters of the elliptical region are generated, ensuring that its major semi-axis aligns with the direction of the ocean current. The boundary is further optimized through spatial positioning accuracy analysis to form the final search area. If the area exceeds a preset threshold, a grid division algorithm is used to subdivide it, resulting in a high-precision search grid.

[0016] Preferably, in step S102, the position and orientation data of the ship's three-dimensional model are obtained from the initial simulation scene as motion parameters;

[0017] If the speed exceeds the threshold, the trajectory is corrected by a vector adjustment algorithm to generate optimized parameters. Combined with sea state characteristics, the ship-sea interaction force is calculated using hydrodynamic simulation. Then, the ship's stability is evaluated by finite element analysis. If the attitude deviation exceeds the threshold, the model orientation is corrected.

[0018] Based on the terrain information, the optimal navigation path is generated using a path planning algorithm. Key nodes are extracted and aligned to the simulation environment through spatial mapping to form the final scene. Finally, the dynamic simulation results are output through real-time rendering.

[0019] Preferably, in step S103, based on the sea state data of the search area, the initial position and orientation of the ship are obtained, and a standard initial state is generated by combining coordinate transformation with the vertical constraint of the probe. Based on this state and the speed threshold, the allowable speed range is calculated. If the speed meets the requirements, the orientation angle is extracted and the heading control parameters are generated by vector mapping. Subsequently, based on the heading control data and the major axis characteristics of the ellipse, a series of initial search paths parallel to the major axis are planned using the A* algorithm. The path nodes are adjusted to align with the sea state through spatial interpolation to optimize the path. Finally, combined with the speed and heading constraints, and through real-time simulation and trajectory correction, the final navigation path covering the search area is generated.

[0020] Preferably, in S104, based on the initial navigation and communication data, the real-time position of the vessel is obtained through spatial mapping. Accordingly, the signal strength is detected in combination with the communication frequency range. If it is lower than the threshold, the frequency parameters are adjusted to optimize communication. The acoustic deck unit configuration is updated using the optimized frequency. Then, the vessel's navigation trajectory is adjusted through path optimization technology. The path and real-time position are comprehensively adjusted, and the vessel's drift or power status is updated using dynamic navigation technology. Finally, based on the navigation status and signal strength, it is evaluated whether the communication task with the mooring releaser has been completed.

[0021] Preferably, in step S105, based on the communication success rate and search time in the task data, a standardized dataset is generated after preprocessing. Then, normalization parameters are extracted and weighted averaged with preset weights to obtain a comprehensive coefficient. Using this coefficient and the current search strategy, a preliminary optimization strategy is generated through a decision tree algorithm. If the path parameters exceed the threshold, the path is corrected.

[0022] If the communication frequency is insufficient, the signal is optimized, and these adjustment parameters are combined to form a comprehensive optimization strategy. Finally, the strategy is dynamically updated in conjunction with real-time task data to generate a task execution plan.

[0023] Preferably, in step S106, the deviation angle is calculated using path comparison technology based on the deviation data between the actual and initial navigation paths. The propulsion power is then adjusted in conjunction with the power parameters to generate a power adjustment dataset. Based on the adjusted power, the navigation path is updated using trajectory correction technology to obtain a regression path. Subsequently, trajectory parameters are extracted and detected in conjunction with probe attitude data. If the probe is found to be not perpendicular, multiple fixed-point grids are generated on the regression path, and acoustic communication signals are attempted to be sent at each grid point. Finally, the signal parameters are analyzed to determine whether they have reached a stable state, thereby judging the communication result.

[0024] Preferably, in step S107, the surface position data of the detached underwater buoy is obtained based on GPS positioning, and standardized positioning is generated through coordinate transformation. If the position deviation exceeds a threshold, a search path is generated through path planning, and the path nodes are optimized by integrating real-time environmental data. Navigation control parameters are extracted based on the optimization results, the heading of the search equipment is adjusted, and its operating status is dynamically updated. Finally, the search task is determined to be completed by verifying the operating status parameters, and the final task dataset is generated.

[0025] The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea described in this application has the following advantages: By introducing simulation modeling and systematic search path planning, this invention achieves scientific prediction and efficient coverage search of mooring drift areas. By constructing a simulation model and combining it with marine environmental data, the search path is systematically planned to avoid blind searching and significantly shorten search time. This invention ensures comprehensive coverage of potential areas through gridded path design and drift control, improving the probability of mooring location. The search strategy can be flexibly adjusted according to the degree of mooring drift (small-scale drift, large-scale drift, anchor breakage), enhancing the practicality and robustness of the method. By outputting control parameters (speed, heading) through simulation, intelligent control of the vessel is achieved, reducing human intervention and lowering the difficulty of operation. By evaluating the search scheme through comprehensive coefficients, the search time and communication success rate are balanced, achieving optimal resource allocation. Attached Figure Description

[0026] Figure 1 This application describes the process of a method for searching for hydrographic moorings around typical islands and reefs in the South China Sea. Figure 1 ;

[0027] Figure 2 This application describes the process of a method for searching for hydrographic moorings around typical islands and reefs in the South China Sea. Figure 2 . Detailed Implementation

[0028] The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea as described in this application includes:

[0029] like Figures 1-2 As shown in Figure S101, an elliptical search area is set with the location of the mooring buoy as the center, based on marine environmental data, and its major semi-axis direction is consistent with the historical ocean current direction.

[0030] Further, in step S101, the coordinates of the center point are obtained from the buoy deployment point, and latitude and longitude data are obtained through the Global Positioning System to determine the coordinates of the center point;

[0031] Based on marine environmental data, data preprocessing techniques are used to extract temperature, salinity, and current velocity data to obtain environmental data analysis results. If the environmental data analysis results include current velocity data, the historical ocean current direction is detected through vector analysis algorithms to determine the current direction angle.

[0032] By employing data fusion processing, combining the center point coordinates and the ocean current direction angle, the shape parameters of the elliptical region are generated, thus obtaining the region's shape parameters.

[0033] Based on the region shape parameters, an elliptical search area is generated, and the angle between the major semi-axis direction and the ocean current direction is set to determine the search area range.

[0034] By analyzing spatial positioning accuracy, the search area is adjusted and the elliptical region boundary is optimized to obtain the final search area. If the final search area boundary exceeds a preset threshold, the region is subdivided using a grid division algorithm to obtain a high-precision search grid.

[0035] Specifically, in step S101, an elliptical search area is set based on marine environmental data with the buoy deployment location as the center, and the direction of the major semi-axis is consistent with the historical ocean current direction. The specific implementation method is as follows:

[0036] First, obtain the latitude and longitude coordinates of the mooring location, assuming it to be (120.5°E, 20.3°N);

[0037] Ocean current data for the past 30 days at this location were extracted from the HYCOM marine environmental database, and the average current direction was calculated.

[0038] Assume that data analysis shows the average ocean current direction is 45° (northeast) and the speed is 0.5 m / s;

[0039] Statistical methods were used to calculate the standard deviation of the ocean current direction as 10° to ensure the reliability of the direction, and the major and minor semi-axes of the elliptical search area were set.

[0040] Based on the drift time of the mooring (assumed to be 48 hours) and the ocean current speed, the length of the major semi-axis is calculated as: 0.5m / s × 48 × 3600s = 86400m, approximately 86.4km. The minor semi-axis is set to 0.5 times the length of the major semi-axis, i.e., 43.2km, to cover the uncertainty of ocean current spread.

[0041] The center of the ellipse marks the mooring location, with its major semi-axis aligned with the ocean current direction (45°). The search area boundary is generated using the ellipse parametric equation.

[0042] x=86.4×cos(θ)×cos(45°)-43.2×sin(θ)×sin(45°);

[0043] y = 86.4 × cos(θ) × sin(45°) + 43.2 × sin(θ) × cos(45°), where θ ranges from 0 to 2π, x, y: coordinates of the boundary point of the ellipse in the Cartesian coordinate system, θ: parameter angle, 86.4: length of the major semi-axis of the ellipse, 43.2: length of the minor semi-axis of the ellipse, 45°: angle between the direction of the major axis of the ellipse and the due north direction (i.e., the historical ocean current direction);

[0044] After generating boundary points, convert them to latitude and longitude coordinates, and perform projection transformation based on the WGS84 ellipsoid model to obtain a set of points in an elliptical region centered on the mooring location;

[0045] To optimize search efficiency, the elliptical region is gridded into 5km × 5km cells using ocean visibility data (assumed to be 5km). The search priority for each cell is calculated using the following formula: Where d is the distance from the center of the unit cell to the center of the ellipse, and D is the length of the major semi-axis;

[0046] The final output includes geographic information data containing elliptical boundary point sets and grid cell priorities, which can be used by the search system.

[0047] like Figures 1-2 As shown, in step S102, a simulation environment is constructed based on the sea state and marine topography data of the search area, and a search vessel model is imported into it.

[0048] Further, in step S102, the position data of the ship's three-dimensional model is obtained from the initial simulation scene, and the position coordinates and orientation information are extracted to obtain the ship's motion parameters;

[0049] If the speed value in the ship's motion parameters exceeds the preset threshold, the ship's motion trajectory is corrected through a vector adjustment algorithm to generate optimized motion parameters.

[0050] Based on the optimized motion parameters and combined with the sea state characteristics in the three-dimensional ocean simulation, fluid dynamics simulation technology is used to calculate the interaction force between the ship and the ocean current, and obtain the ship's force data.

[0051] Mechanical distribution characteristics are extracted from the ship's force data. The stability of the ship in the simulation environment is evaluated using the finite element analysis method. The stability evaluation results are determined. If the stability evaluation results show that the ship's attitude deviation exceeds a preset threshold, the orientation parameters of the ship's three-dimensional model are adjusted using an attitude correction algorithm to generate a corrected ship model.

[0052] Based on the corrected ship model and combined with terrain information, a path planning algorithm is used to generate the optimal navigation path of the ship in the simulation environment, and the navigation path data is obtained.

[0053] Key path nodes are extracted from navigation path data, and spatial mapping technology is used to align the path nodes with the three-dimensional ocean simulation environment to generate the final simulation scene.

[0054] Based on the final simulation scenario, real-time rendering technology is used to update the dynamic display of the simulation environment model, resulting in dynamic simulation output.

[0055] Specifically, in step S102, based on the sea state and ocean topography data of the mooring location (120.5°E, 20.3°N), a simulation environment is constructed and a search vessel model is imported.

[0056] First, sea state data for the past 15 days at this location were obtained from the Copernicus global ocean physical analysis and prediction system, including wind speed, wave height and seabed topography, assuming an average wind speed of 8 m / s, a wave height of 1.5 m, an average seabed depth of 200 m, and the presence of a 50 m deep trench in some areas.

[0057] Using a three-dimensional fluid dynamics model (FVCOM) and sea state data as input, a simulation environment of 100km×100km×200m was constructed with a grid resolution of 1km×1km×10m.

[0058] The seabed topography was analyzed using Gaussian process regression interpolation to generate a continuous depth field and smooth the trench boundaries. The effects of wind and waves on the search vessel were calculated using drag coefficients, assuming a windward drag coefficient of 0.8 and a yaw angle of 5° caused by wave height.

[0059] The search vessel model is based on standard AUV parameters, with a maximum speed of 2m / s and a sensor detection radius of 3km. In the simulation, the vessel path planning adopts the A* algorithm, combined with seabed terrain obstacle avoidance, and prioritizes areas with a depth greater than 100m to reduce the risk of grounding.

[0060] The path cost function is C = 0.6 × d + 0.3 × e + 0.1 × w, where d is the path length, e is the energy consumption, and w is the wind and wave interference weight.

[0061] Assuming the search mission lasts 24 hours, the simulation system generates a set of ship trajectory points in the format (x,y,z,t) and calculates the coverage rate, finding that the covered area accounts for 85% of the target area.

[0062] The final output includes 3D simulation environment mesh data, ship trajectory point sets, and coverage statistics, stored in NetCDF format for subsequent analysis.

[0063] like Figures 1-2 As shown in S103, in the simulation model, with the probe remaining vertical and the speed below a preset threshold as constraints, the ship's speed, heading, and a series of initial search paths parallel to the major axis of the ellipse are output, and the paths cover the entire search area.

[0064] Further, in step S103, the initial position and orientation information of the ship are obtained from the sea state data of the search area. Combined with the vertical constraint of the probe, coordinate transformation technology is used to generate the standard position coordinates and orientation angle of the ship in the simulation environment, and the initial state dataset is obtained.

[0065] Based on the initial state dataset and combined with the preset speed threshold, a speed allocation algorithm is used to calculate the allowable speed range of the ship within the search area, resulting in a speed constraint dataset. If the speed values ​​in the speed constraint dataset meet the preset threshold, the heading angle is extracted from the initial state dataset, and vector mapping technology is used to generate the ship's heading control parameters, resulting in a heading control dataset.

[0066] Based on the heading control dataset and the characteristics of the ellipse's major axis, the A-path planning algorithm is used to generate a series of search paths parallel to the ellipse's major axis, covering the entire search area, thus obtaining the initial path dataset.

[0067] The path node coordinates are extracted from the initial path dataset. Combined with the sea state data from the simulation environment model, spatial interpolation technology is used to adjust the alignment between the path nodes and the sea state features to obtain the optimized path dataset.

[0068] Based on the optimized path dataset, combined with the ship's speed constraint dataset and heading control dataset, real-time simulation technology is used to update the ship's dynamic navigation state in the simulation environment, thus obtaining the dynamic navigation dataset.

[0069] The real-time position and heading changes of ships are extracted from the dynamic navigation dataset. Combined with the search area coverage requirements, a trajectory correction algorithm is used to adjust the ship's navigation path and generate the final search path dataset.

[0070] Specifically, in step S103, based on the mooring deployment location (121.0°E, 19.5°N), sea state data for the past 10 days is obtained from the global ocean database (HYCOM), including ocean current speed of 0.5m / s, wind speed of 6m / s, and wave height of 1.2m. Combined with seabed topographic data, the average depth is 250m, with local ridges reaching a depth of 80m.

[0071] A 50km×50km×250m simulation environment was constructed using the 3D Ocean Simulation Framework (ROMS), with a mesh resolution of 500m×500m×5m. The seabed topography was generated as a continuous depth field through Kriging interpolation, and the mid-ocean ridge boundaries were smoothed to ensure natural terrain transitions. The search vessel was an unmanned underwater vehicle (AUV) with a maximum speed of 1.8m / s. The probe was kept vertical, and the speed was constrained to be below 1.5m / s to ensure stability.

[0072] The search area is set as an ellipse with a major axis of 20km and a minor axis of 10km. The path planning adopts the grid-based D* algorithm to generate an initial search path parallel to the major axis. The path spacing is set to 1km to ensure that the sensor detection radius of 2.5km is covered without omission. The heading is affected by the ocean current. The ocean current yaw angle is calculated to be 3°. The heading is adjusted by vector decomposition to maintain parallelism.

[0073] The path cost function is defined as C = 0.5 × d + 0.4 × v + 0.1 × c, where d is the path length, v is the speed deviation, and c is the ocean current interference weight.

[0074] The simulation ran for 12 hours, generating a path point set in the format (x, y, z, t), which was recorded once per minute, with the speed maintained at 1.4 m / s;

[0075] The coverage rate is calculated by the ratio of the path projection area to the ellipse area, which is 92%. The output data includes the gridded simulation environment, path point set, and coverage statistics, and is stored in HDF5 format for easy subsequent processing. The entire process is implemented through automated scripts, and data processing and path planning are completed by algorithms without human intervention.

[0076] like Figures 1-2 As shown, in step S104, the search vessel is controlled to drift or propel itself along the initial search path, and attempts to communicate with the mooring releaser via the acoustic deck unit during the voyage.

[0077] Further, in step S104, path node coordinates are extracted from the initial navigation and communication dataset, and spatial mapping technology is used to generate the real-time position coordinates of the search vessel within the search area, thus obtaining a real-time position dataset.

[0078] Based on the real-time location dataset and the communication frequency range, signal detection technology is used to determine the communication signal strength between the acoustic deck unit and the buoy release device, and a signal strength dataset is obtained.

[0079] If the signal strength in the signal strength dataset is lower than a preset threshold range, then the communication frequency parameters are optimized using frequency adjustment technology to generate an optimized frequency dataset.

[0080] Based on the optimized frequency dataset and combined with the communication control parameters, the communication configuration of the acoustic deck unit is updated using data fusion technology to obtain the updated communication dataset.

[0081] The communication configuration parameters are extracted from the updated communication dataset, combined with the navigation mode dataset, and the path optimization technique is used to adjust the navigation trajectory of the search vessels, generating an adjusted path dataset.

[0082] Based on the adjusted path dataset and combined with the real-time location dataset, dynamic navigation technology is used to update the drifting or powered navigation status of the search vessels, thus obtaining a dynamic navigation dataset.

[0083] Navigation status parameters are extracted from the dynamic navigation dataset, combined with the signal strength dataset, and status assessment techniques are used to determine whether the search vessel has completed the communication task with the mooring release device, thus obtaining the task completion dataset.

[0084] Specifically, in step S104, based on the mooring deployment location (120.8°E, 19.7°N), sea state data for the past 7 days is obtained from the global ocean database (ECCO2), including ocean current speed of 0.3 m / s, wind speed of 5 m / s, and wave height of 1.0 m.

[0085] Based on seabed topographic data, the average depth is 300m, with some areas having trenches up to 100m deep.

[0086] A simulation environment of 40km×40km×300m was constructed using the 3D ocean simulation framework (MITgcm), with a mesh resolution of 400m×400m×4m. The seabed topography was generated using B-spline interpolation to create a continuous depth field, and the trench boundaries were smoothed to ensure smooth terrain transitions. The search vessel was an unmanned surface vehicle (USV) with a maximum speed of 2.0m / s, and the speed was constrained to be below 1.6m / s to ensure stable acoustic deck unit signals.

[0087] The search area is set as a rectangle with dimensions of 30km × 15km. The path planning uses the A* algorithm to generate an initial search path along the long side of the rectangle with a path spacing of 0.8km to ensure that the acoustic sensor detection radius of 2.0km is covered without omission. The heading is affected by ocean currents, and the calculated yaw angle is 2.5°. The heading is adjusted through vector decomposition to maintain path consistency.

[0088] The path cost function is defined as C = 0.6 × d + 0.3 × v + 0.1 × w, where d is the path length, v is the speed deviation, and w is the wind speed interference weight.

[0089] The simulation ran for 10 hours, generating a path point set in the format (x,y,t), which was recorded every 30 seconds. The speed was maintained at 1.5m / s. The acoustic deck unit communicated with the mooring release device using FSK modulation with a signal frequency of 200Hz. The communication success rate was recorded every 5 minutes and found to be 95%.

[0090] Coverage is calculated by the ratio of the path projection area to the rectangle area, reaching 90%. Output data includes the simulation environment mesh, path point set, and communication success rate, stored in NetCDF format. Automated scripts implement path planning and communication control.

[0091] like Figures 1-2 As shown in step S105, a comprehensive coefficient is calculated based on the communication success rate and search time to dynamically optimize the search strategy.

[0092] Further, in step S105, communication success rate and search time parameters are extracted from the task data, and data preprocessing techniques are used to generate a standardized dataset;

[0093] Normalization parameters are extracted from the standardized dataset, and combined with the preset weight allocation rules, the weighted average technique is used to calculate the comprehensive coefficient to obtain the comprehensive coefficient dataset.

[0094] The coefficient parameters are extracted from the comprehensive coefficient dataset, and combined with the current search strategy, the decision tree algorithm is used to generate a preliminary optimization strategy, resulting in a preliminary strategy dataset.

[0095] Path adjustment parameters are extracted from the initial strategy dataset. If the path adjustment parameters exceed the preset threshold, trajectory correction technology is used to update the search path and obtain the updated path dataset.

[0096] The communication frequency parameters are extracted from the initial strategy dataset. If the communication frequency parameters are lower than the preset threshold, signal optimization technology is used to adjust the communication frequency to obtain an optimized communication dataset.

[0097] Adjustment parameters are extracted from the updated path dataset and the optimized communication dataset. Data fusion technology is used to generate a comprehensive optimization strategy, resulting in the final strategy dataset.

[0098] Policy parameters are extracted from the final policy dataset, combined with real-time task data, and dynamically updated to adjust the search task execution scheme, resulting in a task execution dataset.

[0099] Specifically, in step S105, when the unmanned surface vehicle (USV) performs a search mission based on the mooring position (121.5°E, 20.2°N), the system calculates a comprehensive coefficient by analyzing the communication success rate and search time to dynamically optimize the search strategy.

[0100] The system collects the communication success rate S (percentage, ranging from 0-100) and search time T (minutes) every 5 minutes, and calculates the comprehensive coefficient. 60 is the normalized time base, ensuring that C is between 0 and 100;

[0101] If C is below 70, the system triggers strategy optimization, prioritizing the adjustment of the search grid density;

[0102] The initial grid spacing is 1.5 km. If C < 70, the system optimizes the spacing using a gradient descent algorithm (learning rate 0.01), with the following iterative formula: Calculate partial derivatives By fitting historical data, where d new : Optimized grid spacing, d old : Grid spacing before optimization Partial derivatives of the composite coefficients with respect to grid spacing;

[0103] After adjustment, the grid spacing was reduced to 1.0-1.3 km, and the vehicle moved along the new grid path at a speed of 2.0 m / s;

[0104] The system detects underwater signal strength (in dB) using acoustic sensors. If the strength is below -80 dB, it automatically increases the number of grid points from 10 to 15, evenly distributed over a 10 km × 10 km area.

[0105] Communication uses QAM modulation with a carrier frequency of 300Hz. Signal quality is recorded every 4 minutes and stored in NetCDF format, including timestamps and grid coordinates (x, y, y). i ,y i and signal strength;

[0106] The optimization process is controlled by a Python script, which calls Pandas to process the data and SciPy to perform gradient descent, ensuring that the overall coefficient C is increased to over 75 and the search coverage reaches 90%.

[0107] If the search time T exceeds 30 minutes, the system automatically reduces the speed to 1.5 m / s to reduce energy consumption, while updating the path planning, prioritizing coverage of areas with higher signal strength, generating an optimized path point set (x, y, t), and storing it in a NetCDF file.

[0108] like Figures 1-2 As shown in S106, if the actual navigation path deviates from the initial path, it is brought back by power adjustment. If the probe is not vertical, communication is suspended and multiple grid points are set on the path to attempt fixed-point communication.

[0109] Further, in step S106, path deviation data is extracted from the actual navigation path and the initial path, and the path deviation angle is calculated using path comparison technology to obtain the path deviation dataset.

[0110] Based on the path deviation dataset and combined with the power control parameters, the propulsion power of the search vessel is adjusted using power distribution technology to generate a power adjustment dataset.

[0111] Propulsion power parameters are extracted from the power adjustment dataset, and combined with the actual navigation path. The trajectory correction technique is then used to update the navigation trajectory of the search vessel, resulting in a regression path dataset.

[0112] The navigation trajectory parameters are extracted from the regression path dataset, combined with the probe attitude data, and the attitude detection technology is used to determine whether the probe is in a vertical state, thus obtaining the probe attitude dataset.

[0113] If the probe attitude dataset shows that the probe is not perpendicular, multiple fixed-point grids are generated on the regression path dataset using grid partitioning technology to obtain a fixed-point grid dataset.

[0114] Based on the fixed-point grid dataset and combined with acoustic communication parameters, signal transmission technology is used to conduct communication attempts in each fixed-point grid to obtain the fixed-point communication dataset.

[0115] Communication signal parameters are extracted from the fixed-point communication dataset, and combined with preset thresholds, signal analysis techniques are used to determine whether the communication signal has reached a stable state, thus obtaining the communication status dataset.

[0116] Specifically, in step S106, based on the mooring position (121.5°E, 20.2°N), if the actual navigation path of the unmanned surface vehicle (USV) deviates from the initial path when performing a search mission, the system restores path consistency through real-time dynamic adjustment.

[0117] First, the actual position coordinates (x′, y′) of the vehicle are acquired every 10 seconds using the Global Positioning System (GPS), and compared with the initial path point set (x, y) to calculate the deviation distance. Where D: the deviation distance between the actual navigation path and the initial path, (x′,y′): the actual position coordinates, (x,y): the coordinates of the initial path point;

[0118] If D exceeds the threshold of 0.5km, the power adjustment algorithm is triggered, using a proportional-integral-derivative (PID) controller. The proportional coefficient Kp = 0.8, the integral coefficient Ki = 0.1, and the derivative coefficient Kd = 0.05. The heading adjustment angle θ = atan2(y-y', x-x') is calculated, and the rudder angle control command is sent to the vehicle actuator to adjust the speed to 1.8m / s to quickly return to the path, while limiting the maximum rudder angle to 30° to avoid over-steer.

[0119] After power adjustment, the system verifies the deviation every 2 minutes, repeating the adjustment until D < 0.2km. If the acoustic probe attitude sensor detects that the probe deviates from the vertical angle by more than 5°, the system automatically suspends communication to avoid signal distortion and selects 5 evenly distributed grid points from the path point set, spaced 1.2km apart, to generate fixed-point communication coordinates (x... i ,y i );

[0120] At each grid point, the vehicle hovered at a speed of 0 m / s, adjusted the probe to vertical, used PSK modulation, and the signal frequency was 250 Hz. Communication was attempted every 3 minutes, and the signal strength and reception success rate were recorded. The expected success rate was 92%. The communication data was stored in HDF5 format, including grid point coordinates, attempt time, and success flag.

[0121] The entire process of path deviation adjustment and fixed-point communication is controlled by an automated script. The script is implemented in Python, calls NumPy for vector calculation, and SciPy to optimize PID parameters to ensure that the deviation convergence time is less than 15 minutes, the communication coverage reaches 88%, and the data output is an HDF5 file containing the adjusted path points (x,y,t) and communication records.

[0122] like Figures 1-2 As shown in S107, if the buoy becomes dislodged, a surface search will be conducted using the GPS positioning system.

[0123] Further, in step S107, the surface position data of the detached underwater buoy is obtained through the GPS positioning system to generate a preliminary positioning dataset;

[0124] Latitude and longitude parameters are extracted from the initial positioning dataset, and coordinate transformation technology is used to generate a standardized positioning dataset.

[0125] The location deviation parameter is extracted from the standardized location dataset. If the location deviation parameter exceeds the preset threshold, the path planning technique is used to generate a search path dataset.

[0126] Path node parameters are extracted from the search path dataset, and data fusion technology is used to combine real-time environmental data to generate an optimized path dataset.

[0127] Navigation control parameters are extracted from the optimized path dataset, and the navigation direction of the search device is adjusted through signal transmission technology to generate a navigation instruction dataset.

[0128] The command execution parameters are extracted from the navigation command dataset, and the search device's operating status is adjusted using dynamic update technology to generate a task execution dataset.

[0129] Runtime status parameters are extracted from the task execution dataset, and the completion status of the search task is determined through data verification technology to generate the final task dataset.

[0130] Specifically, in step S107, after the mooring float is dislodged, the system uses GPS positioning to start an unmanned surface vehicle (USV) to search the surface. The initial position of the mooring float is (122.0°E, 19.8°N). The GPS signal is updated every 3 minutes with an accuracy of ±5 meters.

[0131] The system first uses a Kalman filter algorithm to fuse GPS data and inertial navigation data to predict the drift trajectory of the mooring. It assumes that the drift speed is 0.3 m / s and the direction changes with the ocean current. The ocean current speed is provided by an external model and is currently 0.5 m / s, with the direction being 30° north of east.

[0132] Based on this, the system generates an initial search area of ​​8km×8km, with the center point being the latest GPS coordinates;

[0133] The USV searches along a spiral path at a speed of 1.8 m / s, with a path spacing of 0.8 km. The path points are calculated using the polar coordinate formula r = 0.4θ, where θ is the angle (radians), and the interval between each revolution is 0.4 km.

[0134] The acoustic sensor records underwater signals every 2 minutes at a frequency of 400Hz, with a detection range of 1.5km. If the signal strength is below -85dB, the system automatically switches to a high-density search mode, reducing the path spacing to 0.5km. The A* algorithm is used to plan the optimal path, taking into account the influence of ocean currents. The weight is the ratio of distance to energy consumption, calculated as W = 0.7 × D + 0.3 × E, where D is the path length (km) and E is the energy consumption (joules).

[0135] The search data is stored in HDF5 format and includes timestamps, coordinates (x, y), and signal strength.

[0136] The system analyzes the coverage rate every 10 minutes, with a target of 85%. If the coverage rate is lower than the target, it calls NumPy to calculate the signal strength gradient and adjusts the USV to move towards an area with a strength higher than -75dB.

[0137] If the search time exceeds 25 minutes, the system optimizes the flight speed using a genetic algorithm, ranging from 1.2 to 1.8 m / s, with the fitness function being: Where C represents coverage and E represents energy consumption, the optimal speed is selected after 10 iterations, and the final path point set (x,y,t) is stored in an HDF5 file for subsequent analysis.

[0138] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this application.

Claims

1. A method for searching hydrographic moorings around typical islands and reefs in the South China Sea, characterized in that, include: S101. An elliptical search area is set with the location of the mooring buoy as the center, based on marine environmental data, with its major semi-axis direction consistent with the historical ocean current direction. S102. Based on the sea condition and marine topography data of the search area, construct a simulation environment and import the search vessel model into it; S103. In the simulation model, with the constraints of keeping the probe vertical and the speed below a preset threshold, the ship's speed, heading, and a series of initial search paths parallel to the major axis of the ellipse are output, and the paths cover the entire search area. S104. Control the search vessel to drift or propel itself along the initial search path, and attempt to communicate with the buoy releaser via the acoustic deck unit during the voyage. S105. Calculate a comprehensive coefficient based on communication success rate and search time, and dynamically optimize the search strategy, including: The communication success rate and search time parameters are extracted from the task data, and a standardized dataset is generated using data preprocessing techniques. Normalization parameters are extracted from the standardized dataset, and combined with the preset weight allocation rules, the weighted average technique is used to calculate the comprehensive coefficient to obtain the comprehensive coefficient dataset. The coefficient parameters are extracted from the comprehensive coefficient dataset, and combined with the current search strategy, the decision tree algorithm is used to generate a preliminary optimization strategy, resulting in a preliminary strategy dataset. Path adjustment parameters are extracted from the initial strategy dataset. If the path adjustment parameters exceed the preset threshold, trajectory correction technology is used to update the search path and obtain the updated path dataset. The communication frequency parameters are extracted from the initial strategy dataset. If the communication frequency parameters are lower than the preset threshold, signal optimization technology is used to adjust the communication frequency to obtain an optimized communication dataset. Adjustment parameters are extracted from the updated path dataset and the optimized communication dataset. Data fusion technology is used to generate a comprehensive optimization strategy, resulting in the final strategy dataset. The strategy parameters are extracted from the final strategy dataset, combined with real-time task data, and the search task execution plan is adjusted using dynamic update technology to obtain the task execution dataset. S106. If the actual navigation path deviates from the initial path, adjust it back through power. If the probe is not vertical, suspend communication and set multiple grid points on the path to attempt fixed-point communication. S107. If the buoy becomes dislodged, a surface search will be conducted using the GPS positioning system.

2. The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea according to claim 1, characterized in that, The step of setting the elliptical search area includes: The center coordinates of the mooring deployment point are obtained. Based on marine environmental data, preprocessing is performed to extract temperature, salinity, and current velocity data. If current velocity data exists, historical ocean current direction is calculated through vector analysis. Combining the center coordinates and ocean current direction, shape parameters of an elliptical region are generated. An elliptical search area is generated based on the shape parameters, and its major semi-axis direction is aligned with the ocean current direction. The region boundary is optimized through spatial positioning accuracy analysis. If the region exceeds the preset range, a grid division algorithm is used for further subdivision.

3. The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea according to claim 1, characterized in that, The steps for constructing the simulation environment include: The system acquires a 3D model of the search vessel and its motion parameters. If the speed exceeds a threshold, the motion trajectory is corrected using a vector adjustment algorithm. Combined with sea state characteristics, the system uses hydrodynamic simulation to calculate the ship-sea interaction force. The system evaluates the vessel's stability using finite element analysis. If the attitude deviation exceeds the limit, the model orientation is corrected. The system generates the optimal navigation path based on the terrain information and aligns it to the simulation environment through spatial mapping. The system uses real-time rendering technology to output dynamic simulation results.

4. The method for searching hydrographic moorings around typical islands and reefs in the South China Sea according to claim 1, characterized in that, The steps for outputting the initial search path include: The initial state of the ship is obtained based on sea state data. The allowable speed range is calculated based on the speed threshold. The heading angle is extracted and the heading control parameters are generated. A series of search paths parallel to the major axis of the ellipse are generated using a path planning algorithm. The path nodes are adjusted by spatial interpolation to align with the sea state features. Combined with speed and heading constraints, the final navigation path is generated through real-time simulation.

5. The method for searching for hydrographic moorings around typical islands and reefs in the South China Sea according to claim 1, characterized in that, The surface search includes: The GPS positioning system is used to obtain the surface position data of the detached underwater mooring. The position data is then converted into coordinates to generate a standardized position. If the position deviation exceeds a preset threshold, a search path is generated through path planning. The path nodes are optimized by integrating real-time environmental data. The course and operating status of the search equipment are adjusted according to the optimization results. The search task is determined to be completed by verifying the operating status parameters.

Images