An underwater unmanned vehicle formation dynamic area search system

Abstract

The application provides an underwater unmanned underwater vehicle formation dynamic area search system and relates to the field of unmanned underwater vehicle cooperative operation, and comprises the following steps: generating a virtual marine environment according to seabed depth information and land obstacle distribution information; arranging virtual particles in a preset search area to form a gravitational field, and giving the virtual particles an initial mass; constructing a detection pressure field according to the detection activities of the unmanned underwater vehicle formation, determining the probability distribution of the target appearance position based on the reverse distribution of the detection pressure field, and generating a counter target according to the probability distribution; dynamically updating the mass value of the virtual particles according to the detection coverage range of the unmanned underwater vehicle formation; and calculating the gravity and repulsion that each unmanned underwater vehicle in the unmanned underwater vehicle formation receives according to the mass distribution of the virtual particles in the gravitational field, and updating the position and heading of each unmanned underwater vehicle based on the gravity and repulsion. The application improves the efficiency and authenticity of dynamic target search.

Description

Technical Field

[0001] This invention relates to the field of collaborative operation technology for underwater unmanned submersibles, and in particular to a dynamic area search system for underwater unmanned submersible formations. Background Technology

[0002] In military scenarios such as blockading key maritime areas and waterways, and anti-submarine warfare, underwater unmanned vehicle swarms need to rapidly search for and detect dynamic targets with certain movement patterns. Traditional fixed-pattern search methods such as "bow" and "Z" shaped movements have significant shortcomings when facing dynamic targets: on the one hand, these methods cannot adaptively adjust the search path according to the target distribution characteristics, resulting in low search efficiency; on the other hand, fixed search patterns lack a memory mechanism for already searched areas, easily leading to repeated waste of search resources.

[0003] Gravity search algorithms, as optimization algorithms based on the physical law of gravity, have shown great application potential in the field of path planning. This algorithm explores the search space by simulating the gravitational interaction between point masses, possessing global optimization capabilities and distributed cooperative characteristics. However, traditional gravity search algorithms still have limitations when applied to multi-agent dynamic region search tasks: they lack a memory and labeling mechanism for searched regions, making it impossible to effectively achieve dynamic allocation of search resources; they do not fully consider the actual physical characteristics of the search entities, such as motion inertia and velocity constraints; and their detection modeling of dynamic targets is relatively simplified, failing to reflect the influence of the target's motion state on the detection probability.

[0004] Chinese invention patent application number 201910861689.X discloses a method for planning the planar trajectory of underwater unmanned vehicles in formation. This method uses an improved artificial potential field method and a fast expanding random tree method for global trajectory planning, and employs a velocity obstacle method for local collision avoidance planning. This patent mainly solves the path planning and obstacle avoidance problems in formation navigation, and can generate feasible trajectories in both static and dynamic obstacle environments. However, when facing dynamic area searches, it cannot dynamically adjust and avoid unsearched and searched areas. Summary of the Invention

[0005] In view of this, the present invention provides a dynamic area search system for underwater unmanned submersible formations, which can solve the problem that existing technologies cannot achieve effective collaborative detection by multiple agents, and improve the efficiency and realism of dynamic target search.

[0006] The technical solution of this invention is implemented as follows: This invention provides a dynamic area search system for underwater unmanned submersible formations, comprising: The environment simulation module is used to generate a virtual marine environment based on seabed depth information and land obstacle distribution information; The gravitational field construction module is used to deploy virtual particles within a preset search area to form a gravitational field, and to assign initial mass to the virtual particles based on the terrain information of their location. The target generation module is used to construct a detection pressure field based on the detection activities of the unmanned underwater vehicle formation, determine the probability distribution of the target's location based on the inverse distribution of the detection pressure field, and generate adversarial targets based on the probability distribution. The dynamic adjustment module is used to dynamically update the mass value of virtual particles according to the detection coverage of the unmanned underwater vehicle formation, reducing the mass of virtual particles in the covered areas and increasing the mass of virtual particles in the uncovered areas. The path planning module is used to calculate the gravitational and repulsive forces acting on each unmanned underwater vehicle (UUV) in the UUV formation based on the mass distribution of virtual particles in the gravitational field and the gravitational search algorithm, and to update the position and heading of each UUV based on the gravitational and repulsive forces.

[0007] Based on the above technical solutions, preferably, the specific steps of the environment simulation module include: The original depth matrix is ​​generated based on the sea area parameters and complexity parameters. The original depth matrix includes seabed depth data and obstacle distribution data. An interpolation model is established based on the original depth matrix. The interpolation model is used to respond to queries at any coordinate position and return the depth information and obstacle determination results at that position.

[0008] Based on the above technical solutions, preferably, the target generation module includes: The pressure field construction unit is used to construct the detection pressure field based on the spatial distribution of the unmanned underwater vehicle formation and historical search activities. The computing unit is used to calculate the target generation attraction at each location based on the detected pressure field and terrain features; The location sampling unit is used to sample and determine the generation location of the adversarial target based on the probability distribution of the target generation attractiveness.

[0009] Based on the above technical solutions, preferably, the pressure field construction unit specifically includes: Iterate through all unmanned underwater vehicles (UUVs), calculate the coverage contribution of each UUV based on its current position and detection parameters, and sum the coverage intensities of all UUVs to obtain the instantaneous detection intensity: ; ; in For the first An unmanned underwater vehicle is positioned. Coverage intensity, For this position to the th The distance of an unmanned underwater vehicle For the first The detection radius of an unmanned underwater vehicle. Indicates the index of the unmanned underwater vehicle. Indicates the total number of unmanned underwater vehicles; Calculate historical detection intensity using a time-based recursive method: ; in For position At any moment Historical detection intensity, This represents the instantaneous detection intensity at the current moment. For time step, The pressure decay time constant, This indicates the historical detection intensity at the previous moment; The detection pressure field is obtained by summing the instantaneous detection intensity and the historical detection intensity.

[0010] Based on the above technical solutions, preferably, the computing unit specifically includes: According to location The seabed depth information determines the static depth preference factor at that location. ; The pressure suppression factor is obtained by performing a nonlinear mapping on the detected pressure field value. The target generation attractiveness is calculated based on the static depth preference factor and the pressure inhibition factor: ; in, Indicates position At time t, the target generates attractiveness. Indicates position The detected pressure field value at time t, This indicates the reference pressure value.

[0011] Based on the above technical solutions, preferably, the position sampling unit specifically includes: The search area is divided into discrete grids, and the target generation attractiveness at each grid location is calculated. ,in, This represents the x-coordinate of the i-th grid. Represents the ordinate of the j-th grid; The probability distribution is obtained by normalizing the attraction at all grid locations; Candidate generation positions are selected by weighted random sampling based on probability distribution; The candidate generated locations are validated to determine whether they meet the conditions of not being located on obstacles and not being within the immediate detection range of the unmanned underwater vehicle. If the condition is not met, then weighted random sampling is performed again according to the probability distribution; If the conditions are met, the location will be determined as the location where the adversary target is generated.

[0012] Based on the above technical solutions, preferably, the dynamic adjustment module specifically includes: At each time step, traverse all virtual particles in the gravitational field and obtain the current position of each virtual particle; Calculate the spatial distance between each virtual point and all unmanned underwater vehicles (UUVs), and determine whether the virtual point is within the detection coverage area based on the spatial distance and the detection parameters of the corresponding UUVs: For virtual particles identified as being within the coverage area, their mass value is reset to the initial minimum value and their time counter is cleared. For virtual particles determined to be outside the coverage area, their time counters are incremented and their mass values ​​are updated based on the accumulated time.

[0013] Based on the above technical solutions, preferably, the update formula for the dynamic quality is: ; in For the first A virtual particle at time... Dynamic quality, Let be the initial mass of the virtual particle. The quality growth coefficient, This is the cumulative time since the virtual particle was last detected or since the initial moment that it has not been detected.

[0014] Based on the above technical solutions, preferably, the path planning module specifically includes: The gravity calculation unit is used to traverse virtual particles with non-zero mass in the gravitational field, calculate the gravitational effect of each virtual particle on the unmanned underwater vehicle according to the gravity model, and sum them to obtain the total gravity vector. The repulsion calculation unit is used to traverse other unmanned underwater vehicles in the formation, calculate the repulsive force exerted by other unmanned underwater vehicles on the current unmanned underwater vehicle according to the repulsion model, and sum them to obtain the total repulsion vector; The heading update unit is used to vector synthesize the total gravitational vector and the total repulsive vector to obtain the resultant force vector. It then combines the direction of the resultant force vector with the current navigation direction using inertial weights to determine a new navigation direction. Based on the new navigation direction and motion parameters, it updates the position of the unmanned underwater vehicle.

[0015] More preferably, the detection parameter is an adaptive detection radius, and the calculation steps for the adaptive detection radius are as follows: Based on the speed of the opposing target Exposure time threshold and static detection radius Calculate the basic equivalent detection radius, where the detection time constraint is the shortest time that the adversary target needs to stay within the detection range; Based on the standard deviation of the direction of movement of the opposing target Calculate the directional coverage factor: ; in, Indicates the direction coverage factor. For reference angle, Sensitivity index; Based on the adversary target in the historical time window average angular velocity within Computational trajectory complexity correction factor: ; in, This represents the trajectory complexity correction factor. Used as reference angular velocity; The adaptive equivalent detection radius is calculated based on the orientation coverage factor, trajectory complexity correction factor, and basic equivalent detection radius: ; in, Indicates the adaptive equivalent detection radius. This represents the trajectory complexity weighting coefficient.

[0016] The present invention has the following advantages over the prior art: (1) Dynamic region search is constructed by using environmental simulation module, gravitational field construction module, target generation module, dynamic adjustment module and path planning module. The virtual mass mass is updated in real time according to the detection coverage, which guides the unmanned underwater vehicle formation to naturally gather in the undetected area, avoiding repeated search and improving the efficiency and realism of dynamic target search. (2) By constructing a detection pressure field and determining the target generation probability based on the inverse distribution of the pressure field, adversarial targets are made to appear first in areas with weak detection coverage, which improves the adversarial nature of the test scenario and enables a more comprehensive evaluation of the robustness of the search strategy under adverse conditions. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a dynamic area search system for underwater unmanned submersible formations according to the present invention. Figure 2 This is a flowchart illustrating the construction process of the detection pressure field in the dynamic area search system for underwater unmanned submersible formations according to the present invention. Figure 3 This is a block diagram of virtual mass update for a dynamic area search system for underwater unmanned submersible formations according to the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, the present invention provides a dynamic area search system for underwater unmanned submersible formations, comprising: The environment simulation module is used to generate a virtual marine environment based on seabed depth information and land obstacle distribution information; The gravitational field construction module is used to deploy virtual particles within a preset search area to form a gravitational field, and to assign initial mass to the virtual particles based on the terrain information of their location. The target generation module is used to construct a detection pressure field based on the detection activities of the unmanned underwater vehicle formation, determine the probability distribution of the target's location based on the inverse distribution of the detection pressure field, and generate adversarial targets based on the probability distribution. The dynamic adjustment module is used to dynamically update the mass value of virtual particles according to the detection coverage of the unmanned underwater vehicle formation, reducing the mass of virtual particles in the covered areas and increasing the mass of virtual particles in the uncovered areas. The path planning module is used to calculate the gravitational and repulsive forces acting on each unmanned underwater vehicle (UUV) in the UUV formation based on the mass distribution of virtual particles in the gravitational field and the gravitational search algorithm, and to update the position and heading of each UUV based on the gravitational and repulsive forces.

[0021] In one embodiment of the present invention, the specific steps of the environment simulation module include: The original depth matrix is ​​generated based on the sea area parameters and complexity parameters. The original depth matrix includes seabed depth data and obstacle distribution data. An interpolation model is established based on the original depth matrix. The interpolation model is used to respond to queries at any coordinate position and return the depth information and obstacle determination results at that position.

[0022] Understandably, based on the user-defined sea area size parameter AREA_SIZE_KM and seabed complexity level parameter SEABED_COMPLEXITY_LEVEL, a Gaussian feature overlay algorithm is used to generate seabed topographic undulations. This algorithm can simulate common natural topographic features such as ridges and basins. The Gaussian feature overlay algorithm uses several Gaussian distribution functions placed at different locations, each with a random center position, amplitude, and standard deviation, to overlay multiple Gaussian functions to form a depth field with undulations. Based on the obstacle density level parameter OBSTACLE_DENSITY_LEVEL, a Berlin noise algorithm is used to generate the distribution of land obstacles, producing obstacle textures that are naturally random yet continuous. The Berlin noise algorithm generates fractal textures by overlaying multiple harmonic noises, and determines whether a location is land based on a threshold of the noise value. The generated original depth matrix is ​​a two-dimensional array that stores the depth values ​​of discrete sampling points and land markers.

[0023] In one embodiment of the present invention, the interpolation model is used to respond to queries at arbitrary coordinate positions and return the depth information and obstacle determination results for that position. It creates a two-dimensional spline interpolator based on discrete sampling points in the original depth matrix. This interpolator employs a bicubic spline interpolation method, enabling smooth interpolation at any continuous coordinate position. When other modules query a certain coordinate point... When the environmental information is obtained, the interpolator returns the seabed depth value at that location and a Boolean result indicating whether the location is land.

[0024] In one embodiment of the invention, the gravitational field construction module delineates a parallelogram as a responsibility area on the nautical chart based on the user-input search region vertex coordinates SEARCH_AREA_VERTICES_KM. Within this parallelogram region, virtual particles are uniformly distributed in a grid pattern according to the density defined by NUM_GSA_PARTICLES_PER_SIDE to form a gravitational field. Each virtual particle is traversed, and its seabed depth information is obtained by calling the query interface of the environment simulation module. An initial mass is assigned to each virtual particle according to a preset depth-mass mapping rule. For example, virtual particles located in the 500-1000 meter depth region are assigned a higher initial mass, while those in shallow water are assigned a lower initial mass. Virtual particles determined to be located on land have their mass set to zero and do not generate gravitational force in subsequent calculations. The gravitational field construction module provides a spatial framework and initial value distribution for the dynamic search.

[0025] In one embodiment of the present invention, the target generation module includes: The pressure field construction unit is used to construct the detection pressure field based on the spatial distribution of the unmanned underwater vehicle formation and historical search activities. The computing unit is used to calculate the target generation attraction at each location based on the detected pressure field and terrain features; The location sampling unit is used to sample and determine the generation location of the adversarial target based on the probability distribution of the target generation attractiveness.

[0026] This invention constructs a detection pressure field and determines the target generation probability based on the inverse distribution of the pressure field, so that adversarial targets preferentially appear in areas with weak detection coverage, thereby improving the adversarial nature of the test scenario and enabling a more comprehensive evaluation of the robustness of the search strategy under adverse conditions.

[0027] like Figure 2 As shown, in one embodiment of the present invention, the pressure field construction unit specifically includes: Iterate through all unmanned underwater vehicles (UUVs), calculate the coverage contribution of each UUV based on its current position and detection parameters, and sum the coverage intensities of all UUVs to obtain the instantaneous detection intensity: ; ; in For the first An unmanned underwater vehicle is positioned. Coverage intensity, For this position to the th The distance of an unmanned underwater vehicle For the first The detection radius of an unmanned underwater vehicle. Indicates the index of the unmanned underwater vehicle. Indicates the total number of unmanned underwater vehicles; Historical detection intensity is calculated using a time-recursive method, whereby historical detection intensity reflects the cumulative effect of search activities over time and the spatial memory characteristics: ; in For position At any moment Historical detection intensity, This represents the instantaneous detection intensity at the current moment. For time step, The pressure decay time constant, This indicates the historical detection intensity at the previous moment; The detection pressure field is obtained by summing the instantaneous detection intensity and the historical detection intensity. : ; in This is the instantaneous intensity weighting coefficient. The historical intensity weighting coefficients satisfy the following conditions: .

[0028] Understandable, exponential decay factor Controlling the fading rate of historical detection intensity over time, decay time constant A larger value indicates a longer retention time for historical memories, while a smaller decay time constant indicates a faster fading of historical memories. (Instantaneous detection intensity contribution) The current coverage intensity is multiplied by the time step and then added to the historical detection intensity, thus realizing the time integral effect of detection intensity.

[0029] In one embodiment of the present invention, The value is 0.6. The value is set to 0.4, which makes the contribution of the current detection activity to the pressure field slightly greater than the historical cumulative contribution.

[0030] In one embodiment of the present invention, the detection parameter is an adaptive detection radius, and the calculation steps of the adaptive detection radius are as follows: Based on the speed of the opposing target Exposure time threshold and static detection radius Calculate the basic equivalent detection radius, where the detection time constraint is the shortest time that the adversary target needs to stay within the detection range; Based on the standard deviation of the direction of movement of the opposing target Calculate the directional coverage factor: ; in, Indicates the direction coverage factor. For reference angle, Sensitivity index; Based on the adversary target in the historical time window average angular velocity within Computational trajectory complexity correction factor: ; in, This represents the trajectory complexity correction factor. Used as reference angular velocity; The adaptive equivalent detection radius is calculated based on the orientation coverage factor, trajectory complexity correction factor, and basic equivalent detection radius: ; in, Indicates the adaptive equivalent detection radius. This represents the trajectory complexity weighting coefficient.

[0031] This invention overcomes the limitation of traditional equivalent radius formulas, which only consider the ideal situation of the target crossing in a straight line, by adopting an adaptive detection radius. This allows the detection judgment to be dynamically adjusted according to the target's motion state, thereby improving the realism and credibility of the simulation.

[0032] In one embodiment of the present invention, the computing unit specifically includes: According to location The seabed depth information determines the static depth preference factor at that location. ; The pressure suppression factor is obtained by nonlinear mapping of the detected pressure field value. ; The target generation attractiveness is calculated based on the static depth preference factor and the pressure inhibition factor. The target generation attractiveness represents the degree of attraction of a location to adversarial targets. The higher the attractiveness, the more likely a location is to become a target generation site. ; in, Indicates position At time t, the target generates attractiveness. Indicates position The detected pressure field value at time t, This indicates the reference pressure value.

[0033] Understandably, when the detected pressure field value approaches infinity, the hyperbolic tangent function tanh approaches 1, and the pressure suppression factor approaches 0, indicating that the location is completely unattractive to counter-targets. When the detected pressure field value is 0, the hyperbolic tangent function tanh is 0, and the pressure suppression factor is 1, indicating that the location is not suppressed by the detected pressure. The nonlinear characteristics of the hyperbolic tangent function tanh significantly enhance the suppression effect in high-pressure regions, while maintaining a relatively low suppression effect in low-pressure regions, thus forming a sensitive response mechanism to the detected pressure.

[0034] The static depth preference factor provides a basic preference distribution, causing adversarial targets to tend to appear in areas that conform to the depth preference. The pressure suppression factor dynamically adjusts the attractiveness based on the distribution of detection pressure, reducing the attractiveness of areas with high detection pressure (i.e., areas frequently covered by UAV formations) and maintaining a high level of attractiveness in areas with low detection pressure (i.e., weak search areas). Multiplying the static depth preference factor by the pressure suppression factor ensures that adversarial targets preferentially appear in locations that conform to both the depth preference and are in weak search areas, simulating the situational awareness and evasion behavior of adversarial targets. This adversarial target generation mechanism enhances the adversarial nature of the test scenario and enables a more comprehensive evaluation of the robustness of the search strategy under adverse conditions.

[0035] In one embodiment of the present invention, the position sampling unit specifically includes: The search area is divided into discrete grids, and the target generation attractiveness at each grid location is calculated. ,in, This represents the x-coordinate of the i-th grid. Represents the ordinate of the j-th grid; The probability distribution is obtained by normalizing the attraction at all grid locations: ; Candidate generation positions are selected by weighted random sampling based on probability distribution; The candidate generated locations are validated to determine whether they meet the conditions of not being located on obstacles and not being within the immediate detection range of the unmanned underwater vehicle. If the condition is not met, then weighted random sampling is performed again according to the probability distribution; If the conditions are met, the location will be determined as the location where the adversary target is generated.

[0036] The sampling method can employ either the roulette wheel algorithm or the inverse transform sampling method. In the roulette wheel algorithm, a random number between 0 and 1 is generated, and the probability values ​​are accumulated sequentially according to the grid positions. When the accumulated value exceeds the random number, the current grid position is selected. In the inverse transform sampling method, a cumulative distribution function is constructed, and the corresponding position is found based on the random number.

[0037] like Figure 3 As shown, in one embodiment of the present invention, the dynamic adjustment module specifically includes: At each time step, traverse all virtual particles in the gravitational field and obtain the current position of each virtual particle: Calculate the spatial distance between each virtual point and all unmanned underwater vehicles (UUVs), and determine whether the virtual point is within the detection coverage area based on the spatial distance and the detection parameters of the corresponding UUVs: For virtual particles identified as being within the coverage area, their mass value is reset to the initial minimum value and their time counter is cleared. For virtual particles determined to be outside the coverage area, their time counters are incremented and their mass values ​​are updated based on the accumulated time.

[0038] Understandably, the mass value of the covered virtual particle Reset to initial minimum quality Its undetected cumulative time Reset to zero. The initial minimum mass can be set to a small positive value or zero, indicating that the area has just been explored and its attractiveness to unmanned underwater vehicles is minimized. The accumulated unexplored time of uncovered virtual particles is increased by one time step. ,Right now Then, based on the accumulated time, its quality value is updated according to the preset quality growth model.

[0039] In one embodiment of the present invention, the update formula for dynamic quality is: ; in For the first A virtual particle at time... Dynamic quality, Let be the initial mass of the virtual particle. The quality growth coefficient, This is the cumulative time since the virtual particle was last detected or since the initial moment that it has not been detected.

[0040] In one embodiment of the present invention, the path planning module specifically includes: The gravity calculation unit is used to traverse virtual particles with non-zero mass in the gravitational field, calculate the gravitational effect of each virtual particle on the unmanned underwater vehicle according to the gravity model, and sum them to obtain the total gravitational vector. : ; ; ; ; in It is the gravitational constant; Indicates the total number of virtual particles; Represents the unit direction vector. Represents the gravitational vector; Represents the x-coordinate of the virtual particle. The x-coordinate represents the unmanned underwater vehicle. Represents the ordinate of the virtual particle. The vertical coordinate represents the unmanned underwater vehicle. This represents the distance between the virtual point mass and the unmanned underwater vehicle; The repulsion calculation unit is used to traverse other unmanned underwater vehicles (UUVs) in the formation, calculate the repulsive force exerted by other UUVs on the current UUV based on the repulsion model, and sum them to obtain the total repulsion vector. : ; ; ； ; in, It is the repulsive constant. The repulsive force attenuation index is the repulsive force direction away from the opponent's position, that is, the unit direction vector from the opponent's UAV to the current UAV. Represents the repulsive force vector. This represents the unit direction vector of the current unmanned underwater vehicle. This represents the x-coordinate of the k-th unmanned underwater vehicle. This represents the ordinate of the k-th unmanned underwater vehicle. This represents the distance between the s-th unmanned underwater vehicle and the k-th unmanned underwater vehicle; The heading update unit is used to vector synthesize the total gravitational and total repulsive force vectors to obtain the resultant force vector. It then combines the direction of the resultant force vector with the current navigation direction using inertial weights to determine a new navigation direction. Based on this new navigation direction and motion parameters, the position of the unmanned underwater vehicle is updated. ; ; in, Represents the resultant force vector. Indicates a new direction of navigation. Indicates the inertia weighting coefficient. This represents the unit vector indicating the current direction of travel of the unmanned underwater vehicle. Represents the resultant force vector The unit vector of the resultant force direction obtained after normalization; Understandably, this is due to the new course of navigation and the cruising speed of the unmanned underwater vehicle. Calculate the displacement of the unmanned underwater vehicle within the time step Δt, and the position update amount is: The new location of the unmanned underwater vehicle is During the location update process, boundary detection is performed. If the predicted new location exceeds the search area boundary or is located on land, the navigation direction is adjusted to avoid it.

[0041] Understandably, if the x-coordinate of the new location is less than the left boundary of the search area, the x-coordinate is set to the left boundary value, and the lateral component of the heading is inverted. The y-coordinate is handled similarly. Obstacle detection queries the environment simulation module to determine if the new location is land. If it is land, the location update is rejected, the original location is maintained, and the heading is adjusted to avoid obstacles.

[0042] Evasion methods could include rotating the course by a certain angle, such as 90 degrees, and then trying to move again, or adopting a strategy of gliding along the boundary.

[0043] This invention constructs a dynamic region search through an environment simulation module, a gravitational field construction module, a target generation module, a dynamic adjustment module, and a path planning module. It updates the virtual mass mass in real time based on the detection coverage, so that the gravity of undetected areas continuously increases and the gravity of detected areas automatically decreases, forming an adaptive memory of the search space. This guides the unmanned underwater vehicle formation to naturally gather in undetected areas, avoiding repeated searches. Compared with traditional fixed-pattern search methods, it significantly improves the efficiency and realism of dynamic target search.

[0044] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic area search system for underwater unmanned submersible formations, characterized in that, include: The environment simulation module is used to generate a virtual marine environment based on seabed depth information and land obstacle distribution information; The gravitational field construction module is used to deploy virtual particles within a preset search area to form a gravitational field, and to assign initial mass to the virtual particles based on the terrain information of their location. The target generation module is used to construct a detection pressure field based on the detection activities of the unmanned underwater vehicle formation, determine the probability distribution of the target's location based on the inverse distribution of the detection pressure field, and generate adversarial targets based on the probability distribution. The dynamic adjustment module is used to dynamically update the mass value of virtual particles according to the detection coverage of the unmanned underwater vehicle formation, reducing the mass of virtual particles in the covered areas and increasing the mass of virtual particles in the uncovered areas. The path planning module is used to calculate the gravitational and repulsive forces acting on each unmanned underwater vehicle (UUV) in the UUV formation based on the mass distribution of virtual particles in the gravitational field and the gravitational search algorithm, and to update the position and heading of each UUV based on the gravitational and repulsive forces.

2. The underwater unmanned submersible formation dynamic area search system as described in claim 1, characterized in that: The specific steps of the environment simulation module include: The original depth matrix is ​​generated based on the sea area parameters and complexity parameters. The original depth matrix includes seabed depth data and obstacle distribution data. An interpolation model is established based on the original depth matrix. The interpolation model is used to respond to queries at any coordinate position and return the depth information and obstacle determination results at that position.

3. The underwater unmanned submersible formation dynamic area search system as described in claim 1, characterized in that: The target generation module includes: The pressure field construction unit is used to construct the detection pressure field based on the spatial distribution of the unmanned underwater vehicle formation and historical search activities. The computing unit is used to calculate the target generation attraction at each location based on the detected pressure field and terrain features; The location sampling unit is used to sample and determine the generation location of the adversarial target based on the probability distribution of the target generation attractiveness.

4. The underwater unmanned submersible formation dynamic area search system as described in claim 3, characterized in that: The pressure field construction unit specifically includes: Iterate through all unmanned underwater vehicles (UUVs), calculate the coverage contribution of each UUV based on its current position and detection parameters, and sum the coverage intensities of all UUVs to obtain the instantaneous detection intensity: ； ； in For the first An unmanned underwater vehicle is positioned. Coverage intensity, For this position to the th The distance of an unmanned underwater vehicle For the first The detection radius of an unmanned underwater vehicle. Indicates the index of the unmanned underwater vehicle. Indicates the total number of unmanned underwater vehicles; Calculate historical detection intensity using a time-based recursive method: ； in For position At any moment Historical detection intensity, This represents the instantaneous detection intensity at the current moment. For time step, The pressure decay time constant, This indicates the historical detection intensity at the previous moment; The detection pressure field is obtained by summing the instantaneous detection intensity and the historical detection intensity.

5. The underwater unmanned submersible formation dynamic area search system as described in claim 4, characterized in that: The computing unit specifically includes: According to location The seabed depth information determines the static depth preference factor at that location. ; The pressure suppression factor is obtained by performing a nonlinear mapping on the detected pressure field value. The target generation attractiveness is calculated based on the static depth preference factor and the pressure inhibition factor: ； in, Indicates position At time t, the target generates attractiveness. Indicates position The detected pressure field value at time t, This indicates the reference pressure value.

6. The underwater unmanned submersible formation dynamic area search system as described in claim 5, characterized in that: The location sampling unit specifically includes: The search area is divided into discrete grids, and the target generation attractiveness at each grid location is calculated. ,in, This represents the x-coordinate of the i-th grid. Represents the ordinate of the j-th grid; The probability distribution is obtained by normalizing the attraction at all grid locations; Candidate generation positions are selected by weighted random sampling based on probability distribution; The candidate generated locations are validated to determine whether they meet the conditions of not being located on obstacles and not being within the immediate detection range of the unmanned underwater vehicle. If the condition is not met, then weighted random sampling is performed again according to the probability distribution; If the conditions are met, the location will be determined as the location where the adversary target is generated.

7. The underwater unmanned submersible formation dynamic area search system as described in claim 1, characterized in that: The dynamic adjustment module specifically includes: At each time step, traverse all virtual particles in the gravitational field and obtain the current position of each virtual particle; Calculate the spatial distance between each virtual point and all unmanned underwater vehicles (UUVs), and determine whether the virtual point is within the detection coverage area based on the spatial distance and the detection parameters of the corresponding UUVs: For virtual particles identified as being within the coverage area, their mass value is reset to the initial minimum value and their time counter is cleared. For virtual particles determined to be outside the coverage area, their time counters are incremented and their mass values ​​are updated based on the accumulated time.

8. The underwater unmanned submersible formation dynamic area search system as described in claim 7, characterized in that: The update formula for the dynamic quality is: ； in For the first A virtual particle at time... Dynamic quality, Let be the initial mass of the virtual particle. The quality growth coefficient, This is the cumulative time since the virtual particle was last detected or since the initial moment that it has not been detected.

9. The underwater unmanned submersible formation dynamic area search system as described in claim 1, characterized in that: The path planning module specifically includes: The gravity calculation unit is used to traverse virtual particles with non-zero mass in the gravitational field, calculate the gravitational effect of each virtual particle on the unmanned underwater vehicle according to the gravity model, and sum them to obtain the total gravity vector. The repulsion calculation unit is used to traverse other unmanned underwater vehicles in the formation, calculate the repulsive force exerted by other unmanned underwater vehicles on the current unmanned underwater vehicle according to the repulsion model, and sum them to obtain the total repulsion vector; The heading update unit is used to vector synthesize the total gravitational vector and the total repulsive vector to obtain the resultant force vector. It then combines the direction of the resultant force vector with the current navigation direction using inertial weights to determine a new navigation direction. Based on the new navigation direction and motion parameters, it updates the position of the unmanned underwater vehicle.

10. The underwater unmanned submersible formation dynamic area search system as described in claim 9, characterized in that: The detection parameter is an adaptive detection radius, and the calculation steps for the adaptive detection radius are as follows: Based on the speed of the opposing target Exposure time threshold and static detection radius Calculate the basic equivalent detection radius, where the detection time constraint is the shortest time that the adversary target needs to stay within the detection range; Based on the standard deviation of the direction of movement of the opposing target Calculate the directional coverage factor: ； in, Indicates the direction coverage factor. For reference angle, Sensitivity index; Based on the adversary target in the historical time window average angular velocity within Computational trajectory complexity correction factor: ； in, This represents the trajectory complexity correction factor. Used as reference angular velocity; The adaptive equivalent detection radius is calculated based on the orientation coverage factor, trajectory complexity correction factor, and basic equivalent detection radius: ； in, Indicates the adaptive equivalent detection radius. This represents the trajectory complexity weighting coefficient.

Images