An underwater target search and rescue strategy determination method, device, equipment and storage medium

Abstract

The application discloses a kind of underwater target search and rescue strategy determination method, device, equipment and storage medium.It includes: obtaining the marine environment information of target sea area, the motion characteristics of underwater target are calculated according to marine environment information, and dynamic model is constructed;Based on Monte Carlo algorithm, the target trajectory of underwater target is predicted in combination with dynamic model;Grid probability model is constructed, search strategy is primarily determined based on grid probability model, and probability distribution is dynamically updated by Bayes theory to optimize search strategy;Establish cost-benefit analysis model, and the secondary optimization of optimized search strategy is carried out.The uncertainty of environment and target motion is fully considered, and the reliability of trajectory prediction is improved.Grid probability model is constructed, and search strategy is optimized by dynamically updating probability distribution by Bayes theory, so that search is targeted and efficiency is improved.Through secondary optimization of strategy, optimal allocation of resources is realized, search and rescue effect is guaranteed while cost is controlled, and economy and practicability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ship search and rescue, and in particular to a method and device for determining a search and rescue strategy for an underwater target, and a storage medium. BACKGROUND

[0002] In the field of underwater target search and rescue, it is crucial to determine the position of a lost or sunken underwater vehicle such as a submersible or submarine and to carry out search and rescue operations in a timely and efficient manner. The underwater environment is complex and variable, and both the marine environmental information and the motion characteristics of the underwater target can significantly affect the determination of the target position, making underwater target search and rescue a highly challenging task. Search and rescue personnel need to consider a variety of factors in order to develop an effective search and rescue strategy.

[0003] In the prior art, there are some methods and techniques for underwater target search and rescue. For example, some methods use marine environmental information to construct a simple model to predict the target trajectory, and some methods use a probability model to determine the priority of the search area.

[0004] Existing methods often do not adequately consider the uncertainty of the marine environment and the complexity of the motion characteristics of the underwater target, resulting in low accuracy of the predicted target trajectory, which in turn affects the development of subsequent search strategies. In addition, the existing technology lacks an effective mechanism for dynamically updating the probability distribution, making the allocation of search resources unreasonable, inefficient, and costly, and difficult to meet the needs of actual search and rescue tasks. SUMMARY

[0005] The present application provides a method and device for determining a search and rescue strategy for an underwater target, which predicts the target trajectory by constructing a dynamic model, optimizes the search strategy using a grid probability model and Bayesian theory, and performs secondary optimization combined with cost-benefit analysis, thereby solving the technical problems of low search and rescue efficiency and high cost caused by environmental uncertainty and limited search resources in underwater target search and rescue.

[0006] According to an aspect of the present application, a method for determining a search and rescue strategy for an underwater target is provided, the method comprising:

[0007] obtaining marine environmental information of a target sea area, calculating the motion characteristics of the underwater target based on the marine environmental information, and constructing a dynamic model;

[0008] predicting the target trajectory of the underwater target based on the dynamic model using a Monte Carlo algorithm;

[0009] constructing a grid probability model, preliminarily determining a search strategy based on the grid probability model, and dynamically updating the probability distribution using Bayesian theory to optimize the search strategy;

[0010] establishing a cost-benefit analysis model to perform secondary optimization on the optimized search strategy;

[0011] This includes calculating the motion characteristics of underwater targets based on marine environmental information and constructing a dynamic model, including:

[0012] By using correlation analysis algorithms, we can screen out the correlation factors in marine environmental information that are related to the movement of underwater targets.

[0013] Determine the neutral buoyancy point of the underwater target and obtain the seabed position. Compare the neutral buoyancy point with the seabed position to determine the target's motion state and output scene labels, including scenes of drifting with ocean currents and scenes of gradually sinking.

[0014] Deterministic values ​​of ocean current characteristics are extracted from marine environmental information, and a motion reference system is established based on the assumption of a fixed uniform mass distribution.

[0015] The motion equation template is called based on the scene label, and the correlation factor and motion reference system are substituted into the motion equation template to generate the motion equation.

[0016] By acquiring environmental uncertainty parameters and integrating the equations of motion with these parameters, a dynamic model is generated.

[0017] Optionally, based on the Monte Carlo algorithm and combined with a dynamic model, the target trajectory of an underwater target is predicted, including: determining the state vector and state transition equation of the underwater target's motion through the dynamic model; acquiring the position and velocity data of the underwater target before its failure as the initial state, and acquiring the preset initial state uncertainty covariance and process noise covariance; randomly generating a set of particles representing the initial state based on the initial state and the initial state uncertainty covariance; calling the dynamic model at time steps, and superimposing random perturbations on each particle based on the process noise covariance to calculate the predicted state of each particle; statistically analyzing each predicted state to generate a probability distribution map, a mean trajectory, and a confidence region; calculating the position probability of the underwater target through density estimation based on the probability distribution map, the mean trajectory, and the confidence region, and determining the target trajectory based on the highest value of the position probability.

[0018] Optionally, a grid probability model is constructed, including: obtaining the location of the search and rescue area, determining the grid division range based on the location of the underwater target before the accident and the location of the search and rescue area; discretizing the grid division range into individual cells, calculating the number of cells based on preset rules, defining the label and center point coordinates of each cell, and completing the area rasterization; assigning an initial existence probability to each cell based on the target trajectory, combined with Manhattan distance and Poisson distribution, introducing a time dimension to correct the probability, and completing the construction of the grid probability model.

[0019] Optionally, the search strategy is preliminarily determined based on the grid probability model, including: determining the to-be-searched grids in the grid probability model, and extracting real-time probability values of the to-be-searched grids; the to-be-searched grids are sorted in descending order of the real-time probability values to generate a priority sequence; the current search resource number and the device search efficiency are obtained, the current coverage grid number is determined based on the current search resource number and the device search efficiency, and the corresponding to-be-searched grid is selected from the priority sequence as the search strategy according to the current coverage grid number.

[0020] Optionally, the probability distribution is dynamically updated through the Bayesian theory to optimize the search strategy, including: obtaining the grid search result after the current search time interval ends, and extracting the current probability value of the corresponding grid, wherein the grid search result includes successfully finding the target and not finding the target; when the grid search result is not finding the target, the current probability value of the corresponding grid is corrected to 0 through the Bayesian formula, the probabilities of the remaining to-be-searched grids are recalculated, and the updated grid probability distribution is generated; based on the updated grid probability distribution, the to-be-searched grids are sorted in descending order of probability again to generate an optimized search strategy.

[0021] Optionally, a cost-benefit analysis model is established to perform secondary optimization on the optimized search strategy, including: obtaining resource consumption data and benefit data of the optimized search strategy, and determining evaluation indexes; a cost-benefit calculation model is constructed according to the evaluation indexes; the cost-benefit ratio is calculated by substituting the optimized search strategy into the cost-benefit calculation model; the strategies with a cost-benefit ratio lower than a preset threshold are adjusted in resource allocation until the preset threshold is met.

[0022] According to another aspect of the present application, an underwater target search and rescue strategy determination device is provided, which comprises:

[0023] A dynamics model construction module is configured to obtain marine environment information of a target sea area, calculate motion characteristics of an underwater target based on the marine environment information, and construct a dynamics model.

[0024] A target trajectory prediction module is configured to predict a target trajectory of the underwater target based on a Monte Carlo algorithm and in combination with the dynamics model.

[0025] A search strategy optimization module is configured to construct a grid probability model, preliminarily determine a search strategy based on the grid probability model, and dynamically update a probability distribution through the Bayesian theory to optimize the search strategy.

[0026] A secondary optimization module is configured to establish a cost-benefit analysis model and perform secondary optimization on the optimized search strategy.

[0027] The dynamics model construction module is specifically configured to filter out correlation factors related to the motion of the underwater target in the marine environment information through a correlation analysis algorithm.

[0028] Determine the neutral buoyancy point of the underwater target, and obtain the seabed position, compare the neutral buoyancy point with the seabed position to determine the target motion state, and output the scene label, wherein the scene label includes a drift with ocean current scene and a gradual sinking scene;

[0029] Extract the certainty value of the current characteristics in the marine environment information, and establish a motion reference system according to the fixed uniform distribution mass assumption;

[0030] According to the scene label, a motion equation template is called, and the correlation factor and the motion reference system are substituted into the motion equation template to generate a motion equation;

[0031] Obtain the environmental uncertainty parameters, integrate the motion equation and the environmental uncertainty parameters to generate a dynamics model.

[0032] According to another aspect of the present application, an electronic device is provided, which comprises:

[0033] At least one processor;

[0034] And a memory connected in communication with the at least one processor;

[0035] Wherein the memory stores a computer program capable of being executed by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the underwater target search strategy determination method according to any one of the embodiments of the present application.

[0036] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to implement the underwater target search strategy determination method according to any one of the embodiments of the present application when executed.

[0037] The technical scheme of the embodiment of the present application provides accurate environment and motion basis for subsequent underwater target trajectory prediction and search strategy formulation by constructing a dynamics model. The uncertainty of the environment and target motion is fully considered to generate a plurality of possible trajectories, which provides rich reference for determining the area where the target may exist, and improves the reliability of trajectory prediction. The search area is quantized into a grid and assigned a probability, which can intuitively and clearly determine the search priority, and the dynamic updating of the probability distribution according to the Bayesian theory can continuously optimize the search strategy as the search progresses, thereby improving the targeting and efficiency of the search. The cost-benefit analysis model is established to secondarily optimize the strategy, realize optimal allocation of resources, ensure search effect while controlling cost, and improve economic efficiency and practicality.

[0038] It is to be understood that the details set forth in the description contained herein do not limit the scope of the application. Other embodiments of the application will be readily apparent to those skilled in the art from the description herein. With reference to the drawings, embodiments of the application are herein described. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0040] Figure 1 is a flow chart of a method for determining a search and rescue strategy of an underwater target according to an embodiment of the present application;

[0041] Figure 2 is a flow chart of another method for determining a search and rescue strategy of an underwater target according to another embodiment of the present application;

[0042] Figure 3 is a structural schematic diagram of a device for determining a search and rescue strategy of an underwater target according to an embodiment of the present application;

[0043] Figure 4 is a structural schematic diagram of an electronic device for implementing a method for determining a search and rescue strategy of an underwater target according to an embodiment of the present application. DETAILED DESCRIPTION

[0044] In order to make the technical personnel in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, but not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort should be within the scope of protection of the present application.

[0045] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and in the above-described drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0046] Embodiment one

[0047] Figure 1 A flowchart of a method for determining an underwater target search and rescue strategy is provided for the first embodiment of the present application. The present embodiment can be applicable to a target search and rescue scene. The method can be executed by an underwater target search and rescue strategy determination device, which can be realized in the form of hardware and / or software, and can be configured in a computer controller. As shown in Figure 1 the method includes:

[0048] S110, acquiring marine environment information of a target sea area, calculating the motion characteristics of the underwater target according to the marine environment information, and constructing a dynamic model.

[0049] The target sea area refers to a specific marine area where the underwater target is lost or lost, and is the range of search and rescue operations. The marine environment information refers to the key environmental data of the target sea area, including ocean currents, seawater density and geographical position, etc., which is used to analyze the influence of the environment on the motion of the underwater target. The underwater target refers to a water-borne object that is lost or lost, such as a submersible, a submarine, etc., which is the object of search and rescue. The motion characteristics refer to the motion law of the underwater target in the marine environment, including drift speed, sinking trend, etc., which are determined by gravity, buoyancy, ocean currents, etc. The dynamic model refers to a mathematical model constructed based on the motion characteristics of the target, which is used to describe the change of the motion state of the underwater target and is the basis for trajectory prediction.

[0050] Optionally, the motion characteristics of the underwater target are calculated according to the marine environment information, and a dynamic model is constructed, including: through a correlation analysis algorithm, relevant factors in the marine environment information related to the motion of the underwater target are screened out; the neutral buoyancy point of the underwater target is determined, and the seabed position is obtained, and the neutral buoyancy point is compared with the seabed position to determine the target motion state, and a scene label is output, wherein the scene label includes a drifting with ocean current scene and a gradually sinking scene; the certainty value of the ocean current characteristics in the marine environment information is extracted, and a motion reference system is established according to the fixed uniform distribution mass assumption; according to the scene label, a motion equation template is called, the relevant factors and the motion reference system are substituted into the motion equation template to generate a motion equation; the environmental uncertainty parameters are obtained, the motion equation and the environmental uncertainty parameters are integrated to generate a dynamic model.

[0051] The correlation analysis algorithm can be a Pearson correlation coefficient algorithm, which calculates the correlation between each environmental information and the target motion parameter, removes redundant information with very low correlation, such as correlation coefficient absolute value less than 0.1, and selects relevant factors, which are factors such as ocean current velocity and seawater density gradient that directly affect target motion. The neutral buoyancy point is the position where the gravity and buoyancy of the underwater target are in balance. The theoretical neutral buoyancy point depth of the underwater target needs to be calculated first by the design parameters of the target, such as the volume, material density and ballast weight of the underwater target, and then the actual seabed depth of the target sea area is obtained by the sonar detection method. Then the two are compared. When the neutral buoyancy point depth is less than the seabed depth, i.e. the neutral buoyancy point is above the seabed, it means that the target gravity and buoyancy are balanced, and it will slowly move under the action of ocean current, and at this time the drifting with ocean current scene label is output. When the neutral buoyancy point depth is greater than the seabed depth, i.e. the neutral buoyancy point is below the seabed, the target gravity is greater than the buoyancy, and it will gradually sink to the seabed, and at this time the gradually sinking scene label is output.

[0052] Specifically, the current characteristics include flow velocity, flow direction, etc. Due to the fluctuation of the current in the actual marine environment, the deterministic values of the current characteristics are extracted, for example, the average flow velocity and fixed flow direction of the current obtained through the ocean observation station, buoy data or numerical simulation during the target sinking period are taken as the deterministic values. Then, a motion reference system is established according to the assumption of uniform distribution of mass, the mass of the underwater target is assumed to be uniformly distributed, the center of mass coincides with the geometric center, the position at the moment of target sinking is taken as the coordinate origin, the horizontal direction parallel to the sea level is set as x and y axes, and the vertical direction perpendicular to the sea level is set as z axis, and a rectangular coordinate system is constructed as a motion reference system. Then the controller will call the motion equation template according to the scene label and generate the motion equation. For the drifting with the current scene, the driving effect of the current on the horizontal motion of the target is reflected through the motion equation template, and the vertical motion is ignored. For the gradual sinking scene, the motion equation template will contain both the sinking motion in the vertical direction and the drifting motion in the horizontal direction. After calling the corresponding template, the previously screened related factors, such as the current velocity, are substituted into the horizontal direction term, the seawater density affects the buoyancy and is substituted into the vertical direction term, and the established motion reference system is substituted into the template, to generate the motion equation describing the motion law of the target in the current scene.

[0053] Further, the environmental uncertainty parameters mainly include the process noise covariance and the initial state uncertainty covariance matrix. The process noise covariance is set based on the prior knowledge of the target maneuvering ability and the estimation of environmental disturbance. The stronger the target maneuvering ability and the more intense the turbulence, the larger the process noise covariance, representing the stronger random disturbance not considered by the model. The initial state uncertainty covariance matrix is determined according to the error range of the target sinking position and velocity measurement data. The larger the measurement error, the larger the diagonal element value of the initial state uncertainty covariance matrix, representing the lower confidence in the initial state. The uncertainty parameters are integrated into the generated motion equation to finally form a dynamic model that can describe the deterministic motion law of the target and reflect the influence of environmental uncertainty. This model can be directly used for trajectory prediction in the subsequent Monte Carlo algorithm.

[0054] S120, based on the Monte Carlo algorithm, the target trajectory of the underwater target is predicted in combination with the dynamic model.

[0055] Among them, the Monte Carlo algorithm is a probability and statistics method based on random sampling, which simulates the possible initial state and motion trajectory of the target by generating a large number of particle samples, and overcomes the influence of environmental uncertainty on prediction. The target trajectory refers to the possible motion path of the underwater target from sinking to the predicted moment, which is obtained through point cloud generation and statistical analysis by the Monte Carlo algorithm, including the most likely central trajectory and confidence region.

[0056] Optionally, based on the Monte Carlo algorithm, the target trajectory of the underwater target is predicted in combination with the kinetic model, including: determining a state vector and a state transition equation of the underwater target motion through the kinetic model; acquiring position and velocity data of the underwater target before the accident as an initial state, and acquiring a preset initial state uncertainty covariance and process noise covariance; generating a particle set representing the initial state randomly according to the initial state and the initial state uncertainty covariance; calling the kinetic model at a time step, and superimposing random disturbances on each particle based on the process noise covariance to calculate the predicted state of each particle; performing statistics on each predicted state to generate a probability distribution graph, a mean trajectory and a confidence region; based on the probability distribution graph, the mean trajectory and the confidence region, calculating the position probability of the underwater target through density estimation, and determining the target trajectory according to the highest value of the position probability.

[0057] wherein the kinetic model has determined the motion law of the underwater target, and the model can define a state vector of the underwater target at time , which is in the form of , wherein represents the state vector of the underwater target at time , , and correspond to three-dimensional position coordinates of the target, , and are corresponding three-dimensional velocity components, and the kinetic model can comprehensively describe the motion state of the target at a certain time. At the same time, according to the motion law in the kinetic model, a two-dimensional constant velocity model is used to establish a state transition equation, which is expressed by the following formula (1):

[0058] ; (1)

[0059] wherein represents the state vector of the underwater target at time , is a time step, is process noise for reflecting random disturbances not considered in the model, is a state transition matrix, and the matrix form is expressed by the following formula (2): ; (2)

[0060] wherein is a time step. The initial state includes position and velocity data of the underwater target before the accident. The initial state uncertainty covariance is a preset covariance matrix, which quantifies the confidence level of the initial state estimation, and the size of the elements in the matrix is related to the error range of the initial position and velocity measurement. The larger the error is, the larger the corresponding element value is. The process noise covariance Also, the preset parameter is set according to prior knowledge of the target maneuvering ability, such as the maximum acceleration of the target and the estimation of the marine environment interference, The greater the value, the lower the confidence in the target motion model, and the faster the divergence of the subsequent predicted trajectory.

[0061] Specifically, according to the determined initial state and the initial state uncertainty covariance , a random sampling method is used to generate particles, for example =10000 sample points, and each particle corresponds to a possible initial state. The sampling follows the normal distribution rule, that is Each particle , i=1, 2, …, obeys the normal distribution of . Then, the predicted state of each particle can be calculated according to the time step. First, set the prediction time , divide the time into multiple consecutive time steps, that is . For each time step, for each particle in the particle set, call the state transition equation in the dynamic model to perform state prediction. In the prediction process, based on the preset process noise covariance , a sample is randomly extracted from the Gaussian distribution as the process noise of the particle, and is superimposed into the calculation process to obtain the predicted state of each particle at the current time step. Since the random disturbance superimposed by each particle is different, different particles will form divergent predicted trajectories, thereby reflecting the uncertainty of the target motion.

[0062] Further, after calculating the predicted state of all particles at each time step, the position coordinates of all particles are summarized. Plotting the position of all particles at each time step in three-dimensional space will form a point cloud that continuously expands over time, that is, a probability distribution map, which can intuitively show the spatial area where the target may exist at different time points. The mean trajectory is obtained by calculating the average of the position coordinates of all particles at each time step, representing the most likely motion path of the target. The confidence region is determined according to the preset confidence level, such as 95%, and is obtained by finding the minimum elliptical region containing all particles at the confidence level. Finally, using the kernel density estimation method, the probability density of the target appearing at any position at any future time is calculated. In the calculation process, the trend of the mean trajectory and the range of the confidence region will be fully considered to correct the probability density of different positions, ensuring the accuracy of the results. Then, traverse all the probability density values of the positions, find the position point with the highest probability density, and connect these position points with the highest probability in the consecutive time steps in turn to form the path, which is the final determined target trajectory of the underwater target.

[0063] S130, construct a grid probability model, preliminarily determine a search strategy based on the grid probability model, and dynamically update the probability distribution based on Bayesian theory to optimize the search strategy.

[0064] The grid probability model refers to a model that discretizes the search and rescue area into a plurality of small unit grids and assigns a target existence probability to each grid. The probability distribution is constructed based on the Poisson distribution, with the highest probability in the center grid and decaying with distance. The search strategy refers to the execution plan of the search and rescue action, including search area priority, path planning and device deployment. Bayesian theory is a statistical method based on prior probability and observation to update posterior probability, used to dynamically adjust the grid probability during the search process to focus on high-probability areas. The probability distribution refers to the existence probability distribution of the target at each position in the search and rescue area, initially determined based on trajectory prediction and continuously updated based on Bayesian theory to provide a basis for search strategy optimization.

[0065] Optionally, the grid probability model is constructed, including: obtaining the position of the search and rescue area, determining the grid division range based on the position of the underwater target before the accident and the position of the search and rescue area; discretizing the grid division range into unit grids, calculating the number of grids based on a preset rule, defining the labels and center point coordinates of each grid, and completing the area gridding; assigning an initial existence probability to each unit grid according to the target trajectory, combining Manhattan distance and Poisson distribution, and introducing time dimension to correct the probability, and completing the construction of the grid probability model.

[0066] Specifically, first, the specific position information of the search and rescue area is obtained, and then the overall range of grid division is determined in combination with the position of the underwater target before the accident. For example, taking the position of the target before the accident as the reference point, the search and rescue area is set as a reasonable spatial range covering the target that may drift or sink, and the range needs to be adjusted in combination with the previously predicted target trajectory confidence area to ensure that the target may exist in the area, avoid missing the target due to too small range, and avoid increasing unnecessary search cost due to too large range. Then, the area is gridded, and the determined grid division range is discretized into multiple smaller cells. The number of grids is calculated according to the preset rule, which refers to the maximum length of the grid horizontal axis and the size of a single grid to determine the total number of grids through corresponding calculation. Then, the labels of each grid are defined, and the integral function is used to determine the label of the grid to which any point in the grid division range belongs in combination with the x and y coordinates of the point to distinguish different cells. At the same time, according to the label and coordinate range of each grid, the coordinates of the center point of each grid are derived, and the gridding of the search and rescue area is completed, so that the originally continuous search and rescue area becomes a grid system composed of multiple discrete cells.

[0067] Further, the controller determines the grid corresponding to the position with the highest probability in the trajectory as the center grid, assuming that the underwater target is most likely to be located in the predicted center grid, and the probability of locating it decreases as the distance from the grid increases. When the distance between the current grid and the center grid is set as l, the probability of finding the missing underwater target is represented by the following formula (3):

[0068] ; (3)

[0069] wherein, P (x, y) represents the existence probability of the underwater target in the current grid, W (x, y) represents the weight factor of the current grid, P0 represents the initial probability of finding the missing underwater target in the center grid. The spatial probability decay is modeled by Poisson distribution, wherein the intensity parameter controls the steepness of the probability decline, and the farther the cell is from the center grid, the lower the initial existence probability according to the Poisson distribution rule. However, due to the randomness of the marine environment, the uncertainty of the actual position of the submersible increases. Therefore, it can be assumed that the ratio of the probability is integrated with time. Then, the intensity parameter of the Poisson distribution is obtained, which is shown in the following formula (4):

[0070] ; (4)

[0071] wherein, P0 represents the intensity parameter of the Poisson distribution, T represents the actual search and rescue time, Tmin represents the shortest necessary time for the rescue ship to prepare, is a constant directly affecting the initial rescue success rate. Since the randomness of the marine environment, the uncertainty of the actual position of the underwater target will increase over time, so the time dimension is introduced to modify the probability, assuming that the ratio of the probability is related to the time integral, combining the shortest necessary time for the rescue ship to prepare and the constant affecting the initial rescue success rate, the initial existence probability of each cell is modified, and the final existence probability of each cell is obtained after modification, and the construction of the grid probability model is completed.

[0072] Optionally, the search strategy is preliminarily determined based on the grid probability model, including: determining the to-be-searched grid in the grid probability model, and extracting the real-time probability value of each to-be-searched grid; sorting each to-be-searched grid in the order from high to low according to the real-time probability value, and generating a priority sequence; obtaining the current search resource number and the device search efficiency, determining the current coverage grid number based on the current search resource number and the device search efficiency, and selecting the corresponding to-be-searched grid from the priority sequence as the search strategy according to the current coverage grid number.

[0073] Specifically, the controller will explicitly search all the grids in the grid probability model, then extract the real-time existence probability value of each to-be-searched grid from the grid probability model, then compare the real-time probability value of each to-be-searched grid extracted, and arrange the grid with high probability value in front and the grid with low probability value in back, forming a sequence of search priority, and the grid with high priority means that the target exists with a higher probability and should be searched first. The search resource number includes the number of ships, devices and other search resources, and the device search efficiency refers to the number of grids that can be effectively searched by the device in unit time and other related indicators. Based on the current search resource number and the device search efficiency, the current coverage grid number is determined, that is, through calculation, the number of grids that can be covered simultaneously by the current search resources in unit time combined with the search efficiency of the device. Finally, according to the current coverage grid number, the corresponding to-be-searched grid is selected from the priority sequence as the search strategy, for example, if it is determined that 5 grids can be covered, 5 grids are selected from the front of the priority sequence, and the selected grid is the area to be searched first in the current search strategy. By determining the search strategy, the area with high target existence probability can be searched first under limited search resources, and the efficiency and success rate of search and rescue can be improved.

[0074] Optionally, the probability distribution is dynamically updated by Bayesian theory to optimize the search strategy, including: obtaining the grid search result after the end of the current search time interval, extracting the current probability value of the corresponding grid, wherein the grid search result includes successful discovery of the target and no discovery of the target; when the grid search result is no discovery of the target, the current probability value of the corresponding grid is corrected to 0 by the Bayesian formula, the probabilities of the remaining to-be-searched grids are recalculated, and the updated grid probability distribution is generated; based on the updated grid probability distribution, the to-be-searched grids are reordered in descending order of probability, and the optimized search strategy is generated.

[0075] Specifically, first, the grid search result after the end of the current search time interval is obtained, the time interval is the necessary time to complete a single area search, the search result is divided into two cases of successful discovery of the target and no discovery of the target, and the current probability value of the corresponding grid is extracted, the current probability value is the target existence probability recorded in the grid probability model before search. When the grid search result is no discovery of the target, the current probability value of the corresponding grid is corrected to 0 by the Bayesian formula, because according to the Bayesian theory, no target is found in the grid, which means that the probability of the target existing in the grid is 0. Then, the probabilities of the remaining to-be-searched grids are recalculated, that is, the sum of the probabilities of the remaining to-be-searched grids is adjusted to 1, thereby generating the updated grid probability distribution, so that the probabilities of the remaining grids more accurately reflect the possibility of the target existing. Finally, based on the updated grid probability distribution, the to-be-searched grids are reordered in descending order of probability, and the optimized search strategy is generated, so that the subsequent search can be preferentially concentrated in the grid area with high probability, thereby improving the efficiency and pertinence of the search and rescue.

[0076] S140, a cost-benefit analysis model is established to perform secondary optimization on the optimized search strategy.

[0077] The cost-benefit analysis model is an evaluation model that comprehensively considers the performance and cost of the search and rescue equipment, quantifies the performance indicators and cost indicators of the equipment by the entropy weight method, and calculates the cost-benefit score. The secondary optimization refers to further adjusting the search scheme after optimizing the search strategy based on the probability model, in combination with the cost-benefit analysis result, to control the cost while ensuring the success rate of the search and rescue.

[0078] The technical scheme of the embodiment of the present application provides an accurate environment and motion basis for subsequent underwater target trajectory prediction and search strategy formulation by constructing a dynamic model. The uncertainty of the environment and target motion is fully considered, a plurality of possible trajectories are generated, rich reference for determining the area where the target may exist is provided, and the reliability of trajectory prediction is improved. The search and rescue area is quantized into a grid and is assigned a probability, the search priority can be intuitively and clearly determined, the probability distribution is dynamically updated according to the Bayesian theory, the search strategy can be continuously optimized with the progress of the search and rescue, and the search is improved in pertinence and efficiency. The cost-benefit analysis model is established to perform secondary optimization on the strategy, optimal allocation of resources is realized, the search and rescue effect is ensured while the cost is controlled, and the economy and practicability are improved.

[0079] Embodiment two

[0080] Figure 2 A flowchart of a search strategy determination method for an underwater target provided by the second embodiment of the present application, the embodiment adds the specific process of establishing a cost-benefit analysis model to perform secondary optimization on the optimized search strategy on the basis of the first embodiment. The specific content of steps S210-S230 is substantially the same as that of steps S110-S130 in the first embodiment, and therefore will not be described herein again. As shown in the figure, the method comprises the following steps. Figure 2

[0081] S210, acquire marine environment information of a target sea area, calculate motion characteristics of the underwater target according to the marine environment information, and construct a dynamic model.

[0082] Optionally, the motion characteristics of the underwater target are calculated according to the marine environment information, and the dynamic model is constructed, which comprises the following steps: relevant analysis algorithm is used to screen out associated factors related to the motion of the underwater target in the marine environment information; the neutral buoyancy point of the underwater target is determined, and the position of the seabed is acquired; the neutral buoyancy point is compared with the position of the seabed to determine the motion state of the target, and a scene label is output, wherein the scene label comprises a scene of drifting with ocean current and a scene of gradually sinking; the certainty value of the current characteristics in the marine environment information is extracted, and a motion reference system is established according to the fixed uniform distribution mass assumption; the motion equation template is called according to the scene label, the associated factors and the motion reference system are substituted into the motion equation template to generate a motion equation; the environmental uncertainty parameters are acquired, the motion equation and the environmental uncertainty parameters are integrated, and a dynamic model is generated.

[0083] S220, predict the target trajectory of the underwater target based on a Monte Carlo algorithm and in combination with the dynamic model.

[0084] ​Optionally, the target trajectory of the underwater target is predicted based on a Monte Carlo algorithm in combination with a dynamic model, including: determining a state vector and a state transition equation of the underwater target motion through the dynamic model; obtaining position and velocity data of the underwater target before the accident as an initial state, and obtaining a preset initial state uncertainty covariance and a process noise covariance; randomly generating a particle set representing the initial state according to the initial state and the initial state uncertainty covariance; calling the dynamic model at a time step, and superimposing random disturbances on each particle based on the process noise covariance to calculate the predicted state of each particle; performing statistics on each predicted state to generate a probability distribution graph, a mean trajectory and a confidence region; and calculating the position probability of the underwater target based on the probability distribution graph, the mean trajectory and the confidence region through density estimation, and determining the target trajectory according to the highest value of the position probability.

[0085] In S230, a grid probability model is constructed, a search strategy is preliminarily determined based on the grid probability model, and a probability distribution is dynamically updated through Bayesian theory to optimize the search strategy.

[0086] Optionally, the grid probability model is constructed, including: obtaining a search and rescue area position, determining a grid division range based on the position of the underwater target before the accident and the search and rescue area position; discretizing the grid division range into unit cells, calculating the number of grids based on a preset rule, defining the labels and center point coordinates of the grids, and completing the area gridding; according to the target trajectory, combining the Manhattan distance and the Poisson distribution to assign an initial existence probability to each unit cell, introducing a time dimension to correct the probability, and completing the construction of the grid probability model.

[0087] Optionally, the search strategy is preliminarily determined based on the grid probability model, including: determining the to-be-searched grids in the grid probability model, and extracting the real-time probability values of each to-be-searched grid; sorting each to-be-searched grid in descending order of real-time probability value to generate a priority sequence; obtaining the current search resource number and the equipment search efficiency, determining the current covered grid number based on the current search resource number and the equipment search efficiency, and selecting the corresponding to-be-searched grid from the priority sequence as the search strategy according to the current covered grid number.

[0088] Optionally, the probability distribution is dynamically updated through Bayesian theory to optimize the search strategy, including: obtaining the grid search result after the current search time interval ends, and extracting the current probability value of the corresponding grid, wherein the grid search result includes successfully discovering the target and not discovering the target; when the grid search result is not discovering the target, correcting the current probability value of the corresponding grid to 0 through the Bayesian formula, and re-normalizing the probabilities of the remaining to-be-searched grids to generate an updated grid probability distribution; reordering the to-be-searched grids in descending order of probability based on the updated grid probability distribution to generate an optimized search strategy.

[0089] S240. Obtain resource consumption and benefit data of the optimized search strategy, and determine the evaluation indicators.

[0090] Resource consumption data includes individual costs such as purchase cost, maintenance cost, and transportation cost of search equipment, while benefit data includes performance indicators such as equipment search efficiency and target discovery probability. Evaluation indicators are divided into cost indicators and benefit indicators.

[0091] S250. Construct a cost-benefit calculation model based on each evaluation indicator.

[0092] Specifically, first define the original matrix. This includes various performance and cost indicators of the equipment involved. Then, according to standardized formulas... The standardized matrix is ​​obtained and used to calculate the probability matrix, as shown in the following formula (5):

[0093] (5)

[0094] in, This represents the raw data for the i-th type of equipment under the j-th evaluation metric. This represents the standardized data of the i-th type of equipment under the j-th indicator. This represents the total number of devices or search schemes participating in the evaluation. The standardized formula is used to eliminate differences in the dimensions of different indicators, ensuring that each indicator can be compared horizontally. The probability matrix is ​​shown in the following formula (6):

[0095] (6)

[0096] in, This represents the standardized probability value of the i-th type of equipment under the j-th evaluation index, used to eliminate the dimensional differences between different indicators, so that each indicator can be compared horizontally to assess its contribution to cost-effectiveness. This represents the standardized data for the i-th type of equipment under the j-th indicator. This represents the total number of devices or search schemes participating in the evaluation. Then, the entropy value of each indicator is calculated using the following formula (7):

[0097] (7)

[0098] in, Let represent the entropy value of the j-th evaluation indicator, used to measure the information discrimination of this indicator, ranging from [0,1]. The larger the entropy value, the smaller the difference between different devices on this indicator, and the less effective information it can provide for cost-benefit assessment; conversely, the lower the entropy value, the higher the discrimination and the greater the information value. This represents the standardized probability value of the i-th type of equipment under the j-th evaluation index. This represents the total number of devices or search schemes participating in the evaluation. Then, the weights of each indicator are calculated based on the entropy value, using the following formula (8):

[0099] (8)

[0100] in, This represents the weight of the j-th evaluation indicator. Let m represent the entropy value of the j-th evaluation indicator, and m be the number of indicators. The coefficient of variation is the difference coefficient of the indicator. The larger the coefficient of variation, the higher the weight, reflecting the greater the impact of the indicator on cost-benefit assessment. Then, the equipment benefit is calculated using the following formula (9):

[0101] (9)

[0102] in, Indicates the benefit of device i. This represents the weight of the j-th evaluation indicator, where m is the number of indicators. This represents the standardized probability value of the i-th type of equipment under the j-th evaluation index.

[0103] Then calculate the total cost of the equipment using the following formula (10):

[0104] (10)

[0105] in, Let K represent the total cost of equipment i, and K represent the total cost required for the equipment, including purchase cost, maintenance cost, and transportation cost, etc. Display device The individual cost. This ultimately forms the cost-benefit calculation model, the core of which is the cost-benefit score formula, as shown in formula (11):

[0106] (11)

[0107] in, This represents the cost-benefit score for each piece of equipment. This represents the total cost of device i. This represents the benefit of device i.

[0108] S260. Substitute the optimized search strategy into the cost-benefit calculation model to calculate the cost-benefit ratio.

[0109] Specifically, the optimized search strategy includes selected device combination and resource input scheme corresponding to the search path, each resource consumption data and benefit data involved in the strategy is substituted into the above model, the weighted comprehensive benefit and total resource cost of the strategy are calculated first, and then the cost-benefit ratio of the strategy is calculated through the cost-benefit ratio formula. The larger the cost-benefit ratio is, the higher the search and rescue benefit brought by unit cost is.

[0110] S270, adjusting resource allocation for the strategy with a cost-benefit ratio lower than the preset threshold until the preset threshold is met.

[0111] Specifically, when the calculated cost-benefit ratio is lower than the preset threshold, the resource allocation needs to be adjusted. For example, replacing the device with a lower cost-benefit score with a device with a higher cost-benefit score, such as reducing the input proportion of the device with high cost and low benefit, increasing the number of the device with low cost and high benefit; or optimizing the search path to reduce the transportation cost of the device and shorten the search time; or adjusting the index weight. After each adjustment, the cost-benefit ratio is calculated by substituting the model again, and the iteration is repeated until the cost-benefit ratio is higher than the preset threshold, the secondary optimization of the search strategy is completed, and the maximization of the economy of resource input is realized while meeting the success rate requirement of search and rescue.

[0112] The technical scheme of the embodiment of the application provides accurate data basis for cost-benefit analysis by obtaining resource consumption and benefit data of the optimized search strategy and determining evaluation indexes. The cost-benefit calculation model is constructed according to the evaluation indexes, so that the quantitative analysis of cost and benefit is supported by scientific tools; the cost-benefit ratio is calculated by substituting the strategy into the model, so that the economy of the strategy can be judged intuitively. By adjusting the resource allocation of the substandard strategy until it meets the standard, the optimal allocation of resources can be realized while ensuring the search and rescue effect, the search and rescue cost can be effectively controlled, and the economy and practicality of search and rescue can be improved.

[0113] Embodiment three

[0114] Figure 3 A structure schematic diagram of a kind of underwater target search and rescue strategy determination device provided by the third embodiment of the application. As shown in Figure 3 The device includes: a dynamic model construction module 310, for obtaining marine environment information of target sea area, calculating the motion characteristics of underwater target according to marine environment information, and constructing a dynamic model;

[0115] A target trajectory prediction module 320 is used to predict the target trajectory of the underwater target based on the Monte Carlo algorithm in combination with the dynamic model;

[0116] A search strategy optimization module 330 is used to construct a grid probability model, preliminarily determine a search strategy based on the grid probability model, and dynamically update a probability distribution to optimize the search strategy by using Bayesian theory.

[0117] The secondary optimization module 340 is configured to establish a cost-benefit analysis model to perform secondary optimization on the optimized search strategy.

[0118] Optionally, the dynamic model construction module 310 is configured to: filter out the correlation factors related to the underwater target motion from the marine environment information through a correlation analysis algorithm; determine the neutral buoyancy point of the underwater target, and obtain the seabed position, and compare the neutral buoyancy point with the seabed position to determine the target motion state, and output a scene label, wherein the scene label includes a drifting with ocean current scene and a gradually sinking scene; extract the certainty value of the ocean current characteristics in the marine environment information, and establish a motion reference system according to the fixed uniform distribution mass assumption; call a motion equation template according to the scene label, and substitute the correlation factors and the motion reference system into the motion equation template to generate a motion equation; obtain the environmental uncertainty parameters, integrate the motion equation with the environmental uncertainty parameters, and generate a dynamic model.

[0119] Optionally, the target trajectory prediction module 320 is configured to: determine the state vector and the state transition equation of the underwater target motion through the dynamic model; obtain the position and velocity data of the underwater target before the accident as the initial state, and obtain the preset initial state uncertainty covariance and process noise covariance; randomly generate a particle set representing the initial state according to the initial state and the initial state uncertainty covariance; call the dynamic model according to the time step, and add random disturbance to each particle based on the process noise covariance to calculate the predicted state of each particle; perform statistics on each predicted state to generate a probability distribution graph, a mean trajectory and a confidence region; calculate the position probability of the underwater target based on the probability distribution graph, the mean trajectory and the confidence region through density estimation, and determine the target trajectory according to the highest value of the position probability.

[0120] Optionally, the search strategy optimization module 330 includes a grid model construction unit configured to: obtain the search and rescue area position, and determine the grid division range based on the position of the underwater target before the accident and the search and rescue area position; discretize the grid division range into unit cells, and calculate the number of grids based on a preset rule, define the grid label and the center point coordinates, and complete the area gridding; according to the target trajectory, combine the Manhattan distance and the Poisson distribution to assign an initial existence probability to each unit cell, introduce a time dimension to correct the probability, and complete the grid probability model construction.

[0121] Optionally, the search strategy optimization module 330 specifically comprises a search strategy determination unit configured to determine the to-be-searched grids in the grid probability model, extract real-time probability values of each to-be-searched grid, sort each to-be-searched grid according to the order from high to low of the real-time probability values to generate a priority sequence, acquire the current search resource number and the device search efficiency, determine the current coverage grid number based on the current search resource number and the device search efficiency, and select the corresponding to-be-searched grid from the priority sequence as the search strategy according to the current coverage grid number.

[0122] Optionally, the search strategy optimization module 330 specifically comprises a search strategy optimization unit configured to acquire the grid search result after the current search time interval ends, extract the current probability value of the corresponding grid, wherein the grid search result comprises successful discovery of the target and failure to discover the target, correct the current probability value of the corresponding grid to 0 through the Bayes formula when the grid search result is failure to discover the target, perform re-normalization calculation on the probabilities of the remaining to-be-searched grids to generate an updated grid probability distribution, and re-sort the to-be-searched grids according to the probability descending order based on the updated grid probability distribution to generate an optimized search strategy.

[0123] Optionally, the secondary optimization module 340 is specifically configured to acquire the resource consumption data and the benefit data of the optimized search strategy, determine evaluation indexes, construct a cost-benefit calculation model according to the evaluation indexes, calculate the cost-benefit ratio by substituting the optimized search strategy into the cost-benefit calculation model, and perform resource allocation adjustment on the strategies with the cost-benefit ratio lower than a preset threshold until the preset threshold is met.

[0124] The technical scheme of the embodiment of the present application provides accurate environment and motion basis for subsequent underwater target trajectory prediction and search strategy formulation by constructing a kinetic model. The uncertainty of the environment and target motion is fully considered to generate a plurality of possible trajectories, which provides rich reference for determining the area where the target may exist and improves the reliability of trajectory prediction. The search and rescue area is quantized into grids and assigned with probabilities, which can intuitively and clearly determine the search priority, and the dynamic updating of the probability distribution based on the Bayes theory can continuously optimize the search strategy as the search progresses, thereby improving the pertinence and efficiency of the search. The cost-benefit analysis model is established to perform secondary optimization of the strategy, realize optimal allocation of resources, ensure the search and rescue effect while controlling the cost, and improve the economy and practicability.

[0125] The underwater target search and rescue strategy determination device provided in the embodiment of the present application can execute the underwater target search and rescue strategy determination method provided in any embodiment of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0126] Embodiment four

[0127] Figure 4A structural diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0128] As shown in Figure 4 The electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected in communication with the at least one processor 11, where the memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer programs stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0129] Various components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, a speaker, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0130] The processor 11 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as a method for determining an underwater target search and rescue strategy.

[0131] In some embodiments, an underwater target search strategy determination method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 18. In some embodiments, parts or all of the computer program can be loaded and / or installed onto electronic device 10 via, e.g., ROM 12 and / or communication unit 19. When the computer program is loaded onto RAM 13 and executed by processor 11, one or more steps of an underwater target search strategy determination method as described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform an underwater target search strategy determination method by other means, e.g., with the aid of firmware.

[0132] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0133] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0134] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0135] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0136] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0137] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0138] It should be understood that the various forms of flow shown above can be reordered, added to, or have steps deleted. For example, the steps described in the present application can be performed in parallel, in series, or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which are not limited herein.

[0139] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining underwater target search and rescue strategies, characterized in that, include: Obtain marine environmental information of the target sea area, calculate the motion characteristics of underwater targets based on the marine environmental information, and construct a dynamic model; Based on the Monte Carlo algorithm, combined with the aforementioned dynamic model, the target trajectory of underwater targets is predicted. A grid probability model is constructed, and a preliminary search strategy is determined based on the grid probability model. The probability distribution is dynamically updated using Bayesian theory to optimize the search strategy. Establish a cost-benefit analysis model and perform secondary optimization on the optimized search strategy; The step of calculating the motion characteristics of underwater targets based on marine environmental information and constructing a dynamic model includes: By using correlation analysis algorithms, we can screen out the correlation factors in marine environmental information that are related to the movement of underwater targets. Determine the neutral buoyancy point of the underwater target and obtain the seabed position. Compare the neutral buoyancy point with the seabed position to determine the target's motion state and output scene labels. The scene labels include scenes of drifting with ocean currents and scenes of gradually sinking. Deterministic values ​​of ocean current characteristics are extracted from marine environmental information, and a motion reference system is established based on the assumption of a fixed uniform mass distribution. The motion equation template is called based on the scene label, and the correlation factor and motion reference system are substituted into the motion equation template to generate the motion equation. By acquiring environmental uncertainty parameters and integrating the equations of motion with these parameters, a dynamic model is generated.

2. The method according to claim 1, characterized in that, The prediction of underwater target trajectory based on the Monte Carlo algorithm and the dynamic model includes: The state vector and state transition equation of the underwater target motion are determined by a dynamic model. The position and velocity data of the underwater target before the accident are obtained as the initial state, and the preset initial state uncertainty covariance and process noise covariance are obtained. Based on the initial state and the initial state uncertainty covariance, a set of particles representing the initial state is randomly generated; The dynamic model is invoked at each time step, and the predicted state of each particle is calculated by superimposing random perturbations on each particle based on the process noise covariance. Statistical analysis is performed on each predicted state to generate a probability distribution map, mean trajectory, and confidence region; Based on the probability distribution map, mean trajectory, and confidence region, the location probability of the underwater target is calculated through density estimation, and the target trajectory is determined based on the highest value of the location probability.

3. The method according to claim 2, characterized in that, The construction of the grid probability model includes: Obtain the location of the search and rescue area, and determine the grid division range based on the location of the underwater target before the accident and the location of the search and rescue area; The grid area is discretized into individual cells, and the number of cells is calculated based on preset rules. Each cell is labeled and its center point coordinates are defined to complete the area rasterization. Based on the target trajectory, an initial existence probability is assigned to each cell by combining Manhattan distance and Poisson distribution. The probability is then corrected by introducing a time dimension, thus completing the construction of the grid probability model.

4. The method according to claim 1, characterized in that, The preliminary determination of the search strategy based on the grid probability model includes: Identify the grids to be searched in the grid probability model and extract the real-time probability values ​​of each grid to be searched; The grids to be searched are sorted in descending order of their real-time probability values ​​to generate a priority sequence; Obtain the current number of search resources and the device search efficiency, determine the current number of coverage grids based on the current number of search resources and the device search efficiency, and select the corresponding grid to be searched from the priority sequence as the search strategy according to the current number of coverage grids.

5. The method according to claim 4, characterized in that, The method of dynamically updating the probability distribution using Bayesian theory to optimize the search strategy includes: Obtain the grid search results after the current search time interval ends, and extract the current probability value of the corresponding grid, wherein the grid search results include successfully found targets and unfounded targets; When the grid search result is no target found, the current probability value of the corresponding grid is corrected to 0 using Bayes' theorem, and the probability of the remaining grids to be searched is recalculated to generate an updated grid probability distribution. Based on the updated grid probability distribution, the grids to be searched are reordered in descending order of probability to generate an optimized search strategy.

6. The method according to claim 1, characterized in that, The establishment of a cost-benefit analysis model and the subsequent secondary optimization of the optimized search strategy include: Obtain resource consumption and efficiency data for the optimized search strategy to determine evaluation metrics; A cost-benefit calculation model is constructed based on each evaluation indicator; The optimized search strategy is then substituted into the cost-benefit calculation model to calculate the cost-benefit ratio. For strategies with a cost-effectiveness ratio lower than a preset threshold, resource allocation is adjusted until the preset threshold is met.

7. A device for determining underwater target search and rescue strategies, characterized in that, include: The dynamic model construction module is used to acquire marine environmental information of the target sea area, calculate the motion characteristics of underwater targets based on the marine environmental information, and construct a dynamic model. The target trajectory prediction module is used to predict the target trajectory of underwater targets based on the Monte Carlo algorithm and the aforementioned dynamic model. The search strategy optimization module is used to construct a grid probability model, initially determine the search strategy based on the grid probability model, and dynamically update the probability distribution through Bayesian theory to optimize the search strategy. The secondary optimization module is used to establish a cost-benefit analysis model and perform secondary optimization on the optimized search strategy; The dynamic model construction module is specifically used to: use correlation analysis algorithms to screen out correlation factors related to the movement of underwater targets in marine environmental information; Determine the neutral buoyancy point of the underwater target and obtain the seabed position. Compare the neutral buoyancy point with the seabed position to determine the target's motion state and output scene labels. The scene labels include scenes of drifting with ocean currents and scenes of gradually sinking. Deterministic values ​​of ocean current characteristics are extracted from marine environmental information, and a motion reference system is established based on the assumption of a fixed uniform mass distribution. The motion equation template is called based on the scene label, and the correlation factor and motion reference system are substituted into the motion equation template to generate the motion equation. By acquiring environmental uncertainty parameters and integrating the equations of motion with these parameters, a dynamic model is generated.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

9. A computer storage medium, characterized in that, The computer storage medium stores computer instructions that are used to cause a processor to execute the method of any one of claims 1-6.

Images