An underwater path planning method for multi-intelligent AUVs considering ocean currents

Abstract

A multi-intelligent AUVs underwater path planning method considering ocean currents includes the following steps: (1) Based on the ocean data collected in advance, the ocean current function is used to adjust the parameters to obtain the simulated ocean currents, and at the same time, the most valuable waypoints during navigation are determined to establish a guided two-dimensional ocean current map of the zigzag search strategy; (2) Based on the idea of ​​improved DBO, an ocean factor cloud theory is used in the constructed problem space to initialize the position of the AUVs group; (3) Each AUV evaluates the fitness of its surrounding environment according to its current position, and takes actions corresponding to the group to update its position; (4) The groups are divided into four groups according to the proportion of adaptive ocean currents. The intelligent AUV path planning method proposed in this scheme makes the overall path planning method more efficient.

Description

Technical Field

[0001] The invention relates to the field of intelligent underwater robots, and in particular to a method for underwater path planning of multi-intelligent AUVs taking ocean currents into consideration. Background Art

[0002] At present, the deep sea is a field that is less explored by humans. There are many unknown resources hidden in this field. In addition, in this field, the ocean is a huge carbon sink that absorbs and stores carbon dioxide. By studying and searching the deep sea, scholars can better understand the carbon cycle process and its impact on climate change, which has an important impact on the study of the earth's climate system. Intelligent AUVs (unmanned autonomous submersibles) for deep-sea exploration are in urgent need. The exploration of AUVs in the ocean can be divided into important steps such as navigation and positioning, path planning, data acquisition and processing, communication and data transmission, and energy management and maintenance. Among them, path planning technology is an important step in the AUV exploration link. Planning the optimal path can often enable AUVs to complete underwater tasks in deep-sea current disturbances more efficiently, safely, and accurately. Given the particularity of the underwater environment, especially due to the limitations of low-bandwidth submarine channels, effective communication is particularly difficult in this environment. In this challenging context, path planning for underwater robots has become a highly complex and demanding research field. Especially considering the dynamic changes and uncertainty of ocean currents in the deep-sea ocean environment, it is crucial to develop an adaptable and efficient path planning strategy.

[0003] The current AUV path planning algorithms can be divided into geometric model-based, sampling-based, artificial potential field-based and intelligent algorithms according to their basic principles. The path planning algorithm based on the geometric model belongs to discrete optimal planning, which is a traditional and mature method. However, these algorithms require a high level of model construction, and the final path planning results are closely related to the accuracy of the model. When dealing with complex dynamic environments, intelligent algorithms can better solve path planning problems. Compared with traditional algorithms, swarm intelligence optimization algorithms simulate the behavior of biological populations and have higher computational efficiency in large or complex search spaces. In addition, swarm intelligence optimization algorithms are suitable for parallel processing and distributed computing, so they are more effective for large-scale problems. In short, when dealing with ocean current disturbances, uncertainties, and path planning problems in complex environments, swarm intelligence optimization algorithms have stronger adaptability, computational efficiency, and integration.

[0004] Representative algorithms in intelligent optimization algorithms include genetic algorithms, ant colony algorithms, and particle swarm algorithms. Compared with these classic intelligent optimization algorithms, the DBO algorithm uses different search methods based on subpopulations when performing searches, and uses a dynamic boundary search strategy to reduce the search space. Some studies have shown that compared with other classic algorithms, the DBO algorithm has better convergence and stronger global planning capabilities. However, in reality, the DBO algorithm also has some defects, including the difficulty in effectively planning the optimal path in an uncertain environment. Therefore, how to improve the DBO algorithm so that it can better plan the optimal path in an uncertain environment such as ocean current interference is an urgent problem to be solved. Summary of the invention

[0005] In view of the above-mentioned defects, the purpose of the present invention is to propose an intelligent AUV path planning method suitable for submarine current environment, which makes the overall path planning method more efficient.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] A multi-intelligent AUVs underwater path planning method considering ocean currents comprises the following steps:

[0008] (1) Based on the ocean data collected in advance, the ocean current function is used to adjust the parameters to obtain simulated ocean currents, and at the same time, the most valuable waypoints during navigation are determined to establish a two-dimensional ocean current map guided by the field-shaped search strategy;

[0009] (2) Based on the idea of ​​improved DBO, an ocean factor cloud theory is used to initialize the position of the AUVs group in the constructed problem space;

[0010] (3) Each AUV evaluates the fitness of its surrounding environment based on its current position and takes actions corresponding to the population to update its position;

[0011] (4) Divided into four major populations according to the proportion of adaptive ocean currents;

[0012] (5) Population 1 performs the rolling action by integrating the ocean current factor, dancing by integrating the jumping mechanism when encountering obstacles, population 2 by foraging, population 3 by reproduction, and population 4 by stealing by approaching the distance difference.

[0013] (6) Integrate information and make decisions on the optimal path points in each region, and select the optimal path through the ocean current evaluation mechanism.

[0014] Furthermore, in step (1), the two-dimensional guided ocean current map of the zigzag search strategy, the zigzag search strategy steps are as follows: take four populations as units, search throughout the entire two-dimensional plane map, in each zigzag small area, update the positions of different AUVs according to the different strategies of the four major populations, and judge the optimal path point according to the fitness value. The path points of the optimal and worst positions are shown in the following formula:

[0015] The positions of the four AUV populations are represented as: AUVs population 1, AUVs population 2, AUVs population 3, and AUVs population 4;

[0016] G={G g |g=1,2,…,∞} (1)

[0017] F={F f |f=1,2,…,∞} (2)

[0018] H={H h |h=1,2,…,∞} (3)

[0019] T={T t |t=1,2,…,∞} (4)

[0020]

[0021] in, c(x i ,x i+1 ) represents the influence of the ocean current from position xi to xi+1;

[0022]

[0023] ω c =α 3 ·(1+β|C * (P x )|)(5.1)

[0024]

[0025] Formula (1) G represents the AUVs population 1 that performs rolling ball action with the ocean current factor integrated; Formula (2) F represents the AUVs population 2 that performs adaptive reproduction with the ocean current; Formula (3) H represents the AUVs population 3 that performs adaptive foraging with the ocean current; Formula (4) T represents the AUVs population 4 that performs stealing action with the goal of avoiding harm and seeking benefits; the brackets represent the locations of the AUVs one by one;

[0026] In formula (5):

[0027] P b = {P x} indicates the optimal position;

[0028] P w = {P x} indicates the worst position;

[0029] f(P x ) represents the individual optimal degree value;

[0030] The fitness function obtained according to the objective function;

[0031] The weight ω is dynamically adjusted according to the current ocean current conditions. a ,ω b and ω c , where α 1 , α 2 ,α 3 is the initial weight, β is a tuning parameter, |C * (P(x)| represents the intensity of ocean current, F min represents the minimum path, ∈ min It indicates the minimum energy consumed. The smaller the fitness value, the more suitable the AUV individual is.

[0032] Formula (5.1) Dynamic weight adjustment: Dynamically adjust the weight ω according to the current ocean current conditions a ,ω a , and ω a , where α 1 ,α 2 ,α 3 is the initial weight, β is a tuning parameter, |C * (Px)| represents the intensity of ocean current;

[0033] Formula (5.2) Here M(t) is an ocean current prediction model, which can predict the changes of ocean currents in the future. This solution introduces an ocean current model to represent the ocean currents in a sea area. If it is in other sea areas, the ocean current model can be summarized based on the ocean data and then substituted into it.

[0034] in, is a time-varying function, representing the influence of ocean currents from xi to xi+1 at time t.

[0035] Furthermore, in step (2), the ocean factor cloud theory is used for initialization, and the specific implementation steps are as follows: The ocean current formula is as follows:

[0036]

[0037] In formula (6), I x , L x , M x , b xThe expected value and variance of are regarded as individuals of the population, and the expected value and variance of these parameters are generated by fusion ocean current cloud theory as individuals of the initial population, so as to initialize the population of the dung beetle algorithm;

[0038] Cloud theory generation vortex parameters I x , L x , M x , b x The expected value and variance of

[0039] E[I x ]E[L x ]E[M x ]E[b x ] represent the vortex parameters I x , L x , M x , b x The expected value of Denote their variances respectively, then they can be expressed as:

[0040] Vortex intensity I x The expected value E[I x ] and variance

[0041] Vortex position abscissa M x The expected value E[M x ] and variance

[0042] Vortex position ordinate L x The expected value E[L x ] and variance

[0043] Vortex size radius b x The expected value E[b x ] and variance

[0044] Based on the cloud theory normal distribution, its probability density function is as follows: The AUV population individuals usually hope that their distribution conforms to a specific probability density function, such as formula (7). The purpose of this is to ensure that the distribution of the population in the search space can cover the potential solution space and can more effectively search for the optimal solution;

[0045]

[0046] In this formula, x is the random variable, E is the expected value, and σ 2 is the variance;

[0047] By adjusting E and σ 2, we can generate different random numbers to achieve random generation of vortex parameters; according to the population generated by this method, the following formula (8) represents its individuals, which is expressed as follows:

[0048]

[0049] Thus, an individual x i It contains the expected values ​​and variances of all vortex parameters.

[0050] Furthermore, the population 1 adopts the rolling action of integrating the ocean current factor. When it is in the obstacle-free area, this operation mode is adopted. When λ<γ, it is the obstacle-free mode. Combined with the ocean current formula, It is changed to ocean current environmental impact factor, and the formula is as follows:

[0051]

[0052] Formula (9) represents the rolling action of population 1 by integrating the ocean current factor, and formula (10) represents the ocean current environment impact factor;

[0053] Among them, L is the current factor. Since the current factor affects the overall impact of the current on the path planning, it can be considered to be set in the range of [-1,1], indicating the positive or negative impact of the current on the path planning, and 0 means no impact of the current;

[0054] w is the ocean current vortex deflection coefficient. The ocean current vortex deflection coefficient affects the degree of influence of the vortex in the ocean on the path planning. It can be considered to be set in the range of [0,1], indicating the degree of positive influence of the ocean current vortex on the path planning. 0 means no influence by the vortex, and 1 means maximum influence.

[0055] ζ i '(a) and ζ j '(a) represents the position information that does not conform to the ocean current, which can be calculated using the ocean current formula and then these values ​​are compared with the position information that conforms to the ocean current ζ i (a) and ζ j (a) Subtract, then subtract the worst position |s i (t)-S w |, thereby obtaining updated location information;

[0056] When λ≥γ, it is the obstacle mode. At this time, population 1 performs the guided dance step, which integrates an adaptive target point search direction method. First, find the angle α formed by the straight line connecting the current node (x0, y0) and the target point (xg, yg) with the x-axis direction; α is mainly determined by the angle between the starting point, the target point and the x-axis;

[0057]

[0058] in,

[0059] Formula (11) represents the guided dance steps performed by population 1, and formula (12) represents the angle α;

[0060] Current Node Target point Represents obstacles, τ represents an adjustable parameter, which can control the degree of influence of obstacles on path planning, so that the algorithm can avoid obstacles more flexibly and improve the success rate of path planning. By establishing a relationship between the current node and the target point, the AUV can avoid obstacles and no longer search blindly, but determine according to the direction of the target point to avoid falling into local optimization, while also reducing unnecessary search time and improving efficiency;

[0061] Formula (11) incorporates It represents obstacles in the search direction and plays the role of predicting obstacles. By increasing the cost of moving close to obstacles, this mechanism actively guides the path search to avoid nearby obstacles as early as possible.

[0062] Furthermore, population 2 undergoes the reproduction step, the specific process is as follows: the reproduction area is changed, the specific formula is as follows

[0063] Ω'=max(S b ×(1-K),Ω) (13)

[0064] Formula (13) represents the lower limit of the breeding area, S b represents the optimal position in the population;

[0065]

[0066] Formula (14) represents the upper limit of the breeding area;

[0067]

[0068] ζ i (a) and ζ j (a) is the effect of ocean current on individual position a calculated according to the ocean current formula;

[0069] β is a tuning parameter used to control the degree of influence of ocean currents on the boundary scaling factor;

[0070] The parameter K in formulas (13) and (14) is formula (15);

[0071] The adaptive factor K designed in this way can reduce the boundary scale factor when the ocean current has a greater impact, and increase it when the ocean current has a smaller impact, so that the algorithm can adjust the search range more flexibly during the search process, which helps to better find the global optimal solution;

[0072]

[0073] Formula (16) represents the position update formula in the reproduction step of population 2.

[0074] Preferably, population 3 passes the foraging step, and the specific process is as follows: the boundary selection strategy is improved, and the ocean current adaptive factor is added to adapt to the influence of the ocean current on the path planning. The specific formula steps are as follows:

[0075] Ω=max(S b* ×(1-K),Ω) (17)

[0076] Formula (17) represents the lower limit of the breeding area;

[0077]

[0078] Formula (18) represents the upper limit of the breeding area, S b* represents the local optimal position, that is, the optimal position in the current population;

[0079]

[0080] Formula (19) represents the position update formula in the foraging step of population 3;

[0081] The adaptive adjustment of the boundary scale factor allows the algorithm to narrow the search range when it is affected by strong ocean currents to avoid over-exploration, and to expand the search range when the ocean currents are less affected to increase exploration. This flexibility and adaptability helps the algorithm better adapt to different search environments and improve the efficiency and accuracy of the search.

[0082] The influence of ocean currents may cause individuals to deviate from the target in the search space. If the search range is too large, resources may be wasted in invalid search space. On the contrary, if the search range is too small, potential global optimal solutions may be missed. Therefore, by adjusting the boundary scale factor, the search range can be controlled so that the algorithm can focus more on the area that may contain the optimal solution. Appropriate adjustment of the search range can help to better discover the global optimal solution. When the influence of ocean currents is small, increasing the search range can increase exploration and help to find the global optimal solution more comprehensively in the search space. When the influence of ocean currents is large, narrowing the search range can avoid falling into the local optimal solution and improve the efficiency of global search.

[0083] Preferably, population 4 steals by approaching the benefits and avoiding the harms. The idea is that the AUV is close to the optimal position and away from the worst position. The specific process of formula (20) is as follows:

[0084] s i (t+1)=s b +H×Z×((P b -|s i (t)|)-(P w -|s i (t)|)+(s b -|s i (t)|)-(s w -|s i (t)|))(20)

[0085] By subtracting the worst solution (P w -|s i (t)|) and the global worst solution (s w -|s i (t)|), making the overall formula closer to the optimal solution and away from the worst solution.

[0086] Preferably, the AUV group is divided into four groups as a unit to search for the optimal position, namely, group 1 integrates the rolling action and guided dancing mechanism of the ocean current factor, group 2 and group 3 respectively perform ocean current adaptive reproduction and ocean current adaptive foraging through the ocean current adaptive boundary, and group 4 performs the path point of finding the optimal point by stealing in a way of avoiding harm and seeking benefits;

[0087] Adaptive ratio refers to adapting to the changes in ocean currents. The ratio can be dynamically adjusted according to the impact of ocean currents to adapt to ocean currents, making the overall method more efficient. The specific formula is as follows:

[0088] The four proportional coefficients are shown in formula (21):

[0089]

[0090] They represent AUV population 1, AUV population 2, AUV population 3, and AUV population 4 respectively;

[0091] The ocean current adaptive factor is shown in formula (22):

[0092]

[0093] η represents the change of ocean current, k 1 and k 2 is an adjustment parameter; 0<ω<1. By adjusting the adaptive factor ω, the proportional coefficient of each population of AUVs is dynamically adjusted to adapt to ocean current changes and optimize the overall performance.

[0094] Preferably, the optimal path points in each area are integrated and decided, an evaluation mechanism that conforms to the ocean current is introduced, the optimal individuals are screened out according to the satisfaction, and finally the optimal path is obtained by summarizing;

[0095] The evaluation mechanism function of following the ocean current is shown in formula (23):

[0096]

[0097] s represents the coordinates of the path point, V represents the ocean current velocity vector, represents the direction of the unit vector pointing to the path point s, d(s) represents the distance from the path point s to the nearest ocean current line, and c vortex (s) represents the vortex influence factor at the path point s, |V| represents the magnitude of the ocean current speed;

[0098] The first part considers the consistency of the direction of the waypoints with the direction of the ocean current and normalizes it according to the magnitude of the ocean current speed;

[0099] The second part considers the distance from the waypoint to the nearest ocean current line, the closer the distance, the higher the adaptability;

[0100] The third part considers the influence of eddies on the path points. The eddy influence factor is related to the speed of the ocean current, which reflects the influence of eddies on path planning.

[0101] f * (s)>0, indicating that the direction of path point s is consistent with the direction of the ocean current, and the adaptability is high;

[0102] f * (s) = 0, indicating that the direction of path point s is perpendicular to the direction of the ocean current, and the adaptability is medium;

[0103] f * (s)<0, indicating that the direction of path point s is opposite to the direction of the ocean current and has low adaptability.

[0104] Preferably, the fitness value is calculated based on the objective function, and the objective function formula is as follows:

[0105] Position representation formula:

[0106] G={G g |g=1,2,…,∞} (1)

[0107] F={F f |f=1,2,…,∞} (2)

[0108] H={H h |h=1,2,…,∞} (3)

[0109] T={Tt |t=1,2,…,∞} (4)

[0110] These four formulas represent the position sets of four AUV populations, each set contains an infinite number of possible position points, which is set to 30 AUVs per population in the patent of this invention;

[0111]

[0112] This set of formulas (5) defines the optimal position P b and the worst position P w , and the fitness function f(P x ) calculation method;

[0113]

[0114] Formula (al) represents the shortest path;

[0115]

[0116] Formula (a2) represents the minimum energy consumption;

[0117] The two formulas of the objective function define the two main goals of the optimization problem: the shortest path and the minimum energy consumption according to the ocean current adaptation, where C * It is the adaptive ocean current influencing factor;

[0118] Constraints:

[0119]

[0120] Position constraints:

[0121]

[0122] Speed ​​Constraints:

[0123]

[0124] Acceleration constraints:

[0125]

[0126] The above formulas define various constraints for AUV motion. The meanings of the variables are:

[0127] G g ,F f ,H h ,T t : represent the individual positions in the four AUV populations;

[0128] P x ,Py : AUV position point;

[0129] f(P x ):Position P x The fitness value of

[0130] Minimum path length;

[0131] ∈ min :Minimum energy consumption;

[0132] d j : The length of the jth segment of the path;

[0133] Energy consumption of the i-th path;

[0134] ζ i (α),ζ j (α): Ocean current influence factor, including the position, radius and size of the vortex, that is, the generated path points must be within this limit;

[0135] S (x,y) : 2D position coordinates of AUV;

[0136] The speed and acceleration of the AUV;

[0137] r max ,a max : Maximum speed and acceleration limits;

[0138] Fitness function f(P x ) is directly related to the objective function F (shortest path) and ε (minimum energy consumption);

[0139] The four AUV populations (G, F, H, and T) represent different search strategies, each targeting specific environmental factors or constraints;

[0140] Objective functions (a1) and (a2) and F in the fitness function min and ∈ min Direct correspondence;

[0141] Constraints (a3)-(a6) ensure that the motion of the AUV meets practical physical and environmental constraints;

[0142] The Tianzi search strategy divides the entire search space into small areas, applies different AUV population strategies in each area, and judges the optimal path point by the fitness value.

[0143] One of the above technical solutions includes the following beneficial effects: 1. An intelligent AUV path planning method suitable for the seabed ocean current environment is designed. By dividing the AUVs into four populations and searching for the optimal position at the same time, the efficiency is improved. At the same time, each position update step is optimized and improved, making the overall path planning method more efficient. 2. By introducing the ocean current adaptive factor, the population ratio is adjusted through real-time ocean current data to ensure that the AUV group can quickly adapt to environmental changes. At the same time, by following the ocean current for path planning, the overall AUVs energy efficiency is low, which improves the energy utilization rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0144] Figure 1 is an overall method flow chart of an embodiment of the present invention;

[0145] Figure 2 is an ocean current path planning diagram of an embodiment of the present invention;

[0146] Figure 3 It is an ocean current field-shaped diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0147] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0148] like Figure 1-3 As shown, a multi-intelligent AUVs underwater path planning method considering ocean currents includes the following steps:

[0149] (1) Based on the ocean data collected in advance, the ocean current function is used to adjust the parameters to obtain simulated ocean currents, and at the same time, the most valuable waypoints during navigation are determined to establish a two-dimensional ocean current map guided by the field-shaped search strategy;

[0150] (2) Based on the idea of ​​improved DBO, an ocean factor cloud theory is used to initialize the position of the AUVs group in the constructed problem space;

[0151] (3) Each AUV evaluates the fitness of its surrounding environment based on its current position and takes actions corresponding to the population to update its position;

[0152] (4) Divided into four major populations according to the proportion of adaptive ocean currents;

[0153] (5) Population 1 performs the rolling action by integrating the ocean current factor, dancing by integrating the jumping mechanism when encountering obstacles, population 2 by foraging, population 3 by reproduction, and population 4 by stealing by approaching the distance difference.

[0154] (6) Integrate information and make decisions on the optimal path points in each region, and select the optimal path through the ocean current evaluation mechanism.

[0155] In view of the long running time of the algorithm, a parallel search method, the Tianzi search strategy, is adopted. As the algorithm has many population update steps, the four populations are randomly divided into arbitrary areas in the Tianzi shape, and searches are performed in the corresponding areas to improve the efficiency of the algorithm and save time.

[0156] The ocean factor cloud theory is aimed at the initialization of the population in the ocean current environment. It is a method of changing the cloud theory into one that adapts to the ocean current environment. The cloud theory can initialize the population according to the strength, position and radius of the vortex, thereby generating an AUVs population that adapts to the ocean current environment.

[0157] The AUV in this step has intelligent sensing sensors and the ability to self-evaluate. Each AUV is regarded as a particle in the algorithm, and then takes corresponding update steps according to the steps in the algorithm. Modern AUVs are usually equipped with various sensors, such as sonar, pressure sensors, gyroscopes, etc., to perceive the surrounding environment. The control system of the AUV usually includes self-diagnosis and evaluation functions, which can evaluate its own status and performance. In the path planning problem, the AUV needs to be able to evaluate the fitness of the current position, which corresponds to the fitness evaluation in the algorithm.

[0158] This is aimed at the ocean current environment, so that the proportion of the population is adjusted according to the different ocean currents, such as the ocean current adaptive factor formula in formula 22, combined with formula 21, to obtain this operation.

[0159] The algorithm is improved to address the impact of ocean current disturbances and adapt it to path planning under ocean currents.

[0160] In recent years, machine learning algorithms, especially deep learning and reinforcement learning, have made significant progress in the field of path planning. However, in a complex and changeable marine environment, the use of reinforcement learning for path planning faces many challenges. For example, the dimensionality explosion problem causes reinforcement learning algorithms to need to deal with high-dimensional state space and action space, and the demand for data and computing resources grows exponentially. In addition, the high cost of experiments is also a major problem. Interactive learning in a real environment requires a lot of time, manpower and money, while simulation experiments have modeling difficulties and inherent model uncertainty. The present invention proposes a new AUV path planning method, which aims to utilize ocean current energy, avoid underwater obstacles, and plan a path with minimal energy consumption, shortest time and shortest travel. This method fully considers the constraints of the uncertainty of the seabed environment and the dynamic characteristics of the ocean current, and uses biological intelligence algorithms to solve the non-deterministic polynomial (NP) path planning problem.

[0161] This step is to summarize the optimal points obtained by different updating methods in the four populations, and obtain the optimal path through the evaluation mechanism. This step is to prevent the optimal point from not conforming to the trend of ocean currents due to the summary. This part is through formula 23. For the problem of ocean current environment, the relationship between the direction of the path point and the direction of the ocean current and the speed of the ocean current is considered; secondly, the distance from the path point to the nearest ocean current line is considered. The ocean current line refers to the path of the water flowing in the ocean. There are a large number of ocean currents in the ocean, which are caused by factors such as wind, the rotation of the earth, and differences in water temperature. Ocean current lines can describe the direction and path of the movement of ocean currents; in addition, the impact of vortex problems on path points is considered. In general, the purpose of this evaluation mechanism is to confirm that the final generated path can fully adapt to the ocean current environment. The details of the formula can be seen in the description of point 10.

[0162] Among them, in step (1), the two-dimensional guided ocean current map of the Tianzi search strategy, the Tianzi search strategy steps are as follows: take four populations as units, search throughout the entire two-dimensional plane map, in each Tianzi small area, update the positions of different AUVs according to different strategies of the four major populations, and judge the optimal path point according to the fitness value. The path points of the optimal and worst positions are shown in the following formula:

[0163] The positions of the four AUV populations are represented as: AUVs population 1, AUVs population 2, AUVs population 3, and AUVs population 4;

[0164] G={G g |g=1,2,…,∞} (1)

[0165] F={F f |f=1,2,…,∞} (2)

[0166] H={H h|h=1,2,…,∞} (3)

[0167] T={T t |t=1,2,…,∞} (4)

[0168]

[0169] in, c(x i ,x i+1 ) represents the influence of the ocean current from position xi to xi+1;

[0170]

[0171] ω c =α 3 ·(1+β|C * (P x )|) (5.1)

[0172]

[0173] Formula (1) G represents the AUVs population 1 that performs rolling ball action with the ocean current factor integrated; Formula (2) F represents the AUVs population 2 that performs adaptive reproduction with the ocean current; Formula (3) H represents the AUVs population 3 that performs adaptive foraging with the ocean current; Formula (4) T represents the AUVs population 4 that performs stealing action with the goal of avoiding harm and seeking benefits; the brackets represent the locations of the AUVs one by one;

[0174] In formula (5):

[0175] P b = {P x} indicates the optimal position;

[0176] P w = {P x} indicates the worst position;

[0177] f(P x ) represents the individual optimal degree value;

[0178] The fitness function obtained according to the objective function

[0179] The weight ω is dynamically adjusted according to the current ocean current conditions. a ,ω b and ω c , where α 1 , α 2 ,α 3 is the initial weight, β is a tuning parameter, |C * (P(x)| represents the intensity of ocean current, F min represents the minimum path,min It indicates the minimum energy consumed. The smaller the fitness value, the more suitable the AUV individual is.

[0180] Formula (5.1) Dynamic weight adjustment: Dynamically adjust the weight ω according to the current ocean current conditions a ,ω a , and ω a , where α 1 ,α 2 ,α 3 is the initial weight, β is a tuning parameter, |C * (Px)| represents the intensity of ocean current;

[0181] Formula (5.2) Here M(t) is an ocean current prediction model, which can predict the changes of ocean currents in the future. The present invention introduces an ocean current model to represent the ocean currents in a sea area. If it is in other sea areas, the ocean current model can be summarized based on the ocean data and then substituted;

[0182] in, is a time-varying function, representing the influence of ocean currents from xi to xi+1 at time t.

[0183] The role of this strategy is to search in units of four populations on a two-dimensional plane map. In each small Tianzi-shaped area, the positions of different AUVs are updated according to different strategies, the optimal path point is judged by the fitness value, and the path points with the best and worst positions are determined. Tianzi-shaped search strategy, which divides the search area into multiple small areas and adopts different search strategies in each area. This method allows AUVs to explore in multiple directions at the same time, thereby covering the entire search area more efficiently. In addition, the search within each small Tianzi-shaped area may be adjusted according to the direction and strength of the ocean current to ensure that the ocean current can be used to boost or avoid its resistance during the search. This approach allows the AUV to move to favorable areas while avoiding those path points where the ocean current interference leads to low efficiency.

[0184] In general, the roles played are as follows:

[0185] Improve search efficiency: By searching in multiple small areas simultaneously, the entire search area can be covered more quickly.

[0186] Taking advantage of ocean currents: By taking into account the effects of ocean currents, AUVs can take advantage of downstream currents or avoid upstream currents, thereby saving energy and improving navigation efficiency.

[0187] Dynamic Adjustment: Based on real-time fitness evaluation, AUVs can dynamically adjust their paths to cope with environmental changes and unknown obstacles.

[0188] Improved path optimization: By continuously evaluating and updating the best and worst-case pathpoints, AUVs can find more optimized paths to achieve their mission objectives.

[0189] In addition, in step (2), the ocean factor cloud theory is used for initialization, and the specific implementation steps are as follows:

[0190] The ocean current formula is as follows:

[0191]

[0192] In formula (6), I x , L x , M x , b x The expected value and variance of are regarded as individuals of the population, and the expected value and variance of these parameters are generated by fusion ocean current cloud theory as individuals of the initial population, so as to initialize the population of the dung beetle algorithm;

[0193] Cloud theory generation vortex parameter I x , L x , M x , b x The expected value and variance of

[0194] E[I x ]E[L x ]E[M x ]E[b x ] represent the vortex parameters I x , L x , M x , b x The expected value of Denote their variances respectively, then they can be expressed as:

[0195] Vortex intensity I x The expected value of $q_i$ is E[I x ] and variance

[0196] Vortex position abscissa M x The expected value E[M x ] and variance

[0197] Vortex position ordinate L x The expected value E[L x ] and variance

[0198] Vortex size radius b x The expected value E[b x ] and variance

[0199] Based on the cloud theory normal distribution, its probability density function is as follows: For the AUV population individuals, it is usually hoped that their distribution conforms to a specific probability density function, such as formula (7). The purpose of this is to ensure that the distribution of the population in the search space can cover the potential solution space and can more effectively search for the optimal solution.

[0200]

[0201] In this formula, x is the random variable, E is the expected value, and σ 2 is the variance. By adjusting E and σ 2 , we can generate different random numbers to achieve random generation of vortex parameters; according to the population generated by this method, the following formula (8) represents its individuals, which is expressed as follows:

[0202]

[0203] Thus, an individual x i It contains the expected values ​​and variances of all vortex parameters.

[0204] In addition, the population 1 adopts the rolling action of integrating the ocean current factor. When it is in the obstacle-free area, it adopts this operation mode. When λ<γ, it is the obstacle-free mode. Combined with the ocean current formula, It is changed to ocean current environmental impact factor, and the formula is as follows:

[0205]

[0206] Formula (9) represents the rolling action of population 1 by integrating the ocean current factor, and formula (10) represents the ocean current environment impact factor;

[0207] Among them, L is the current factor. Since the current factor affects the overall impact of the current on the path planning, it can be considered to be set in the range of [-1,1], indicating the positive or negative impact of the current on the path planning, and 0 means no impact of the current;

[0208] w is the ocean current vortex deflection coefficient. The ocean current vortex deflection coefficient affects the degree of influence of the vortex in the ocean on the path planning. It can be considered to be set in the range of [0,1], indicating the degree of positive influence of the ocean current vortex on the path planning. 0 means no influence by the vortex, and 1 means maximum influence.

[0209] ζ i '(a) and ζ j '(a) represents the position information that does not conform to the ocean current, which can be calculated using the ocean current formula and then these values ​​are compared with the position information that conforms to the ocean current ζ i (a) and ζ j (a) Subtract, then subtract the worst position |si (t)-S w |, thereby obtaining updated location information;

[0210] When λ≥γ, it is the obstacle mode. At this time, population 1 performs the guided dance step, which integrates an adaptive target point search direction method. First, find the current node (x 0 ,y 0 ) and the target point (x g ,y g ) is an angle α formed by a straight line connected to the x-axis direction; wherein α is mainly determined based on the angle between the starting point, the target point and the x-axis;

[0211]

[0212] in,

[0213] Formula (11) represents the guided dance steps performed by population 1, and formula (12) represents the angle α;

[0214] Current Node Target point Represents obstacles, τ represents an adjustable parameter, which can control the degree of influence of obstacles on path planning, so that the algorithm can avoid obstacles more flexibly and improve the success rate of path planning. By establishing a relationship between the current node and the target point, the AUV can avoid obstacles and no longer search blindly, but determine according to the direction of the target point to avoid falling into local optimization, while also reducing unnecessary search time and improving efficiency;

[0215] Formula (11) incorporates It represents obstacles in the search direction and plays the role of predicting obstacles. By increasing the cost of moving close to obstacles, this mechanism actively guides the path search to avoid nearby obstacles as early as possible.

[0216] In addition, population 2 undergoes a breeding step, the specific process is as follows: the breeding area is changed, the specific formula is as follows

[0217] Ω'=max(S b ×(1-K),Ω) (13)

[0218] Formula (13) represents the lower limit of the breeding area, S b represents the optimal position in the population;

[0219]

[0220] Formula (14) represents the upper limit of the breeding area;

[0221]

[0222] ζ i (a) and ζ j (a) is the effect of ocean current on individual position a calculated according to the ocean current formula;

[0223] β is a tuning parameter used to control the degree of influence of ocean currents on the boundary scaling factor;

[0224] The parameter K in formulas (13) and (14) is formula (15);

[0225] The adaptive factor K designed in this way can reduce the boundary scale factor when the ocean current has a greater impact, and increase it when the ocean current has a smaller impact, so that the algorithm can adjust the search range more flexibly during the search process, which helps to better find the global optimal solution;

[0226]

[0227] Formula (16) represents the position update formula in the reproduction step of population 2.

[0228] In addition, population 3 goes through the foraging step. The specific process is as follows: the boundary selection strategy is improved and the ocean current adaptive factor is added to adapt to the influence of ocean currents on path planning. The specific formula steps are as follows:

[0229] Ω=max(S b* ×(1-K),Ω) (17)

[0230] Formula (17) represents the lower limit of the breeding area;

[0231]

[0232] Formula (18) represents the upper limit of the breeding area, S b* represents the local optimal position, that is, the optimal position in the current population;

[0233]

[0234] Formula (19) represents the position update formula in the foraging step of population 3;

[0235] The adaptive adjustment of the boundary scale factor allows the algorithm to narrow the search range when it is affected by strong ocean currents to avoid over-exploration, and to expand the search range when the ocean currents are less affected to increase exploration. This flexibility and adaptability helps the algorithm better adapt to different search environments and improve the efficiency and accuracy of the search.

[0236] The influence of ocean currents may cause individuals to deviate from the target in the search space. If the search range is too large, resources may be wasted in invalid search space. On the contrary, if the search range is too small, potential global optimal solutions may be missed. Therefore, by adjusting the boundary scale factor, the search range can be controlled so that the algorithm can focus more on the area that may contain the optimal solution. Appropriate adjustment of the search range can help to better discover the global optimal solution. When the influence of ocean currents is small, increasing the search range can increase exploration and help to find the global optimal solution more comprehensively in the search space. When the influence of ocean currents is large, narrowing the search range can avoid falling into the local optimal solution and improve the efficiency of global search.

[0237] Among them, population 4 steals by avoiding harm and approaching benefits. The idea is that the AUV is close to the optimal position and away from the worst position. The specific process of formula (20) is as follows:

[0238] s i (t+1)=s b + H×Z×((P b -∣s i (t)∣)-(P w -∣s i (t)∣)+(s b -∣s i (t)∣)-(s w -∣s i (t)∣)) (20)

[0239] By subtracting the worst solution (P w -∣s i (t)|) and the global worst solution (s w -∣s i (t)∣), making the overall formula closer to the optimal solution and away from the worst solution.

[0240] In addition, the AUV group is divided into four groups as a field unit to search for the optimal position. Group 1 integrates the rolling action and guided dancing mechanism of ocean current factors, groups 2 and 3 respectively perform ocean current adaptive reproduction and ocean current adaptive foraging through ocean current adaptive boundaries, and group 4 performs the path point of finding the optimal point through stealing by avoiding harm and seeking benefits.

[0241] Adaptive ratio refers to adapting to the changes in ocean currents. The ratio can be dynamically adjusted according to the impact of ocean currents to adapt to ocean currents, making the overall method more efficient. The specific formula is as follows:

[0242] The four proportional coefficients are shown in formula (21):

[0243]

[0244] They represent AUV population 1, AUV population 2, AUV population 3, and AUV population 4 respectively;

[0245] The ocean current adaptive factor is shown in formula (22):

[0246]

[0247] η represents the change of ocean current, k 1 and k 2 is an adjustment parameter; 0<ω<1. By adjusting the adaptive factor ω, the proportional coefficient of each population of AUVs is dynamically adjusted to adapt to ocean current changes and optimize the overall performance.

[0248] In addition, the optimal path points in each area are integrated and decided, and an evaluation mechanism that conforms to the ocean current is introduced to screen out the best individuals according to the satisfaction, and finally the optimal path is obtained by summarizing;

[0249] The evaluation mechanism function of following the ocean current is shown in formula (23):

[0250]

[0251] s represents the coordinates of the path point, V represents the ocean current velocity vector, represents the direction of the unit vector pointing to the path point s, d(s) represents the distance from the path point s to the nearest ocean current line, and c vortex (s) represents the eddy influence factor at path point s. |V| represents the magnitude of the ocean current velocity.

[0252] The first part considers the consistency of the direction of the waypoints with the direction of the ocean current and normalizes it according to the magnitude of the ocean current speed;

[0253] The second part considers the distance from the waypoint to the nearest ocean current line, the closer the distance, the higher the adaptability;

[0254] The third part considers the influence of eddies on the path points. The eddy influence factor is related to the speed of the ocean current, which reflects the influence of eddies on path planning.

[0255] f * (s)>0, indicating that the direction of path point s is consistent with the direction of the ocean current, and the adaptability is high;

[0256] f * (s) = 0, indicating that the direction of path point s is perpendicular to the direction of the ocean current, and the adaptability is medium;

[0257] f * (s)<0, indicating that the direction of path point s is opposite to the direction of the ocean current and has low adaptability.

[0258] In addition, the fitness value is calculated based on the objective function, and the objective function formula is as follows:

[0259] Position representation formula:

[0260] G={G g |g=1,2,…,∞} (1)

[0261] F={F f |f=1,2,…,∞} (2)

[0262] H={H h |h=1,2,…,∞} (3)

[0263] T={T t |t=1,2,…,∞} (4)

[0264] These four formulas represent the position sets of four AUV populations. Each set contains an infinite number of possible position points, which is set to 30 AUVs per population in the patent of this invention;

[0265]

[0266] This set of formulas (5) defines the optimal position Pb and the worst position Pw, as well as the fitness function f(P x ) calculation method;

[0267]

[0268] Formula (al) represents the shortest path;

[0269]

[0270] Formula (a2) represents the minimum energy consumption;

[0271] The two formulas of the objective function define the two main goals of the optimization problem: to find the shortest path and minimize energy consumption according to the ocean current. * It is the adaptive ocean current influencing factor;

[0272] Constraints:

[0273]

[0274] Position constraints:

[0275]

[0276] Speed ​​Constraints:

[0277]

[0278] Acceleration constraints:

[0279]

[0280] The above formulas define various constraints for AUV motion. The meanings of the variables are:

[0281] G g ,F f ,H h ,T t : represent the individual positions in the four AUV populations;

[0282] P x ,P y : AUV position point;

[0283] f(P x ):Position P x The fitness value of

[0284] Minimum path length;

[0285] ∈ min :Minimum energy consumption;

[0286] d j : The length of the jth segment of the path;

[0287] Energy consumption of the i-th path;

[0288] ζ i (α),ζ j (α): Ocean current influence factor, including the position, radius and size of the vortex, that is, the generated path points must be within this limit;

[0289] S (x,y) : 2D position coordinates of AUV;

[0290] The speed and acceleration of the AUV;

[0291] r max ,a max : Maximum speed and acceleration limits;

[0292] Fitness function f(P x ) is directly related to the objective function F (shortest path) and ε (minimum energy consumption);

[0293] The four AUV populations (G, F, H, and T) represent different search strategies, each targeting specific environmental factors or constraints;

[0294] Objective functions (a1) and (a2) and F in the fitness function min and ∈min Direct correspondence;

[0295] Constraints (a3)-(a6) ensure that the motion of the AUV meets practical physical and environmental constraints;

[0296] The field-shaped search strategy divides the entire search space into small areas, applies different AUV population strategies in each area, and judges the optimal path point by the fitness value.

[0297] The technical principle of the present invention is described above in conjunction with specific embodiments. These descriptions are only for explaining the principle of the present invention and cannot be interpreted as limiting the scope of protection of the present invention in any way. Based on the explanations herein, those skilled in the art can associate other specific implementations of the present invention without paying creative labor, and these methods will fall within the scope of protection of the present invention.

Claims

1. A multi-intelligent AUVs underwater path planning method considering ocean currents, characterized in that: The following steps are involved: (1) Based on the ocean data collected in advance, the ocean current function is used to adjust the parameters to obtain simulated ocean currents, and at the same time, the most valuable waypoints during navigation are determined to establish a two-dimensional ocean current map guided by the field-shaped search strategy; (2) Based on the idea of ​​improved DBO, an ocean factor cloud theory is used to initialize the position of the AUVs group in the constructed problem space; (3) Each AUV evaluates the fitness of its surrounding environment based on its current position and takes actions corresponding to the population to update its position; (4) According to the adaptive ocean current ratio, it is divided into four major populations. The adaptive ocean current ratio refers to adapting to the changes in ocean currents. The ratio can be dynamically adjusted according to the influence of ocean currents to adapt to ocean currents, making the overall method more efficient. The specific formula is as follows: The four proportional coefficients are shown in formula (21): They represent AUV population 1, AUV population 2, AUV population 3, and AUV population 4 respectively; The ocean current adaptive factor is shown in formula (22): η represents the change of ocean current, k1 and k2 are adjustment parameters; 0<ω<1, by adjusting the adaptive factor ω, the proportional coefficient of each population AUV is dynamically adjusted to adapt to the change of ocean current and optimize the overall performance; (5) Population 1 performs the rolling action by integrating the ocean current factor, dancing by integrating the jumping mechanism when encountering obstacles, population 2 by foraging, population 3 by reproduction, and population 4 by stealing by approaching the distance difference. (6) Integrate information and make decisions on the optimal path points in each region, and select the optimal path through the ocean current evaluation mechanism.

2. The multi-intelligent AUVs underwater path planning method considering ocean currents according to claim 1 is characterized in that: In step (1), the two-dimensional guided ocean current map of the Tianzi search strategy, the Tianzi search strategy steps are as follows: take four populations as units, search throughout the entire two-dimensional plane map, in each Tianzi small area, update the positions of different AUVs according to the different strategies of the four major populations, and judge the optimal path point according to the fitness value. The path points of the optimal and worst positions are shown in the following formula: The positions of the four AUV populations are represented as: AUVs population 1, AUVs population 2, AUVs population 3, and AUVs population 4; G={G g |g=1,2,…,∞} (1) F={F f |f=1,2,…,∞} (2) H={H h |h=1,2,…,∞} (3) T={T t |t=1,2,…,∞} (4) in, c(x i ,x i+1 ) represents the influence of the ocean current from position xi to xi+1; Formula (1) G represents the AUVs population 1 that performs rolling ball action with the ocean current factor integrated; Formula (2) F represents the AUVs population 2 that performs adaptive reproduction with the ocean current; Formula (3) H represents the AUVs population 3 that performs adaptive foraging with the ocean current; Formula (4) T represents the AUVs population 4 that performs stealing action with the goal of avoiding harm and seeking benefits; the brackets represent the locations of the AUVs one by one; In formula (5): P b = {P x } indicates the optimal position; P w = {P x } indicates the worst position; f(P x ) represents the individual optimal degree value; The fitness function obtained according to the objective function; The weights ωa, ωb, and ωc are dynamically adjusted according to the current ocean current conditions, where α1, α2, α3 are the initial weights, β is a tuning parameter, |C * (P(x)| represents the intensity of ocean current, F min represents the minimum path, ε min It represents the minimum energy consumed. The smaller the fitness value, the more suitable the AUV individual is. Formula (5.1) Dynamic weight adjustment: Dynamically adjust the weights ωa, ωb, and ωc according to the current ocean current conditions, where α1, α2, α3 are the initial weights, β is a tuning parameter, |C * (Px)| represents the intensity of ocean current; Formula (5.2) Here M(t) is an ocean current prediction model, which can predict the changes of ocean currents in the future. This solution introduces an ocean current model to represent the ocean currents in a sea area. If it is in other sea areas, the ocean current model can be summarized based on the ocean data and then substituted into it. in, is a time-varying function, representing the influence of ocean currents from xi to xi+1 at time t.

3. The multi-intelligent AUVs underwater path planning method considering ocean currents according to claim 2 is characterized in that: In step (2), the ocean factor cloud theory is used for initialization. The specific implementation steps are as follows: The ocean current formula is as follows: In formula (6), I x , L x , M x , b x The expected value and variance of are regarded as individuals of the population, and the expected value and variance of these parameters are generated by fusion ocean current cloud theory as individuals of the initial population, so as to initialize the population of the dung beetle algorithm; Cloud theory generation vortex parameters I x , L x , M x , b x The expected value and variance of E[I x ]E[L x ]E[M x ]E[b x ] represent the vortex parameters I x , L x , M x , b x The expected value of Denote their variances respectively, then they can be expressed as: Vortex intensity I x The expected value E[I x ] and variance Vortex position abscissa M x The expected value E[M x ] and variance Vortex position ordinate L x The expected value E[L x ] and variance Vortex size radius b x The expected value E[b x ] and variance Based on the cloud theory normal distribution, its probability density function is as follows: The AUV population individuals usually hope that their distribution conforms to a specific probability density function, such as formula (7). The purpose of this is to ensure that the distribution of the population in the search space can cover the potential solution space and can more effectively search for the optimal solution; In this formula, x is the random variable, E is the expected value, and σ 2 is the variance; By adjusting E and σ 2 , we can generate different random numbers to achieve random generation of vortex parameters; according to the population generated by this method, the following formula (8) represents its individuals, which is expressed as follows: Thus, an individual x i It contains the expected values ​​and variances of all vortex parameters.

4. The multi-intelligent AUVs underwater path planning method considering ocean currents according to claim 3 is characterized by: The population 1 adopts the rolling action of integrating the ocean current factor. When it is in the obstacle-free area, this operation mode is adopted. When λ<γ, it is the obstacle-free mode. Combined with the ocean current formula, It is changed to ocean current environmental impact factor, and the formula is as follows: Formula (9) represents the rolling action of population 1 by integrating the ocean current factor, and formula (10) represents the ocean current environment impact factor; Among them, L is the current factor. Since the current factor affects the overall impact of the current on the path planning, it can be considered to be set in the range of [-1,1], indicating the positive or negative impact of the current on the path planning, and 0 means no impact of the current; w is the ocean current vortex deflection coefficient. The ocean current vortex deflection coefficient affects the degree of influence of the vortex in the ocean on the path planning. It can be considered to be set in the range of [0,1], indicating the degree of positive influence of the ocean current vortex on the path planning. 0 means no influence by the vortex, and 1 means maximum influence. ζ′ i (a) and ζ′ j (a) represents the position information that does not conform to the ocean current, which can be calculated using the ocean current formula and then these values ​​are compared with the position information ζ that conforms to the ocean current i (a) and ζ j (a) Subtract, then subtract the worst position |s i (t)-S w |, thereby obtaining updated location information; When λ≥γ, it is the obstacle mode. At this time, population 1 performs the guided dance step, which integrates an adaptive target point search direction method. First, find the angle α formed by the straight line connecting the current node (x0, y0) and the target point (xg, yg) with the x-axis direction; α is mainly determined by the angle between the starting point, the target point and the x-axis; in, Formula (11) represents the guided dance steps performed by population 1, and formula (12) represents the angle α; Current Node Target point Represents obstacles, τ represents an adjustable parameter, which can control the degree of influence of obstacles on path planning, so that the algorithm can avoid obstacles more flexibly and improve the success rate of path planning. By establishing a relationship between the current node and the target point, the AUV can avoid obstacles and no longer search blindly, but determine according to the direction of the target point to avoid falling into local optimization, while also reducing unnecessary search time and improving efficiency; Formula (11) incorporates It represents obstacles in the search direction and plays the role of predicting obstacles. By increasing the cost of moving close to obstacles, this mechanism actively guides the path search to avoid nearby obstacles as early as possible.

5. The multi-intelligent AUVs underwater path planning method considering ocean currents according to claim 4 is characterized in that: Population 2 goes through the reproduction steps, the specific process is as follows: the reproduction area is changed, the specific formula is as follows Ω'=max(S b ×(1-K),Ω) (13) Formula (13) represents the lower limit of the breeding area, S b represents the optimal position in the population; Formula (14) represents the upper limit of the breeding area; ζ i (a) and ζ j (a) is the effect of ocean current on individual position a calculated according to the ocean current formula; β is a tuning parameter used to control the degree of influence of ocean currents on the boundary scaling factor; The parameter K in formulas (13) and (14) is formula (15); The adaptive factor K designed in this way can reduce the boundary scale factor when the ocean current has a greater impact, and increase it when the ocean current has a smaller impact, so that the algorithm can adjust the search range more flexibly during the search process, which helps to better find the global optimal solution; Formula (16) represents the position update formula in the reproduction step of population 2.

6. The multi-intelligent AUVs underwater path planning method considering ocean currents according to claim 5 is characterized in that: Population 3 goes through the foraging step. The specific process is as follows: the boundary selection strategy is improved and the ocean current adaptive factor is added to adapt to the influence of ocean currents on path planning. The specific formula steps are as follows: Ω=max(S b* ×(1-K),Ω) (17) Formula (17) represents the lower limit of the breeding area; Formula (18) represents the upper limit of the breeding area, S b* represents the local optimal position, that is, the optimal position in the current population; Formula (19) represents the position update formula in the foraging step of population 3; The adaptive adjustment of the boundary scale factor allows the algorithm to narrow the search range when it is affected by strong ocean currents to avoid over-exploration, and to expand the search range when the ocean currents are less affected to increase exploration. This flexibility and adaptability helps the algorithm better adapt to different search environments and improve the efficiency and accuracy of the search. The influence of ocean currents may cause individuals to deviate from the target in the search space. If the search range is too large, resources may be wasted in invalid search space. On the contrary, if the search range is too small, potential global optimal solutions may be missed. Therefore, by adjusting the boundary scale factor, the search range can be controlled so that the algorithm can focus more on the area that may contain the optimal solution. Appropriate adjustment of the search range can help to better discover the global optimal solution. When the influence of ocean currents is small, increasing the search range can increase exploration and help to find the global optimal solution more comprehensively in the search space. When the influence of ocean currents is large, narrowing the search range can avoid falling into the local optimal solution and improve the efficiency of global search.

7. The method for underwater path planning of multi-intelligent AUVs considering ocean currents according to claim 6, characterized in that: Population 4 steals by avoiding harm and approaching benefits. The idea is that the AUV is close to the optimal position and away from the worst position. The specific process of formula (20) is as follows: s i (t+1)=s b +H×Z×((P b -|s i (t)|)-(P w -|s i (t)|)+(s b -|s i (t)|)-(s w -|s i (t|)) (20) By subtracting the worst solution (P w -|s i (t)|) and the global worst solution (s w -|s i (t)|), making the overall formula closer to the optimal solution and away from the worst solution.

8. The method for underwater path planning of multi-intelligent AUVs considering ocean currents according to claim 7, characterized in that: The AUV group is divided into four groups as a T-shaped unit to search for the optimal position. Group 1 integrates the rolling action and guided dancing mechanism of ocean current factors, groups 2 and 3 respectively perform ocean current adaptive reproduction and ocean current adaptive foraging through ocean current adaptive boundaries, and group 4 performs the path point of finding the optimal point through stealing by avoiding harm and seeking benefits. Adaptive ratio refers to adapting to the changes in ocean currents. The ratio can be dynamically adjusted according to the impact of ocean currents to adapt to ocean currents, making the overall method more efficient. The specific formula is as follows: The four proportional coefficients are shown in formula (21): They represent AUV population 1, AUV population 2, AUV population 3, and AUV population 4 respectively; The ocean current adaptive factor is shown in formula (22): η represents the change of ocean current, k1 and k2 are adjustment parameters; 0<ω<1, by adjusting the adaptive factor ω, the proportional coefficient of each population AUV is dynamically adjusted to adapt to the change of ocean current and optimize the overall performance.

9. The method for underwater path planning of multi-intelligent AUVs considering ocean currents according to claim 8, characterized in that: The optimal path points in each area are integrated and decided, and an evaluation mechanism that conforms to the ocean current is introduced to screen out the best individuals according to the satisfaction, and finally the optimal path is obtained by summarizing; The evaluation mechanism function of following the ocean current is shown in formula (23): s represents the coordinates of the path point, V represents the ocean current velocity vector, represents the direction of the unit vector pointing to the path point s, d(s) represents the distance from the path point s to the nearest ocean current line, and c vortex (s) represents the vortex influence factor at the path point s, |V| represents the magnitude of the ocean current speed; The first part considers the consistency of the direction of the waypoints with the direction of the ocean current and normalizes it according to the magnitude of the ocean current speed; The second part considers the distance from the waypoint to the nearest ocean current line, the closer the distance, the higher the adaptability; The third part considers the influence of eddies on the path points. The eddy influence factor is related to the speed of the ocean current, which reflects the influence of eddies on path planning. f * (s)>0, indicating that the direction of path point s is consistent with the direction of the ocean current, and the adaptability is high; f * (s) = 0, indicating that the direction of path point s is perpendicular to the direction of the ocean current, and the adaptability is medium; f * (s)<0, indicating that the direction of path point s is opposite to the direction of the ocean current and has low adaptability.

10. The method for underwater path planning of multi-intelligent AUVs considering ocean currents according to claim 2, characterized in that: The fitness value is calculated based on the objective function. The objective function formula is as follows: Position representation formula: G={G g |g=1,2,…,∞} (1) F={F f |f=1,2,…,∞} (2) H={H h |h=1,2,…,∞} (3) T={T t |t=1,2,…,∞} (4) These four formulas represent the position sets of four AUV populations, each set contains an infinite number of possible position points, which is set to 30 AUVs per population in the patent of this invention; This set of formulas (5) defines the optimal position Pb and the worst position Pw, as well as the fitness function f(P x ) calculation method; Formula (al) represents the shortest path; Formula (a2) represents the minimum energy consumption; The two formulas of the objective function define the two main goals of the optimization problem: the shortest path and the minimum energy consumption according to the ocean current adaptation, where C * It is the adaptive ocean current influencing factor; Constraints: Position constraints: Speed ​​Constraints: Acceleration constraints: The above formulas define various constraints for AUV motion. The meanings of the variables are: Gg, Ff, Hh, Tt: represent the individual positions in the four AUV populations; Px, Py: AUV position point; f(Px): fitness value of position Px; ω1, ω2: weight coefficients, which can be used to balance the relationship between the shortest path and energy consumption according to the adjustment parameters; Minimum path length; ∈ min :Minimum energy consumption; d j : The length of the jth segment of the path; Energy consumption of the i-th path; ζi(α),ζj(α): ocean current influence factors, including the position, radius and size of the vortex, that is, the generated path points must meet this limit; s(x,y): 2D position coordinates of AUV; vjAUV,ajAUV: speed and acceleration of AUV; rmax, amax: maximum speed and acceleration limits; The fitness function f(Px) is directly associated with the objective function F (shortest path) and ε (minimum energy consumption); The four AUV populations (G, F, H, and T) represent different search strategies, each targeting specific environmental factors or constraints; The objective functions (a1) and (a2) directly correspond to Fmin and εmin in the fitness function; Constraints (a3)-(a6) ensure that the motion of the AUV meets practical physical and environmental constraints; The field-shaped search strategy divides the entire search space into small areas, applies different AUV population strategies in each area, and judges the optimal path point by the fitness value.

Images