A dolphin swarm algorithm-based global path planning method for surface unmanned surface vehicle

Abstract

This invention discloses a global path planning method for unmanned surface vessels (USVs) based on the dolphin swarm algorithm, comprising the following steps: Step 1: Modeling the USV path planning task environment using a grid method, designing the path length as the objective function for the global path, and calculating the shortest path; Step 2: Using the standard dolphin swarm algorithm to search for points within the path area to determine possible routes, and determining a suitable route based on all possible routes; Step 3: Setting a strategy for the standard dolphin swarm algorithm, setting a fixed search time to search for a suitable route, and adding a jump step after the search to jump to the optimal position within the suitable route; Step 4: Using a cubic B-spline curve fitting optimization method to smooth the feasible path, obtaining a smooth curve that conforms to the dynamic characteristics of the USV. This planning method has strong universality in both simple and complex environments.

Description

Technical Field

[0001] This invention relates to the field of unmanned surface vessel (USV) path planning technology, specifically to a global path planning method for USVs based on the dolphin swarm algorithm. Background Technology

[0002] As science and technology and shipbuilding technology mature, the automation level of ships is increasing. Traditional ship equipment requiring human operation is gradually being replaced by intelligent unmanned maritime equipment, and fully autonomous unmanned surface vessels (USVs) have become the future trend of the shipbuilding industry. Because they do not require operators, USVs can replace time-consuming and dangerous manual operations when handling complex and hazardous civilian or military missions at sea, offering significant advantages in long-term operations. Therefore, developing more autonomous USV technology and achieving unmanned ship operation has important practical significance.

[0003] A feasible path planning system is the core and key to improving the autonomy of unmanned surface vessels (USVs). Path planning not only needs to avoid obstacles and ensure a smooth and safe arrival, but also needs to meet constraints such as short path distance, good smoothness, and high safety. With the in-depth research and optimization of various path planning algorithms by scholars, most planning algorithms have been widely applied in the fields of mobile robots, drones, and USVs. The essence of path planning technology is to use a wealth of sensors to acquire pose information and the surrounding environment in real time, and then use appropriate optimization algorithms under a map environment model to plan the path trajectory with the optimal fitness value. In short, path planning technology determines the level of intelligence of unmanned equipment.

[0004] Path planning for unmanned surface vessels (USVs) has always been a research hotspot in autonomous navigation. From an algorithmic perspective, evolutionary algorithms can adapt to complex environments, and improvement methods are more diverse, but they are prone to getting trapped in local optima, often finding only suboptimal solutions. Traditional heuristic algorithms can find the shortest path, but as the computational scope increases, they consume more time and memory, lacking adaptability. Research on USV path planning algorithms tends towards the fusion of different algorithms or the application of emerging intelligent algorithms. From the perspective of path practicality, previous smooth route designs mostly involve splicing straight lines and arc segments or directly using spline curves, with little consideration for the dynamic state of the USV. Therefore, it is a problem that needs to be solved to address the global path planning problem of USVs, fully considering the dynamic constraints of USVs, represented by the turning radius, and proposing a safe, efficient, and fast optimal feasible path. Summary of the Invention

[0005] The purpose of this invention is to provide a global path planning method for unmanned surface vessels based on the dolphin swarm algorithm, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a global path planning method for unmanned surface vessels based on the dolphin swarm algorithm, comprising the following steps:

[0007] Step 1: Model the unmanned surface vessel path planning task environment using the grid method, design the path length as the objective function for the global path, and calculate the shortest path;

[0008] Step 2: Use the standard dolphin swarm algorithm to search for points within the path area to determine possible routes, and determine the appropriate route based on all possible routes;

[0009] Step 3: Set a strategy for the standard dolphin swarm algorithm, set a fixed search time to search for a suitable route, and add a jump step after the search is completed to jump to the optimal position in the suitable route;

[0010] Step 3.1: The optimal position of all individual dolphins does not contact the obstacle. The optimal positions of all dolphin neighborhoods are connected in sequence, and the path points are smoothed to obtain the optimal movement route of the unmanned surface vessel from the starting point to the end point.

[0011] Step 4: Use a cubic B-spline curve fitting optimization method to smooth the feasible path and obtain a smooth curve that conforms to the dynamic characteristics of the unmanned surface vessel.

[0012] Furthermore, in step 1, the environment modeling method is to map the environment into a grid map and set the side length of the grid, distinguishing between feasible areas and obstacle areas in the grid map. For ease of calculation, the side length of each small square pixel is set to 1, where white represents feasible areas and black represents obstacles.

[0013] Furthermore, in step 1, the path with the minimum fitness function value is selected as the optimal path. The fitness function is expressed as: F = ω1f1 + ω2f2, where f1 represents the path length, f2 represents the collision detection cost function, ω1 represents the weight of the path length, ω2 represents the weight of the collision detection cost function, and ω1 + ω2 = 1.

[0014] Furthermore, in step 2, the dolphin randomly emits sound waves in several directions, records the new locations searched within a specified time, and calculates the individual optimal solution and the neighborhood optimal solution searched within the maximum search time; the calling and receiving phases are carried out simultaneously to update the neighborhood optimal solution, and in the hunting phase, the new location of the dolphin is obtained.

[0015] Furthermore, in step 3, the unmanned surface vessel (USV) path planning design steps based on the dolphin swarm algorithm are as follows: First, initialize parameters and model the USV path planning task environment according to the grid method; second, initialize the population, randomly generate the population's location information, calculate the fitness of the initial population, and find the optimal position and fitness of individual dolphins; third, update by sequentially selecting a better neighborhood optimal solution through search, jump, call and receive, and hunt steps.

[0016] The trajectory of the neighborhood optimal solution does not contact the obstacle. Based on the point that does not contact the obstacle, the unmanned surface vessel path planning is performed again to realize the iteration of unmanned surface vessel path planning. The positions of all neighborhood optimal solutions are connected to obtain the unmanned surface vessel route.

[0017] Furthermore, in step 4, the expression for the B-spline curve is: Among them, u min ≤u≤u max , represents the curve parameters; p is the curve degree, Q j (j = 0, 1, 2, ..., n) are control points, and the number of control points is n+1. These are B-spline basis functions.

[0018] Compared with the prior art, the beneficial effects of the present invention are:

[0019] (1) Based on the dolphin swarm algorithm, the unmanned surface vessel route is planned. Unlike the traditional path planning algorithms that lack adaptability and are prone to getting trapped in local optima, this method has strong universality in both simple and complex environments and shortens the route planning time.

[0020] (2) By combining grid-based environment modeling, swarm intelligence optimization algorithm and cubic spline curve theory, an improved dolphin swarm algorithm for smooth path planning was designed, which ensured the convergence speed and accuracy of the path optimization algorithm and improved the global search efficiency and mission safety of the unmanned surface vessel in complex environments.

[0021] (3) By improving the strategy through swarm intelligence optimization algorithm, the convergence speed and robustness of unmanned surface vessel route planning are improved. The introduction of cubic spline curve theory comprehensively considers the dynamic characteristics of unmanned surface vessels, making the planned path safer. Attached Figure Description

[0022] Figure 1 This is a flowchart of the global path planning method for unmanned surface vessels of the present invention;

[0023] Figure 2 This is a schematic diagram of the two-dimensional workspace simulation environment of the present invention;

[0024] Figure 3This is a flowchart illustrating the key steps of map environment rasterization in this invention;

[0025] Figure 4 This is an example image of map image binarization generated according to the present invention;

[0026] Figure 5 This is a schematic diagram of the cubic B-spline curve of the present invention;

[0027] Figure 6 This is a simulation diagram of unmanned surface vessel path planning in a simple environment according to the present invention;

[0028] Figure 7 This is a schematic diagram illustrating the trend of fitness function value changes in a simple environment according to the present invention;

[0029] Figure 8 This is a simulation diagram of unmanned surface vessel path planning in a complex environment according to the present invention;

[0030] Figure 9 This is a schematic diagram illustrating the trend of fitness function value changes under complex environments according to the present invention;

[0031] Figure 10 This is a simulation diagram of the improved dolphin swarm algorithm path planning of the present invention;

[0032] Figure 11 This is a simulation diagram of the ant colony algorithm path planning of the present invention;

[0033] Figure 12 This is a simulation diagram of the particle swarm optimization algorithm path planning of the present invention;

[0034] Figure 13 This is a simulation diagram of the path planning of the Grey Wolf algorithm of the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Example:

[0037] Please see Figure 1-13 This invention provides a technical solution: a global path planning method for unmanned surface vessels based on the dolphin swarm algorithm, the process of which is as follows: Figure 1 As shown, the specific steps include the following:

[0038] Step 1: Model the unmanned surface vessel path planning task environment using the grid method, and design an objective optimization function for the global path optimization problem;

[0039] Step 1 specifically includes: studying the path planning problem of unmanned surface vessels, which requires establishing an environmental model of their workspace. In this paper, the environment is mapped into a grid map.

[0040] In a grid map, assuming the location and size of static obstacles are known, for ease of calculation, the side length of each small square pixel is set to 1, where white represents feasible areas and black represents obstacles;

[0041] Two-dimensional space simulation environment such as Figure 2 As shown, the specific implementation steps of map rasterization are as follows: Figure 3 As shown, the static marine environment is captured in Baidu Maps, and the visible area of ​​the map is saved as an image in a local file. The local image is read, converted into a grayscale image, and its size is adjusted. Then, it is binarized to obtain a logical matrix. Finally, the path planning algorithm simulation is completed on the generated raster map.

[0042] Examples of map image binarization generation are as follows: Figure 4 As shown, the fitness function is an indicator for evaluating path quality. In this embodiment, the path length is used as the objective function, and the constraint is the safety of the path, including the unmanned surface vessel (USV) not coming into contact with obstacles and the radar and electromagnetic interference threats during the USV's navigation.

[0043] Assume node P in the map environment i =(x i ,y i ) represents the coordinates of the current path point, node P e =(x e ,y e If ) represents the coordinates of the target point, then the path length function f1 is:

[0044]

[0045] Where, d ie For node P i To the target point P e The distance between;

[0046] Assuming there are K waypoints and Q obstacles in the environment, the collision detection cost function f2 is:

[0047]

[0048] Among them, W i jLet d(i,j) represent the cost function of the i-th path point relative to the j-th obstacle, and let λ represent the distance of the i-th path point relative to the j-th obstacle. j Let represent the weight corresponding to the j-th obstacle, and satisfy λ1+λ2+…λ j =1 (j=1,2,…,Q). Therefore, the fitness function in this embodiment can be expressed as:

[0049] F=ω1f1+ω2f2 (3)

[0050] Here, ω1 represents the weight of the path length, ω2 represents the weight of the collision detection cost function, and ω1 + ω2 = 1. The path length weight factor k is introduced to dynamically and adaptively adjust ω1 to satisfy this condition. In summary, the path with the minimum fitness function value is the optimal path.

[0051] Step 2: Briefly describe the principle of the standard dolphin swarm algorithm. Step 2 specifically includes:

[0052] The dolphin swarm algorithm simulates the dolphin hunting process, mainly including four key stages: searching, calling, receiving, and hunting. In the optimization problem, each individual dolphin i represents a feasible solution, denoted as Dol. i = [x1 x2 … x D ] Τ (i = 1, 2, ..., N), where N is the population size, D is the dimension of the problem being optimized, and x j (j=1,2,…,D) represents the value of the i-th dolphin individual in the j-th dimension. N dolphin individuals Dol are randomly generated according to equation (4). i :

[0053] Dol i,j =F j +rand×(H j -F j ),j=1,2,…,D (4)

[0054] In equation (4): H j and F j represents the upper and lower bounds of the search range for the j-th dimension variable, respectively, and rand is a random number between [0,1]. During the search phase, Dolphin... i The dolphin randomly emits sound waves in M ​​directions. i The new location found at time t is X. ijt It can be expressed by equation (5):

[0055] X ijt =Dol i +V j ×t (5)

[0056] In equation (5): t = 1, 2, 3, ..., T1, V j =[v1 v2 … v D ] Τ (j=1,2,…,M) represents a segment of sound waves emitted by a dolphin in the j-th direction, satisfying ||V j || = speed, where speed is a constant. Find the dolphin Dol within the maximum search time T1. i The optimal solution L found by the individual i and neighborhood optimal solution K i ;

[0057] Where L i satisfy Maximum search radius R1 = T1 × speed. K i Dol, representing dolphins i The optimal position found by itself and other neighboring individuals in this stage K i Update according to formula (6):

[0058]

[0059] The calling and receiving phases occur simultaneously to update the dolphin Dol. i The neighborhood optimal solution K i Sound wave transmission requires time; an N×N transmission time matrix TS is defined. i,j Indicates sound waves from dolphin Dol i To Dol j The remaining transmission time, initially set to the maximum transmission matrix T2 (human-defined), is TS in each algorithm iteration. i,j Decrease by 1, representing the sound wave propagating in one unit of time. In each iteration, when Dol j The neighborhood optimal solution K j Superior to Dol i The neighborhood optimal solution K i And TS i,j Greater than the sound wave propagation time between the two At that time, according to formula (7) for TS i,j Update:

[0060]

[0061] In equation (7): DD i,j Dol means dolphin i With Dolphin j The distance between them, DD i,j =||Dol i -Dol j||,i,j=1,2,…,N,i≠j,A is the acceleration constant that adjusts the speed of sound wave propagation;

[0062] TS i,j After the update is complete, determine whether the TS requirement is met. i,j =0, if the condition is met, it means dolphin Dol j The sound waves emitted have already been detected by the dolphin Dol i Receive, at this time TS i,j Reassign the value to T2 and select K. i and K j The better one updates K i Specifically, as shown in equation (8):

[0063]

[0064] During the hunting phase, according to dolphin Dol... i ,L i ,K i The positional relationship between the three and the size of R1 (where Dol) i With K i Between and K i With L i The distances between them are DK i =||Dol i -K i ||and DKL i =||L i -K i The hunting of dolphins will be conducted in three ways to obtain their new location, NewDol. i to compare it with Dol i The neighborhood optimal solution K i Comparison, if fitness(NewDol) i ) < fitness(K i If K is updated, then K is updated. i K i =NewDol i When DK i ≤R1 indicates that the dolphin Dol i The neighborhood optimal solution K i Within the search range, K is considered to be... i =L i Dolphin i The new position obtained is shown in equation (9):

[0065]

[0066] In equation (9), e > 2 is the radius attenuation coefficient, which is usually taken as 3 or 4. When DK i >R1 and DKi ≥DKL i , indicating dolphin Dol i The neighborhood optimal solution K i Outside the search range, and L i By Dol i Closer to K i Dolphin i The new position obtained is shown in equation (10).

[0067]

[0068] When DK i >R1 and DK i <DKL i , indicating dolphin Dol i The neighborhood optimal solution K i Outside the search scope, and Dol i Compared to L i Closer to K i Dolphin i The new position obtained is shown in equation (11):

[0069]

[0070] Step 3: Improve the standard dolphin swarm algorithm with a strategy, design the unmanned surface vessel path planning process based on the improved dolphin swarm algorithm, realize the safe collision avoidance between the unmanned surface vessel and obstacles, and obtain the global optimal route from the starting position to the ending position.

[0071] Step 3 specifically includes:

[0072] During the search phase, dolphins search the surrounding area by emitting sound waves. The search time is a positive integer constant. If the dolphin does not find a better position when T1=1, it indicates that the probability of finding a better solution in that direction is low. Continuing to extend the search time may result in an invalid search, thereby reducing the search efficiency. Therefore, it is more reasonable to use a fixed search time during the search process.

[0073] After the search phase, the dolphins obtain the neighborhood optimal solution in the standard dolphin swarm algorithm. In the subsequent calling and receiving process, the neighborhood optimal solution is updated according to Equation (8). This update method is conducive to speeding up the convergence speed of the algorithm, but it is also very likely to cause multiple dolphins to have the same neighborhood optimal solution. The subsequent search is based on the neighborhood optimal solution, which will obviously greatly reduce the population differentiation in the subsequent search phase, thus causing the path planning to fall into a local optimum.

[0074] Therefore, to better apply the dolphin swarm algorithm to unmanned surface vessel (USV) path planning problems, a skip step is added after the search phase to discard the K values. iPreserve the individual dolphin Dol during the search phase. i After finding the neighborhood optimal solution, Dol i Jump directly to the neighborhood optimal solution K i The position is such that it maintains population diversity while ensuring convergence speed. The specific update method is shown in equation (12):

[0075] Dol i =K i (12)

[0076] The design steps for global path planning of unmanned surface vessels based on the dolphin swarm algorithm are as follows: First, initialize the parameters, setting the population size M for the dolphin swarm algorithm, the dimension D of the optimization problem, and the upper and lower bounds H of the search range. max and F min Basic parameters such as the maximum number of iterations (MaxIter) and the transmission time matrix (TS) are included. The environment for the unmanned surface vessel (USV) path planning task is modeled using the grid method.

[0077] The second step is population initialization, which involves randomly generating the location information of the population, calculating the fitness of the initial population, and finding the optimal location and fitness of individuals.

[0078] The third step is to search, update the dolphin's new position according to formula (5), select the individual with the smallest fitness as the individual's optimal solution according to formula (3), and complete the update of the neighborhood optimal solution by formula (6).

[0079] The fourth step is to jump to the position of the neighborhood optimal solution according to formula (12);

[0080] The fifth step is calling and receiving. The transmission time matrix TS is updated according to formula (7). Each time the algorithm iterates, each element in TS is reduced by 1. After TS is updated, the best neighborhood optimal solution is selected according to formula (8) to complete the update.

[0081] Step Six: Hunting. (According to DK) i DKL i The size relationship between R1 and the other three corresponds to three cases. Choose one of the formulas (9), (10), and (11) to obtain the new position of the dolphin. By comparing it with the previous better neighborhood optimal solution, select the better neighborhood optimal solution to complete the update.

[0082] In step seven, assuming that the trajectories of the neighborhood optimal solutions of all dolphin individuals do not come into contact with obstacles, we determine whether the path planning constraints (reaching the target point or reaching the maximum number of iterations) are met. If they are met, we jump to step nine; otherwise, we record the optimal position of the neighborhood optimal solution in this iteration and go to step eight.

[0083] The eighth step takes the new position updated in the previous iteration as the starting position of the next iteration and jumps to the third step to continue execution;

[0084] The ninth step connects the positions of the optimal solution in the dolphin's neighborhood in each iteration and smooths the path points to obtain the optimal moving route of the unmanned surface vessel from the starting point to the end point.

[0085] Step 4: Use a cubic B-spline curve fitting optimization method to smooth the feasible path and obtain a smooth curve that better matches the dynamic characteristics of the unmanned surface vessel.

[0086] Step 4 specifically includes:

[0087] The cubic B-spline curve fitting optimization method can be used to optimize the polyline path. The generalized B-spline curve expression is as follows:

[0088]

[0089] In equation (13), u min ≤u≤u max , represents the curve parameters; p is the curve degree, Q j (j = 0, 1, 2, ..., n) are control points, and the number of control points is n+1. The basis functions are B-spline functions. The node vectors are composed of a set of non-decreasing sequences u. i The structure is as follows:

[0090]

[0091] The endpoints of node vector U have p+1 overlaps, with the intermediate nodes increasing sequentially. Here, m = n + p + 1; the number of node vectors is n + p + 2; in this embodiment, u is taken as... min =0, u max =1, that is, 0≤u≤1. Basis functions The de Boor-Cox recurrence relation is typically used. Its principle is to construct a recurrence formula that builds higher-order formulas from lower-order ones. The expression is as follows:

[0092]

[0093]

[0094] In equation (15), Let u represent the p-th B-spline basis function; jLet j represent the j-th parameter point; by convention, 0 / 0 = 0; from the above equation, it can be seen that the B-spline basis function can be solved by the degree p of the spline curve and the node vector U. From equation (3-13), it can be seen that as long as the basic information of the B-spline curve, the control points, the node vectors, and the degree of the curve are obtained, the B-spline curve can be constructed.

[0095] Given control point coordinates Q = [(1, 3); (2, 2); (3, 2); (4, 5); (5, 6); (6, 2); (7, 1)], and nodal vector U = [0, 0, 0, 0, 0.25, 0.5, 0.75, 1, 1, 1, 1], the B-spline curve with p = 3 is as follows: Figure 5 As shown.

[0096] During the optimization process, the motion characteristics of the unmanned surface vessel must also be considered. R is the radius of curvature at each point on the B-spline curve, which must satisfy the following constraints regarding orientation and curvature continuity:

[0097]

[0098] In equation (16), k is the curvature, and y' and y" are the first and second derivatives of each point on the B-spline curve, respectively. The feasibility of an unmanned surface vessel's navigation path must be considered in relation to its minimum turning radius R. min That is, R≥R min The corresponding curvature constraint k max =1 / R min .

[0099] To verify the effectiveness of the method of the present invention, a simulation experiment was conducted on the unmanned surface vessel path planning based on the improved dolphin swarm algorithm proposed above using Matlab2022a.

[0100] The basic parameters and environment of the experimental scenario are set as follows: The unmanned surface vessel (USV) is assumed to travel at a constant speed, with an initial speed of 0 and a speed constant of 1. The starting coordinates of the USV are (1,1), the ending coordinates are (30,30), the population size M is 20, the maximum search time T1 is 1, the maximum transmission time T2 is 1000, the maximum number of iterations MaxIter is 100, the path length weight factor k is 0.02, and the acceleration constant A is 5. The simulation environment is set as a 30×30 square grid as the workspace of the USV, with each grid consisting of a 1×1 grid cell.

[0101] like Figure 6-9 The figure shows simulation diagrams of path planning for unmanned surface vessels (USVs) in both simple and complex environments. The simulation results demonstrate that the generated USV path can plan a globally optimal trajectory that simultaneously satisfies smoothness, economy, and safety, without contacting obstacles in static sea areas. As the number of iterations increases, the average value of the fitness function gradually decreases and approaches the population optimum.

[0102] Depend on Figure 6 It can be seen that in a simple environment, this algorithm converges to the final path in only 11 iterations. Figure 8 It can be seen that when the iteration reaches 19 times in a complex environment, the average value of the fitness function overlaps with the optimal solution. At this point, both the average value of the fitness function and the optimal solution gradually approach zero until they stabilize.

[0103] like Figure 10-13 The diagram shows a comparison of the paths of four algorithms under identical parameter settings and the same complex environment. Simulation results show that the improved dolphin swarm algorithm generates fewer waypoints, and its path planning is significantly better than the other three algorithms. In the path planning of the ant colony algorithm, particle swarm algorithm, and gray wolf algorithm, the unmanned surface vessel's trajectory more or less contacts the edges of obstacles on the grid map. Figure 2 During the quantization process, pixels are added to the obstacle boundaries, which expands the actual obstacle edges on the nautical chart, making the obstacle range at the edge larger than the actual range. Although no collisions have occurred in reality, this still poses a significant safety hazard.

[0104] In the planning method of this embodiment, the unmanned surface vessel can more safely pass through narrow areas to reach the destination. The path trajectory it generates can always maintain a relatively safe distance from surrounding obstacles. After smoothing, such a path can better meet the actual navigation requirements of the unmanned surface vessel.

[0105] Table 1 Comparison of results from different unmanned surface vessel path planning algorithms.

[0106]

[0107] As shown in the table above, compared to traditional ant colony optimization, particle swarm optimization, and gray wolf optimization algorithms, the improved dolphin swarm algorithm achieves shorter path lengths, the best obstacle avoidance performance, and fewer iterations. Therefore, selecting a suitable optimization algorithm can effectively shorten the travel distance, accelerate the convergence speed, and improve the global search efficiency and mission safety of unmanned surface vessels.

[0108] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A global path planning method for unmanned surface vessels based on the dolphin swarm algorithm, characterized in that, Includes the following steps: Step 1: Model the unmanned surface vessel path planning task environment using the grid method, design the path length as the objective function for the global path, and calculate the shortest path; Step 2: Use the standard dolphin swarm algorithm to search for points within the path area to determine possible routes, and determine the appropriate route based on all possible routes; Step 3: Set a strategy for the standard dolphin swarm algorithm, set a fixed search time to search for a suitable route, and add a jump step after the search is completed to jump to the optimal position in the suitable route; Step 3.1: The optimal position of all individual dolphins does not contact the obstacle. The optimal positions of all dolphin neighborhoods are connected in sequence, and the path points are smoothed to obtain the optimal movement route of the unmanned surface vessel from the starting point to the end point. Step 4: The feasible path is smoothed using a cubic B-spline curve fitting optimization method to obtain a smooth curve that conforms to the dynamic characteristics of the unmanned surface vessel. In step 2, the dolphin randomly emits sound waves in several directions, records the new locations searched within a specified time, and calculates the individual optimal solution and the neighborhood optimal solution searched within the maximum search time. The calling and receiving phases are carried out simultaneously to update the neighborhood optimal solution. In the hunting phase, the new location of the dolphin is obtained. In step 3, the unmanned surface vessel (USV) path planning design steps based on the dolphin swarm algorithm are as follows: First, initialize parameters and model the USV path planning task environment according to the grid method; second, initialize the population, randomly generate the population's location information, calculate the fitness of the initial population, and find the optimal position and fitness of individual dolphins; third, update by sequentially selecting a better neighborhood optimal solution through the steps of searching, jumping, calling and receiving, and hunting.

2. The global path planning method for unmanned surface vessels based on the dolphin swarm algorithm according to claim 1, characterized in that: In step 1, the environment modeling method is to map the environment into a grid map and set the side length of the grid, and distinguish between the feasible area and the obstacle area of ​​the grid map.

3. The global path planning method for unmanned surface vessels based on the dolphin swarm algorithm according to claim 2, characterized in that: In step 1, the path with the minimum fitness function value is selected as the optimal path. The fitness function is expressed as: ,in, Represents the length of the path. This represents the collision detection cost function. Weights representing path length. The weights represent the collision detection cost function, and .

4. The global path planning method for unmanned surface vessels based on the dolphin swarm algorithm according to claim 1, characterized in that: In step 4, the expression for the B-spline curve is: ,in, , represents the curve parameters; Let Q be the degree of the curve. j Let j = 0, 1, 2, ..., n be the control points, and the number of control points be n+1. These are B-spline basis functions.

5. The global path planning method for unmanned surface vessels based on the dolphin swarm algorithm according to claim 4, characterized in that: Based on the motion characteristics of unmanned surface vessels Let be the radius of curvature at each point on the B-spline curve, satisfying the following constraints for orientation and curvature continuity: ,in, For curvature, and These are the first and second derivatives of the B-spline curve at each point, respectively. The minimum turning radius of the unmanned surface vessel is... , Corresponding curvature constraint .

6. The global path planning method for unmanned surface vessels based on the dolphin swarm algorithm according to claim 1, characterized in that: The trajectory of the neighborhood optimal solution does not contact the obstacle. Based on the point that does not contact the obstacle, the unmanned surface vessel path planning is performed again to realize the iteration of unmanned surface vessel path planning. The positions of all neighborhood optimal solutions are connected to obtain the unmanned surface vessel route.

Images