A method for distributed coverage path planning of multiple unmanned boats based on electronic nautical charts

By using the area division algorithm of electronic chart and the minimum spanning tree template method in the multi-unmanned boat dispersed coverage path planning, the initial position and path turning rate of the unmanned boat are optimized, and the uneven area division in the prior art is solved. The problem of high path turning rate has achieved efficient multi-unmanned boat coverage path planning.

CN115437380BActive Publication Date: 2024-11-15DALIAN MARITIME UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211125165.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-09-15
Publication Date
2024-11-15
Estimated Expiration
2042-09-15

AI Technical Summary

Technical Problem

The existing multi-unmanned boat scattered coverage path planning methods have shortcomings in area division and path optimization, resulting in uneven coverage area of ​​unmanned boats, long task time and high energy consumption.

Method used

The electronic chart-based area division algorithm (DARP) and the minimum spanning tree template method are used to optimize path planning, combine the roulette selection method to optimize the initial position and the spanning tree algorithm to optimize the path turning rate, and build the hardware through the Raspberry Pi and the flight control source code. The ring simulation platform performs path planning.

Benefits of technology

The division of multiple unmanned boat areas into sub-regions with approximately equal areas and regular shapes is realized, which reduces the number of turns and energy consumption of unmanned boats, and improves mission efficiency and coverage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115437380B_ABST
    Figure CN115437380B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for planning the dispersed coverage paths of multiple unmanned boats based on electronic nautical charts, comprising the following steps: processing of electronic nautical chart information; establishment of a grid map and obstacle expansion; dispersed coverage of regions, obtaining an optimized set of starting positions of multiple unmanned boats according to a roulette selection algorithm; and planning the coverage paths of multiple unmanned boats based on a minimum spanning tree. The user of the present invention can use the electronic nautical chart to manually select different mission areas to perform coverage path planning operations for unmanned boats for different mission scenarios. The present invention designs a roulette selection algorithm for the regional division of multiple unmanned boats, and provides an optimized set of initial positions for the regional segmentation of multiple unmanned boats, so that the sub-regions after the regional division can maintain the regularity of the regional shape on the basis of maintaining approximately equal areas. The present invention optimizes the algorithm for coverage path planning, and optimizes the turning rate of the unmanned boat using a spanning tree template method, so that the coverage path has fewer turns.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of shipbuilding and marine engineering, and in particular to a method for planning dispersed coverage paths of multiple unmanned boats based on electronic nautical charts. Background Art

[0002] The exploration, development and utilization of the ocean are the current development priorities of various countries. Unmanned boats have become an important tool for people to explore and develop the ocean because of their small size, light weight and autonomy. Since a single unmanned boat can only complete limited tasks and is inefficient, in order to solve the above problems, the coordination of multiple unmanned boats is mostly used. In the path planning of dispersed coverage of multiple unmanned boats, most studies do not design the dispersed coverage path planning of multiple unmanned boats based on electronic chart information, so there is a problem of dispersed coverage path planning of multiple unmanned boats based on electronic charts; at the same time, in the path planning stage, there are problems such as regional division and path optimization, so it is necessary to optimize the methods of regional division and path optimization. Therefore, it is of great significance to study the dispersed coverage path planning and control problem of multiple unmanned boats based on electronic charts.

[0003] Domestic and foreign scholars have conducted a lot of relevant research on the coverage path planning of multiple unmanned boats, and proposed many control methods, including the plowing round-trip planning based on regional segmentation, the rotation path planning based on the Chinese character, and the coverage path planning based on the spanning tree. From the perspective of the design of the coverage path planning algorithm, the existing multi-unmanned boat coverage path planning method still has the following shortcomings: First, the existing unmanned boat coverage path planning algorithms are mostly concentrated on the development of a single unmanned boat. In practical applications, the use of multiple unmanned boats to cover the area will greatly improve the task efficiency. Therefore, it is necessary to study the distributed coverage path planning of multiple unmanned boats. Second, the existing multi-unmanned boat path planning algorithm does not solve the problem of regional division of distributed coverage. In the case of regional coverage, it is impossible to make the coverage area of ​​each unmanned boat approximately equal, and the overall mission time is long. This is also an important research direction for the regional coverage of multiple unmanned boats. Third, the existing research on the distributed coverage path planning of multiple unmanned boats does not perform path optimization operations for each sub-area after division. Different sub-areas use the same spanning tree expansion direction, which will increase the energy consumption of multiple unmanned boats in performing distributed coverage path tracking control. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention proposes to establish a multi-unmanned boat dispersed coverage path planning method based on electronic nautical charts for unmanned boat clusters, optimized area division and path optimization.

[0005] The technical ideas of the present invention are as follows: first, the regional information of the selected electronic chart mission area is acquired, the longitude and latitude information of the chart is converted into bitmap information by the coordinate conversion method, and the binarization processing and rasterization processing of the electronic chart data are completed; secondly, for the problem of decentralized coverage planning of multiple unmanned boats, a regional partitioning (DARP) algorithm is used to divide the coverage area, and a roulette partitioning algorithm is proposed to optimize the initial position selection problem in the process of dividing the coverage area of ​​multiple unmanned boats; then, for the method based on the minimum spanning tree (MST), a template method optimization algorithm is adopted to reduce the turning rate of the generated path and save the time and energy consumption of the unmanned boat to perform the task; finally, for the decentralized coverage path planning of multiple unmanned boats, a combination of Raspberry Pi and flight control source code is used to build a hardware-in-the-loop simulation platform for simulation experiments, and the path planning task of multiple unmanned boats for a coastal mission area is completed.

[0006] The technical solution of the present invention is as follows: A method for planning a distributed coverage path for multiple unmanned boats based on an electronic nautical chart comprises the following steps:

[0007] A. Process electronic chart information and create raster maps

[0008] A1. Processing of electronic chart information

[0009] Extract the required chart data of the mission area from the electronic chart, including latitude and longitude information and sea and land information. The sea and land information refers to the information that each longitude and latitude coordinate position is sea or land, and package the information and send it to the coverage path planning algorithm program;

[0010] A2. Grid map creation and obstacle expansion

[0011] First, according to the information of the received electronic chart, the ocean information, land information, length and width information of the mission area are extracted by traversal, and a 0-1 binary grid map with a pixel size of m*n is obtained, where m is the height of the grid map; n is the width of the grid map; the value of the grid is 0, which represents land, that is, an obstacle grid; the value of the grid is 1, which represents the ocean, that is, a free grid for navigation. Each grid is a square grid with a side length of d. At this time, the grid map obtained has an irregular boundary, so the grid map obstacles are expanded before executing the path planning algorithm. Since the planning algorithm is designed to perform a spanning tree algorithm for a 2*2 grid, the obstacles in the figure are expanded according to the 2*2 grid, that is, starting from the origin, there are obstacles in the four adjacent grids, and this 2*2 grid is designed as an obstacle grid. The obstacle expansion algorithm obtains a binary grid map of obstacle expansion. Extract all obstacle grids from the grid map and aggregate all grid coordinates in the obstacle area into Points obs :

[0012] Point obs ={p o1 , p o2 ,…,p og}, g = 1, 2, 3…i

[0013] Among them, p o1 =[x obs1 ,y obs1 ] T 、p o2 =[x obs2 ,y obs2 ] T ,…,p og =[x obsg ,y obsg ] T are the 1st, 2nd, …, gth obstacle grids respectively, [x obs1 ,y obs1 ] T ,[x obs2 ,y obs2 ] T ,…,[x obsg ,y obsg ] T They are the horizontal and vertical coordinates of the 1st, 2nd, …, gth obstacle grids respectively, and i is the number of obstacle grids.

[0014] The free grid coordinates for the unmanned boat to navigate are collected as Point free :

[0015] Point free ={p f1 , p f2 ,…p fk}, k = 1, 2, 3…j

[0016] Among them, p f1 =[x free1 ,y free1 ] T 、p f2 =[x free2 ,y free2 ] T ,…,p fk =[x freek ,y freek ] T are the 1st, 2nd, …, gth free grids, [x free1 ,y free1 ] T ,[x free2 ,y free2 ] T ,…,[xfreek ,y freek ] T They are the horizontal and vertical coordinates of the 1st, 2nd,…, kth free grids respectively, and j is the number of free grids.

[0017] B. Regional Dispersed Coverage

[0018] B1. Roulette selection method is based on the length and width of the task area. An inscribed circle is drawn in the task area, and the center of the inscribed circle is the center position of the task area. After the inscribed circle is drawn, points are taken on the inscribed circle according to the number of unmanned boats, and the points are evenly distributed on the circle, thereby ensuring that the initial positions have no overlap and are evenly distributed in the task area. When the grid taken is in the obstacle area, the distance calculation method is used to find the nearest free grid to this position and use it as the initial point of this grid. The series of initial points obtained in this way ensure that the sub-regions are approximately equal in area when the region is segmented, and also ensure the regularity of the segmented region.

[0019] The task area height H of the roulette partitioning method is set to the height m of the grid map, the task area width W of the roulette partitioning method is set to the width n of the grid map, and the number of unmanned boats is num, where num is a positive integer.

[0020] Set the diameter of the wheel to the smaller of the mission area's height and width:

[0021] R = min(H, W)

[0022] p o =[x o ,y o ] T

[0023] [x o ,y o ] T =[H / 2, W / 2] T

[0024] θ=2π / r

[0025] where p o =[x o ,y o ] T is the center point of the roulette wheel, [x o ,y o ] T is the horizontal and vertical coordinates of the center point, θ is the angle of the wheel evenly divided, and the initial position of the r∈num unmanned boat needs to traverse the wheel according to the number of unmanned boats:

[0026] When r is in num:

[0027] xr =x o +(R / 2)*cos(r*θ)

[0028] y r =y o +(R / 2)*sin(r*θ)

[0029] If p r =[x r ,y r ] T is the selected initial position on the wheel, [x r ,y r ] T is the horizontal and vertical coordinates of the initial position, if p r Point free Then p r Add the unmanned boat starting point collection Point start ,use" " operation refers to adding elements to a collection:

[0030] If [x r ,y r ] T ∈Point free :

[0031]

[0032] If pr at this time does not belong to Point free , then find Point free In and p r The closest point is used as the initial point of this sub-region. free For the zth grid in, each grid is related to p r It can be expressed as:

[0033] When z is at Point free China Times:

[0034]

[0035]

[0036] p fz =[x freez ,y freez ] T ,[x freez ,y freez ] T ∈Point free

[0037] In the formula, p fz =[xfreez ,y freez ] T is the zth free grid, [x freez ,y freez ] T is the coordinate of the zth free grid, and distance is the distance of all free grids p fz With p r The distance set is selected with p r The distance is the minimum value d minr Point p minr As the starting point of this sub-area:

[0038] d minr =min{d1, d2, d3, ... d j}

[0039]

[0040] p minr =[x minr ,y minr ] T ,[x minr ,y minr ] T ∈Point free

[0041]

[0042] where p minr =[x minr ,y minr ] T is the calculated starting point, [x minr ,y minr ] T The horizontal and vertical coordinate values ​​of the starting point.

[0043] B2. Obtain the optimized starting position set Point of multiple unmanned boats according to the roulette selection algorithm start , based on Point start The entire area is divided into num sub-areas of approximately equal area using the region segmentation algorithm. For the rth sub-area, the set of free grids in the sub-area is set to Point freer , Point freer ∈Point free .

[0044] C. Multi-unmanned boat coverage path planning based on minimum spanning tree

[0045] C1. Once each unmanned boat is assigned its sub-area in the entire area, a coverage path planning problem for a single unmanned boat is planned for each sub-area. For the path generation problem of each sub-area, the minimum spanning tree algorithm is used. That is, a minimum spanning tree is constructed for each mission area, and then a path is generated to bypass it. The minimum spanning tree algorithm generates a coverage path, with l as the minimum distance for the unmanned boat to move, and makes the following assumptions:

[0046] First, the initial position of the r∈num unmanned boat is set to p startr :

[0047] p startr =[x startr ,y startr ] T ,[x startr ,y startr ] T ∈Point freer

[0048] [x startr ,y startr ] T is the horizontal and vertical coordinates of the initial position of each sub-region, Point freer A collection of free grids for each sub-region

[0049] Secondly, the unmanned boat is only allowed to move in the four directions of up, down, left, and right of the grid where the unmanned boat is located. Set the set of path points of the rth unmanned boat as Point shipr , set the previous position of the rth unmanned boat to p lastr :

[0050] p lastr =[x lastr ,y lastr ] T ,[x lastr ,y lastr ] T ∈Point freer

[0051] where [x lastr ,y lastr ] T is the horizontal and vertical coordinates of the previous position of the r-th unmanned boat, and the current position of the r-th unmanned boat is set to p nowr :

[0052] p nowr =[x nowr ,y nowr ] T ,[x nowr ,y nowr ] T∈Point freer

[0053] where [x nowr ,y nowr ] T is the horizontal and vertical coordinates of the current position of the rth unmanned boat, and the initial settings are as follows:

[0054] p nowr =p startr

[0055] p lastr =p nowr

[0056] The number of free grids j in the rth sub-area for the rth unmanned boat r Traverse the free grid area and set the number of loops count = 0:

[0057] If count < j r :

[0058] Each cycle selects in a counterclockwise direction, that is, searching in the right, up, left, and down directions.

[0059] First, if the right grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the right grid:

[0060] If [x nowr +l,y nowr ] T ∈Point freer The conditions are met:

[0061] [x nowr ,y nowr ] T =[x nowr +l,y nowr ] T

[0062]

[0063] If the right grid of the rth unmanned boat is an obstacle grid, and the upper grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the upper grid:

[0064] If [x now r,y nowr +l] T ∈Point freer The conditions are met:

[0065] [x nowr ,y nowr ] T =[x nowr ,ynowr +l] T

[0066]

[0067] If the right and upper grids of the rth unmanned boat are obstacle grids, and the left grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the left grid:

[0068] If [x nowr -l,y nowr ] T ∈Point freer The conditions are met:

[0069] [x nowr ,y nowr ] T =[x nowr -l,y nowr ] T

[0070]

[0071] If the right, upper and left grids of the rth unmanned boat are obstacle grids, and the lower grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the lower grid:

[0072] If [x nowr ,y nowr -l] T ∈Point freer The conditions are met:

[0073] [x nowr ,y nowr ] T =[x nowr ,y nowr -l] T

[0074]

[0075] At this time, it has moved to the next grid, so the number of loops count = count + 1. If count < j r Then continue the loop, if count = j r The loop ends.

[0076] Finally, the spanning tree generated by the above algorithm is represented by p startr Around Point in counterclockwise direction from the starting point shipr The waypoints generate a closed path.

[0077] C2. The optimization algorithm based on the template method optimizes the number of turns of the coverage path of multiple unmanned boats in the sub-area. In order to deal with this problem, four different combinations of minimum spanning trees at the upper, lower, rightmost and leftmost levels are tested for each sub-area, and finally the optimal spanning tree path is selected by calculating the number of turns of the path of each template.

[0078] Furthermore, the grid side length d is 1 to 2 m.

[0079] Compared with the prior art, the present invention has the following beneficial effects:

[0080] First, compared with the existing unmanned boat path planning algorithm, the present invention develops a multi-unmanned boat distributed coverage path planning platform based on the electronic nautical chart. Users can use the electronic nautical chart to manually select different mission areas to perform unmanned boat coverage path planning operations for different mission scenarios. Therefore, the application scenarios of the present invention are broader, which is conducive to platform transplantation and secondary development.

[0081] Second, compared with the existing multi-unmanned boat dispersed coverage path planning algorithm platform, the present invention designs a roulette selection algorithm for the regional division of multiple unmanned boats, which provides an optimized initial position set for the regional segmentation of multiple unmanned boats; then the regional division algorithm is used to divide the multi-unmanned boat mission area into sub-areas with approximately equal areas and relatively regular shapes. After the regional division, each unmanned boat is responsible for covering a specific part of the grid. Having exclusive sub-areas ensures the generation of non-intersecting paths, allowing multiple unmanned boats to operate safely at the same time and avoiding potential collisions between them. The proposed roulette division method uses a roulette-like method to optimize the initial positions of the unmanned boats before regional division, so that the initial positions of each unmanned boat are evenly dispersed in various positions of the region, so that the sub-areas after regional division can maintain the regularity of the regional shape on the basis of maintaining approximately equal areas.

[0082] Third, compared with the existing multi-unmanned boat distributed coverage path planning algorithm platform, the present invention optimizes the coverage path planning algorithm and uses the spanning tree template method to optimize the turning rate of the unmanned boat, so that the coverage path has fewer turns. On the one hand, the total time to complete the task is reduced, and on the other hand, the energy consumption of the unmanned boat is reduced by reducing the number of turns of the unmanned boat. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 It is a flow chart of the present invention.

[0084] Figure 2 The image is rasterized.

[0085] Figure 3 Augment the processed image for obstacles.

[0086] Figure 4 Schematic diagram of the roulette wheel selection method.

[0087] Figure 5 These are the four templates for the spanning tree template method.

[0088] Figure 6 The paths generated around the spanning tree based on the four spanning tree template methods. DETAILED DESCRIPTION

[0089] The present invention will be further described below in conjunction with the accompanying drawings.

[0090] like Figure 1 As shown, a multi-unmanned boat distributed coverage path planning method based on an electronic nautical chart includes the following steps:

[0091] The first step is to extract the required task area chart data from the electronic chart according to the user's task requirements. Based on the intercepted electronic chart data, the information is globally thresholded using a binarization function, and the intercepted chart is processed into a 0-1 binary map; based on the data information of the 0-1 binary map, the map is rasterized using a rasterization function (such as Figure 2 ); To address the problem of irregular obstacle areas in raster images, an obstacle expansion function is used to expand the obstacle boundaries of the rasterized image (e.g. Figure 3 ).

[0092] The second step is to obtain the optimized starting position set Point of multiple unmanned boats according to the roulette selection algorithm. startr , based on Point startr The entire region is divided into num sub-regions of approximately equal area (such as Figure 4 ).

[0093] In the third step, the regional coverage path planning is based on the minimum spanning tree algorithm. Four different combinations of minimum spanning trees at the upper, lower, rightmost and leftmost levels are tested for each sub-area. Finally, the optimal spanning tree path is selected by calculating the number of path turns for each template. The four templates of the template method are as follows: Figure 5 As shown; a closed loop coverage path is generated around the spanning tree in a counterclockwise direction, and this path is used as the result of the unmanned boat coverage path planning (such as Figure 6 ).

[0094] Before starting planning, the number of unmanned boats to perform the task is determined in advance. The path planning platform is connected to the electronic chart through the communication protocol to obtain the initial state information of the unmanned boat. The area where the task needs to be performed can be obtained by dragging the electronic chart with the mouse, and the initial position information of the unmanned boat can be obtained by clicking the mouse or inputting the longitude and latitude coordinates. Click "Send Information" with the mouse to package the electronic chart information of the task area, the number of unmanned boats required for the path planning of multiple unmanned boats, the position of the unmanned boats, and other information and send them to the path planning platform. After obtaining the packaged data, the path planning algorithm uses a rasterization program to convert the information of the electronic chart into a 0-1 binary map by traversal search, and then obtains the raster map of the corresponding area through the rasterization function. The regional division of the task area is completed by using the regional division program based on the raster map, and then the minimum spanning tree algorithm is used to obtain the planning path of each task sub-area. The total number of grids in the task area is 768, of which 692 are free grids, 76 are obstacle grids, and 22 are expanded obstacle grids. For the simulation of the area division of the four unmanned boats, the four initial positions of the four unmanned boats are obtained by the algorithm as p1[20,24] T , p2[4,24] T 、p3[4,8] T 、p4[20,8] T After the regional division algorithm is performed, the number of grids in each sub-region is 168, 168, 167, and 167, respectively, and the overall coverage rate reaches 97%; before optimization, the number of turns of the unmanned boat in each sub-region is 82, 82, 88, and 86, respectively, and after optimization, the number of turns of the unmanned boat in each sub-region is 60, 70, 68, and 64, respectively, and the number of turns is reduced by 27%, 15%, 23%, and 26%, respectively. From the simulation results, it can be seen that the optimization algorithm greatly reduces the number of turns of the unmanned boat and reduces the energy consumption of the unmanned boat at the turn.

[0095] The present invention is not limited to this embodiment, and any equivalent concepts or changes within the technical scope disclosed by the present invention are included in the protection scope of the present invention.

Claims

1. A method for planning paths for dispersed coverage of multiple unmanned boats based on electronic nautical charts, characterized in that: The following steps are involved: A. Process electronic chart information and create raster maps A1. Processing of electronic chart information Extract the required chart data of the mission area from the electronic chart, including latitude and longitude information and sea and land information. The sea and land information refers to the information that each longitude and latitude coordinate position is sea or land, and package the information and send it to the coverage path planning algorithm program; A2. Grid map creation and obstacle expansion First, according to the information of the received electronic chart, the ocean information, land information, length and width information of the mission area are extracted by traversal to obtain a 0-1 binary grid map with a pixel size of m*n, where m is the height of the grid map; n is the width of the grid map; the value of the grid is 0, which represents land, that is, it is an obstacle grid; the value of the grid is 1, which represents the ocean, that is, it is a free grid for navigation; each grid is a square grid with a side length of d. At this time, the grid map obtained has an irregular boundary, so the grid map obstacles are expanded before executing the path planning algorithm; since the planning algorithm is designed to perform a spanning tree algorithm for a 2*2 grid, the obstacles in the figure are expanded according to the 2*2 grid, that is, starting from the origin, there are obstacles in the four adjacent grids, so this 2*2 grid is designed as an obstacle grid, and the obstacle expansion algorithm obtains the binary grid map of obstacle expansion; all obstacle grids of the grid map are extracted, and all grid coordinates of the obstacle area are set as Point obs : Point obs ={p o1 ,p o2 ,…,p og },g=1,2,3…i Among them, p o1 =[x obs1 ,y obs1 ] T 、p o2 =[x obs2 ,y obs2 ] T ,…,p og =[x obsg ,y obsg ] T are the 1st, 2nd, …, gth obstacle grids respectively, [x obs1 ,y obs1 ] T ,[x obs2 ,y obs2 ] T ,…,[x obsg ,y obsg ] T are the horizontal and vertical coordinates of the first, second, …, and gth obstacle grids, respectively, and i is the number of obstacle grids; The free grid coordinates for the unmanned boat to navigate are collected as Point free : Point free ={p f1 ,p f2 ,…p fk },k=1,2,3…j Among them, p f1 =[x free1 ,y free1 ] T 、p f2 =[x free2 ,y free2 ] T ,…,p fk =[x freek ,y freek ] T are the 1st, 2nd, …, gth free grids, [x free1 ,y free1 ] T ,[x free2 ,y free2 ] T ,…,[x freek ,y freek ] T are the horizontal and vertical coordinates of the first, second, …, kth free grids, respectively, and j is the number of free grids; B. Regional Dispersed Coverage B1. Roulette selection method: Based on the length and width of the task area, an inscribed circle is drawn in the task area, and the center of the inscribed circle is the center position of the task area. After the inscribed circle is drawn, points are taken on the inscribed circle according to the number of unmanned boats, and the points are evenly distributed on the circle, thereby ensuring that the initial positions have no overlap and are evenly distributed in the task area. When the grid taken is in the obstacle area, the distance calculation method is used to find the nearest free grid to this position and use it as the initial point of this grid. The series of initial points obtained in this way ensure that the sub-regions are approximately equal in area when the region is segmented, and also ensure the regularity of the segmented region. Set the mission area height H of the roulette partitioning method to the height m of the grid map, set the mission area width W of the roulette partitioning method to the width n of the grid map, and the number of unmanned boats to num, where num is a positive integer; Set the diameter of the wheel to the smaller of the mission area's height and width: R = min(H, W) p o =[x o ,y o ] T [x o ,y o ] T =[H / 2,W / 2] T θ=2π / r where p o =[x o ,y o ] T is the center point of the roulette wheel, [x o ,y o ] T is the horizontal and vertical coordinates of the center point, θ is the angle of the wheel evenly divided, and the initial position of the r∈num unmanned boat needs to traverse the wheel according to the number of unmanned boats: When r is in num: x r =x o +(R / 2)*cos(r*θ) and r =and o +(R / 2)*sin(r*θ) If p r =[x r ,y r ] T is the selected initial position on the wheel, [x r ,y r ] T is the horizontal and vertical coordinates of the initial position, if p r Point free Then p r Add the unmanned boat starting point collection Point start ,use The operation refers to adding elements to the collection: If [x r ,y r ] T ∈Point free : If p r Not a Point free , then find Point free In and p r The closest point is used as the initial point of this sub-region. free For the zth grid in , each grid and pr can be expressed as: When z is at Point free China Times: p fz =[x freez ,and freez ] T ,[x freez ,and freez ] T ∈Point free In the formula, p fz =[x freez ,y freez ] T is the zth free grid, [x freez ,y freez ] T is the coordinate of the zth free grid, and distance is the distance of all free grids p fz With p r The distance set is selected with p r The distance is the minimum value d minr Point p minr As the starting point of this sub-area: <h2 style=";text-align:left;direction:ltr">d<h2 style=";text-align:left;direction:ltr"> minr <h2 style=";text-align:left;direction:ltr"> = min{d1, d2, d3,...d<h2 style=";text-align:left;direction:ltr"> j <h2 style=";text-align:left;direction:ltr">} p minr =[x minr ,and minr ] T ,[x minr ,and minr ] T ∈Point free where p minr =[x minr ,y minr ] T is the calculated starting point, [x minr ,y minr ] T The horizontal and vertical coordinates of the starting point; B2. Obtain the optimized starting position set Point of multiple unmanned boats according to the roulette selection algorithm start , based on Point start The entire area is divided into num sub-areas of approximately equal area using the region segmentation algorithm. For the rth sub-area, the set of free grids in the sub-area is set to Point freer , Point freer ∈Point free ; C. Multi-unmanned boat coverage path planning based on minimum spanning tree C1. Once each unmanned boat is assigned its sub-area in the entire area, a coverage path planning problem for a single unmanned boat is planned for each sub-area; for the path generation problem of each sub-area, the minimum spanning tree algorithm is used; that is, a minimum spanning tree is constructed for each mission area, and then a path is generated to bypass it; the minimum spanning tree algorithm generates a coverage path, with l as the minimum distance for the unmanned boat to move, and makes the following assumptions: First, the initial position of the r∈num unmanned boat is set to p startr : P startr =[x startr ,and startr ] T ,[x startr ,and startr ] T ∈Point freer [x startr, y startr ] T is the horizontal and vertical coordinates of the initial position of each sub-region, Point freer A collection of free grids for each sub-region Secondly, the unmanned boat is only allowed to move in the four directions of up, down, left, and right of the grid where the unmanned boat is located. Set the set of path points of the rth unmanned boat as Point shipr , set the last position of the rth unmanned boat to p lastr : p lastr =[x lastr ,and lastr ] T ,[x lastr ,and lastr ] T ∈Point freer where [x lastr ,y lastr ] T is the horizontal and vertical coordinates of the previous position of the r-th unmanned boat, and the current position of the r-th unmanned boat is set to p nowr : p nowr =[x nowr, and nowr ] T ,[x nowr ,and nowr ] T ∈Point freer where [x nowr ,y nowr ] T is the horizontal and vertical coordinates of the current position of the rth unmanned boat, and the initial settings are as follows: p nowr =p startr p lastr =p nowr The number of free grids j in the rth sub-area for the rth unmanned boat r Traverse the free grid area and set the number of loops count = 0: If count < j r : Each cycle selects in the counterclockwise direction, that is, searching in the right, up, left, and down directions; First, if the right grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the right grid: If [x nowr +l,y nowr ] T ∈Point freer The conditions are met: [x nowr ,and nowr ] T =[x nowr +l,y nowr ] T If the right grid of the rth unmanned boat is an obstacle grid, and the upper grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the upper grid: If [x now r,y nowr +l] T ∈Point freer The conditions are met: [x nowr ,and nowr ] T =[x nowr ,and nowr [+l] T If the right and upper grids of the rth unmanned boat are obstacle grids, and the left grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the left grid: If [x nowr -l,y nowr ] T ∈Point freer The conditions are met: [x nowr ,and nowr ] T =[x nowr -l,y nowr ] T If the right, upper and left grids of the rth unmanned boat are obstacle grids, and the lower grid of the rth unmanned boat is a free grid, the rth unmanned boat moves to the lower grid: If [x nowr ,y nowr -l] T ∈Point freer The conditions are met: [x nowr ,and nowr ] T =[x nowr ,and nowr -l] T At this time, it has moved to the next grid, so the number of loops count = count + 1. If count < j r Then continue the loop, if count = j r Then the loop ends; Finally, the spanning tree generated by the above algorithm is represented by p startr Around Point in counterclockwise direction from the starting point shipr The waypoints generate a closed loop path; C2. The optimization algorithm based on the template method optimizes the number of turns of the coverage path of multiple unmanned boats in the sub-area; in order to deal with this problem, four different combinations of minimum spanning trees at the upper, lower, rightmost and leftmost levels are tested for each sub-area, and finally the optimal spanning tree path is selected by calculating the number of turns of the path of each template.

2. According to claim 1, a method for planning a distributed coverage path for multiple unmanned boats based on an electronic nautical chart is characterized by: The side length d of the grid is 1 to 2 m.