A global route planning method based on ship historical track

Abstract

The application discloses a kind of global route planning methods based on ship historical track, belong to ship planning technical field.This method first collects ship historical track point data, constructs and dynamically updates two-dimensional grid environment model, extracts safe navigation area by grid extension, connectivity analysis etc.;Based on safe navigation area, the optimal safe node path from starting point to target point is searched by constructing safe node network;Construct the route planning model of single safe navigation area, optimize the route in region;Based on the order of optimal safe node path, local route in each safe navigation area is generated in turn, and finally global planning route is formed.The application deeply mines and utilizes ship historical track information, and provides global navigation route in accordance with actual navigation experience for ship under the condition of lacking complete static environment information, improves the safety and rationality of autonomous navigation of ship under complex sea conditions, and is suitable for ship autonomous return and other scenes.

Description

Technical Field

[0001] This invention relates to the field of ship planning technology, and specifically to a global route planning method based on a ship's historical voyage. Background Technology

[0002] With the continuous development of intelligent ship technology, the autonomous navigation capability of ships in complex waters has become a research hotspot. Path planning, as one of the core technologies of autonomous navigation, is directly related to navigation safety, economy, and intelligence level. Common global route planning methods mostly rely on static electronic charts, real-time environmental perception information, or shore-based support systems to generate feasible paths by constructing a global environment model and combining it with obstacle information. However, in actual navigation, ships often face the following challenges: first, static charts cannot reflect dynamic changes such as underwater obstacles and temporary no-navigation zones in a timely manner; second, in environments such as open seas, limited communication, or without shore-based support, the difficulty of acquiring real-time external environmental data increases significantly.

[0003] In contrast, ships accumulate a wealth of historical navigation data during long voyages. This data not only records the actual geographical locations the ships have passed through, but also implicitly contains navigation experience that conforms to the ship's kinematic characteristics and adapts to the local water environment, possessing high reliability and practical value. Therefore, this invention proposes a global route planning method based on ship historical navigation tracks. By mining and utilizing ship historical navigation track information and combining it with the global route planning process, the generated routes not only meet safety and feasibility requirements but also more closely resemble the real environment, thereby improving the rationality and engineering application value of global ship route planning. Summary of the Invention

[0004] This invention provides a global route planning method based on historical ship tracks, aiming to solve the problem of global route planning for ships in complex environments. By analyzing and utilizing historical ship track data, a global navigation route that conforms to actual navigation experience is provided for ships in the absence of complete static environmental information, thereby improving the safety, rationality, and engineering application adaptability of the planned route.

[0005] To achieve the above objectives, this invention discloses a global route planning method based on historical ship tracks, comprising the following steps: S1. Collect historical navigation point data of the ship, construct and dynamically update a two-dimensional grid environment model based on the historical navigation point data, and extract multiple safe navigation areas; S2. Set the latitude and longitude coordinates of the starting point and target point of this voyage mission; S3. Define each safe navigation area extracted in step S1 as a safe node, and construct a safe node network by combining the starting point and target point set in step S2. S4. On the safe node network constructed in step S3, the shortest path search algorithm is used to find the optimal safe node path from the starting node to the target node. S5. For each safe navigation area in the optimal safe node path obtained in step S4, construct a route planning model within a single safe navigation area. S6. Generate a global planned route from the starting point to the destination point based on the optimal safe node path sequence.

[0006] Furthermore, step S1 specifically includes the following steps: S1.1 Select a reference origin and convert the latitude and longitude coordinates of the reference origin into plane rectangular coordinates; convert the latitude and longitude coordinates of the ship's historical track points into local plane coordinates based on the reference origin; S1.2 Construct a grid map, set the grid step size and grid map coverage, map the local planar coordinates of historical track points to grid coordinates, and mark the corresponding grid cells as safe; when the distance between the ship's current position and the reference origin exceeds a preset threshold, clear the current grid map and reinitialize and record it with the current position as the new reference origin; S1.3. Expand the neighborhood of the grid cells marked as safe, perform connectivity analysis on all expanded safe grids, and extract several connected regions formed by adjacent safe grids.

[0007] Furthermore, step S1.2 also includes track continuity guarantee processing: after mapping track points to grid coordinates, searching for marked safe grids in the neighborhood with the current grid as the center, filling the connection between the current grid and the nearest safe grid, and marking the grids through which the connection passes as safe.

[0008] Furthermore, step S3, constructing the secure node network, specifically includes the following steps: S3.1 Perform geometric shrinkage processing on the extracted safe navigation area polygons. When the shape difference between the shrunken polygon and the original polygon is less than a preset threshold, the shrunken polygon is used as the corresponding safe navigation area representation. Each safe navigation area corresponds to a safe node, forming a safe area node set. S3.2 Construct a secure node network based on the secure nodes, determine the connection relationships between the nodes, calculate the connection weights between the secure nodes, and update the node network.

[0009] Furthermore, in step S3.1, the Hausdorff distance metric is used to measure the geometric difference between the shrunken polygon and the original polygon.

[0010] Furthermore, step S4 specifically includes the following steps: S4.1. Based on the secure node network, initialize the path cost matrix and path record matrix; S4.2 Select each node in the secure node network as a candidate relay node in turn; S4.3 For any pair of nodes, calculate the cost of the combined route formed through the candidate relay nodes; S4.4 Update the route cost and route record matrix; S4.5. Based on the path record matrix, recursively backtrack from the starting node index to the target node index to generate the optimal safe node path.

[0011] Furthermore, step S5 specifically includes the following steps: S5.1 Determine the starting and ending points of the currently planned safe navigation area; S5.2 Determine whether the line segment connecting any two points satisfies the safety constraints; S5.3. Based on the line connecting the first and last points and boundary constraints, perform route planning within the region, calculate the intersection of the line connecting the first and last points and the boundary of the safe navigation area, and calculate the route between adjacent intersection points based on the region boundary. S5.4. Connect the local routes generated between all intersection pairs in sequence according to the order of the intersection pairs to form the final planned route of a single safe navigation area; S5.5. For the initial planned route obtained within a single safe navigation area, perform local route simplification and overall route simplification based on the head and tail buffer zones to reduce redundant waypoints and lower the complexity of the route polyline.

[0012] The beneficial effects of the above-described technical solution of the present invention are as follows: 1. This invention constructs a grid environment model based on the ship's accumulated historical track data and extracts safe navigation areas. It is applicable to complex waters such as the open sea, areas with limited communication or no shore-based coverage, and significantly enhances the ship's environmental adaptability for autonomous navigation.

[0013] 2. This invention effectively utilizes the actual geographical location information of ships by mapping historical track points to grid safety status, performing connectivity analysis and extracting continuous safety areas. It conforms to the navigation experience of ship kinematics and local water environment, and the planned route has higher reliability and engineering practicality.

[0014] 3. This invention generates a global route that can avoid potential danger areas by expanding the neighborhood of the safety grid, geometrically shrinking the safe navigation area, and judging the safety line segments based on the region overlap ratio. At the same time, by simplifying the head and tail buffers and the overall path, redundant track points and polyline complexity are reduced, making the route smoother and more reasonable, and improving the safety of autonomous navigation of ships.

[0015] 4. By setting a reference origin distance threshold, this invention automatically clears and reconstructs the grid map when a ship enters a new water area, ensuring the consistency between the environmental model and the current active water area; at the same time, the use of track point connection filling process reduces the discontinuity caused by GPS measurement errors, enabling the model to be dynamically updated and adapt to changes in the navigation environment. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] Combination Figure 1 The flowchart shown illustrates a global route planning method based on a ship's historical course, comprising the following steps: 1. Collection of historical ship routes and construction of safe navigation areas; 2. Set the starting point and target point through the integrated display and interaction module; 3. Construct a secure node network based on safe navigation areas; 4. Search for the optimal safe node path from the starting point to the target point based on the safe node network; 5. Construct a route planning model for a single safe navigation area; 6. Generate the final global route based on the optimal safe node path order; The specific details of each step are as follows: 1. The steps of collecting historical ship tracks and constructing safe navigation areas include the following: 1.1 Establishing a two-dimensional raster environment model for storing historical flight tracks 1.1.1 Raster Model Coordinate Transformation To model and calculate historical ship tracks in a two-dimensional planar grid environment, before constructing the grid map, the track points, represented in latitude and longitude, need to be converted into Cartesian coordinates, and a local planar coordinate system needs to be established through coordinate translation. First, the latitude and longitude coordinates are converted into Cartesian coordinates. Then, using a selected reference point within the planning area as the origin, the converted planar coordinates are translated to construct a local planar coordinate system for grid modeling.

[0019] 1.1.1.1 Calculation of Plane Coordinates of the Reference Origin Using the ship's berthing position at the dock as the reference origin, its latitude and longitude coordinates are expressed as follows: ,in Indicates longitude. Representing latitude, the reference point is converted to Cartesian coordinates according to the Mercator projection. The calculation formula is as follows: ; ; in, The radius is the Earth's radius.

[0020] 1.1.1.2 Calculation of Planar Coordinates of Track Points Based on Reference Origin For the historical track points of a ship, their latitude and longitude coordinates are represented as follows: Convert these coordinates to planar coordinates and construct a local planar coordinate system with the reference point as the origin to obtain the local planar coordinates. The calculation formula is: ; ; 1.1.2 Initialize the environment raster map First, set the spatial resolution of the environment raster map, i.e., the raster stride. Centered on the reference origin, a certain spatial range is selected around it as the raster map area, and the coverage diameter of the raster map is set to... Based on the grid step size and grid diameter, the physical dimensions of the grid are converted into the number of discrete grid cells, expressed as: ; in, This indicates that the calculation result is rounded to the nearest integer. Therefore, the dimensions of the raster image can be obtained as follows: ; After determining the grid size, the grid map is initialized by setting the initial value of all grid cells to 1 to indicate that the corresponding area is in an unknown state.

[0021] 1.2 Historical Data Collection and Update of Ship Raster Maps 1.2.1 Mapping of waypoints to grid coordinates Ships are equipped with a Global Positioning System (GPS) to obtain their current latitude and longitude information in real time. Based on the coordinate transformation method described above, latitude and longitude coordinates are converted to Cartesian coordinates. .

[0022] The raster coordinates are defined as follows: The raster position corresponding to the reference origin is taken as the center of the raster map. ; The coordinates of any track point in the raster map are calculated as follows: ; ; The above grid coordinates are rounded down and used as the index position of the track point in the grid map. The corresponding grid cell is then assigned a value of 0 to indicate that the area is a safe zone that the ship has passed through.

[0023] 1.2.2 Special processing mechanism for raster map updates To ensure consistency between the grid map and the vessel's current operating waters, the distance between the vessel's current position and the reference origin is first calculated. If this distance exceeds a preset threshold... When the vessel has entered a new water area, the currently saved raster map and reference origin information are cleared, and the vessel's current position is used as the new reference origin to restart the recording and reconstruction of the historical track raster map.

[0024] After the waypoints are written into the raster map, to reduce the problem of track discontinuity caused by GPS measurement errors or slight jumps in latitude and longitude, adjacent graticets are searched within a certain range around the current waypoint's corresponding graticet. The search radius is determined by the number of graticets. Confirmed. If there is a grid cell with a value of 0 within the neighborhood, select the grid cell closest to the current track point, fill the grid line between them, and set the value of all grid cells passed through by the line to 0 to ensure the continuity of the track in the grid map.

[0025] 1.3 Safe navigation area extraction based on historical track grid map 1.3.1 Safety Grid Extension Based on Track Influence Range Historical navigation tracks appear as linear distributions of discrete grid cells in a grid map. In actual navigation, the area surrounding the track is typically navigable water. To more accurately depict safe navigation areas, the safety grid cells corresponding to historical navigation tracks need to be spatially expanded.

[0026] For grid cells with a value of 0 in the historical track grid, a grid expansion operation is performed according to a preset neighborhood range, and all grid cells in the neighborhood are uniformly marked as 0, indicating that the area also possesses safe navigation attributes. The neighborhood range can be set according to factors such as ship size, positioning uncertainty during navigation, and grid resolution.

[0027] 1.3.2 Safe navigation area extraction based on extended safety grid After expanding the safety grid, grid cells with a value of 0 in the grid map are defined as safety grids. Connectivity analysis is performed on all safety grids to identify several connected regions formed by adjacent safety grids. For each connected region, its outer boundary contour is extracted, and the corresponding boundary grid points are obtained, thus forming multiple safety grid regions. Each safety grid region corresponds to a candidate safe navigation area.

[0028] The safe navigation area based on raster coordinates is converted into a geometric representation in a plane coordinate system, with the center coordinates of the raster map being... The coordinates of the boundary points of the safe zone in the raster map are: The corresponding plane coordinates are calculated as follows: ; ; This provides a spatial representation of the safe navigation area in a two-dimensional plane coordinate system, offering a basic input for subsequent route planning.

[0029] 2. The steps for setting the starting point and target point through the integrated display and interaction module include the following steps: After the integrated display and interaction module completes the task configuration and data loading, the operator can issue a global route planning task start command through this module. The task data includes the latitude and longitude coordinates of the starting point and the target point.

[0030] 3. The step of constructing a safe node network based on a safe navigation area includes the following steps: 3.1 Node-based processing of safe navigation areas For each extracted safe navigation area polygon, geometric shrinkage processing is performed. If the shrunken polygon is still a valid polygon and the difference between its geometry and the original safe navigation area polygon meets the preset conditions, then the shrunken polygon is used as the corresponding safe navigation area representation.

[0031] The geometric shape difference is measured using the Hausdorff distance. A Hausdorff distance between the shrunken polygon and the original polygon is considered a threshold. If the shrinkage process does not significantly affect the overall shape of the safe navigation area, the shrinkage result is considered a valid area.

[0032] The safe navigation areas after the above processing are denoted as: ; in, This represents the number of safe navigation zones. Each safe navigation zone corresponds to a safe node, thus forming a set of safe zone nodes.

[0033] 3.2 Construction of a Secure Node Network 3.2.1 Network Model Initialization Construct a secure node network based on the number of secure nodes, and introduce [the following into the network model]: There are 1 safe node, and two additional nodes are introduced to represent the starting point and the destination point, respectively, thus constructing a network matrix with dimensions of . .

[0034] Each element in the matrix represents the connection cost between the corresponding pair of nodes. During initialization, all elements in the matrix are set to infinity to indicate that no feasible secure connection has yet been established between the nodes.

[0035] 3.2.2 Secure Node Connections and Weight Settings Based on Inter-Regional Distance After initializing the network model, the weights between safe nodes are set based on the distance relationships between safe navigation areas. The distance between two safe nodes is defined as the minimum geometric distance between their corresponding safe navigation areas.

[0036] Iterate through the safe nodes, updating the network weights row by row. Taking a single safe node as an example, calculate the distance between the current safe node and all other safe nodes, and record the minimum distance as . Set certain buffer parameters The distance between regions is no greater than Node pairs are considered as node pairs that can establish a direct connection, and the weights of the corresponding positions in the network are updated to the actual distance between the two nodes.

[0037] The starting node and the target node are processed separately. The distance between the starting node or target node and a safe node is represented as the minimum geometric distance from that point to the corresponding safe navigation area. The distances between the starting point or target point and all safe nodes are calculated, and the corresponding weights in the network are updated according to the same weight update rules as for safe nodes.

[0038] After completing all node connections and weight settings, the constructed weighted network is symmetric to ensure that the connection relationships between nodes meet the requirements of an undirected graph, thus providing a foundation for subsequent shortest node path search.

[0039] 4. The step of searching for the optimal safe node path from the starting point to the target point based on the safe node network includes the following steps: Based on the constructed secure node network, an optimal path search method based on iterative updates of multiple relay nodes is used to solve for the optimal secure node path between the starting node and the target node. Specifically, the steps are as follows: 4.1 Initialization of Path Cost Matrix and Path Record Matrix The secure node network includes The weighted adjacency matrix of the network is represented by the following: (Number of nodes: safe node, starting node, and target node) .

[0040] Construct the path cost matrix Path record matrix The row and column indices of the matrix correspond to the node numbers in the secure node network.

[0041] The initialization rules for the path cost matrix are as follows: ; Route Record Matrix Used to record node pairs The relay node numbers traversed in the optimal route are initially set to null values ​​for all elements.

[0042] 4.2 Selecting candidate relay nodes Following the preset node numbering order, each node in the secure node network is selected sequentially. As a candidate relay node. For the current candidate relay node... Using it as a route split point, for any node pair in the network The route will be updated.

[0043] 4.3 Calculate the cost of combined routes For any pair of nodes Calculate via candidate relay nodes The cost of the combined route is calculated using the following formula: ; in, Indicates from node To the node The cost of the shortest route Indicates from node To the node The cost of the shortest route.

[0044] 4.4 Route Cost and Route Record Matrix Update Compare the combined route cost with the current route cost, and select the option that meets the following conditions: ; Then update the route cost matrix and route record matrix: ; ; Otherwise, the cost of maintaining the original route remains unchanged.

[0045] For all nodes By sequentially executing the above route combination and update steps, each node participates in the path optimization calculation as a candidate relay node, thereby completing the global update of the shortest path cost between any pair of nodes in the secure node network.

[0046] 4.5 Optimal Safe Node Path Generation In the path cost matrix With path record matrix After the update is complete, given the starting node index index of target node The optimal path is traced back iteratively using the path recording matrix.

[0047] when When empty, it indicates a node. With nodes There is a direct optimal connection between them; when When that happens, the path will be split into and The process involves recursively backtracking the path and then concatenating the nodes in sequence to obtain the optimal safe path between the starting point and the target point. ; in, Corresponding to the starting point of the voyage, The corresponding navigation target point is represented by the other nodes, which are the safe nodes to be passed through in sequence.

[0048] 5. The step of constructing a route planning model for a single safe navigation area includes the following steps: After obtaining the optimal safe node path from the starting point to the target point, a route planning model is constructed for each safe navigation area traversed along the path. This model uses the line connecting the first and last planned points as the guiding direction and generates a feasible navigation route with a relatively optimal route length under the constraints of the safe navigation area.

[0049] 5.1 Determine the fore and stern planning points of the safe navigation area Assuming the current security node to be planned is The corresponding safe navigation area is Determine the corresponding planning start point and planning end point respectively: The previous path safety node is The corresponding safe navigation area is The first point of the plan is defined as arrive The closest point is denoted as ; The next path safety node is The corresponding safe navigation area is Then the tail point of the planning is defined as arrive The closest point is denoted as .

[0050] 5.2 Safety Assessment of Line Segments To determine any two points and To determine whether connected line segments meet safety constraints, a safety judgment method based on the region overlap ratio is designed.

[0051] Let the safe navigation area be line segment In the region The effective length within is denoted as The total length of the line segment is denoted as The overlap ratio of the line segments is defined as follows: ; A line segment is considered a safe line segment when the following conditions are met: ; in To offset the effects of numerical calculation errors, a preset safety threshold is set. .

[0052] 5.3 Regional Route Planning Based on Connecting the First and Last Points and Boundary Constraints 5.3.1 Feasibility assessment of connecting the first and last points In safe navigation area Inside, first construct the planning starting point. With the end of the planning The connection: ; Determine whether the line segment meets the safety requirements. If it does, the line is directly used as the planned route within the safe navigation area. Otherwise, subsequent route planning is carried out based on the area boundary constraints.

[0053] 5.3.2 Calculation of the intersection point of the line connecting the first and last points with the region Calculate the line connecting the first and last points The intersection points with the boundary of the safe navigation area are sorted according to their distance from the starting point. The set of all intersection points is as follows: ; in: ; To avoid missing intersection points due to numerical errors, the first point of the planning is calculated and... Distance between points, planning the tail point and The distance between points, if satisfying: ; ; If the first or last point of the plan is not included in the set of intersection points, it is considered to be added to the set. The preset judgment threshold can be set to 0.01. The final route result is: ; 5.3.3 Local route generation based on region boundaries Select adjacent pairs of intersection points from the intersection point set in sequence. And for each pair of intersections, the following route generation strategy is executed: Determine the intersection point line segment If the safety requirements are met, then the line segment is used as the planned route between the intersection pairs.

[0054] If the security requirements are not met, then from the intersection... Two candidate routes are generated along the boundary of the safe navigation area: from the direction of the boundary points. marching to The boundary route is denoted as ; from the boundary points in reverse order marching to The boundary route is denoted as .

[0055] and Divide the region boundary into two segments, calculate the lengths of the two candidate routes for each segment, and select the shorter segment as the intersection point. The planned route between them.

[0056] 5.4 Single-area navigation route generation The local routes generated between all intersection pairs are sequentially spliced ​​together according to the order of the intersection pairs to form a continuous navigation route from the planned starting point to the planned ending point. This route lies entirely within the corresponding safe navigation area and serves as the final planned route for a single safe navigation area. The planned route within the area is represented as follows: ,in For planning route nodes within the region.

[0057] 5.5 Route optimization within the region For the initial planned route obtained within a single safe navigation area, the route is simplified and optimized in a structured manner while meeting safety requirements, so as to reduce redundant waypoints and reduce the complexity of the route polygon.

[0058] 5.5.1 Local route simplification based on head and tail buffers To avoid large turning angles at the beginning and end points of the route within the region, local simplification processing is performed on the beginning and end points of the route respectively.

[0059] 5.5.1.1 Local simplification of the first point Starting point of the route Based on this, select a length not exceeding [a certain value] along the route direction after it. The route interval, and according to the interpolation interval Perform equidistant interpolation on this interval to obtain the set of interpolation points: ; in, distance The farthest, distance recent.

[0060] From distance Starting from the more distant interpolation point, the line segments are evaluated sequentially. Safety. When the safety criteria are met for the first time, this line segment is used to replace the original route. arrive All intermediate route points are used to simplify the route in the initial point area.

[0061] 5.5.1.2 Local simplification of the tail point Tail point The simplified processing method is symmetrical to the local simplification process at the first point. Using the tail point as a reference, a length not exceeding the buffer parameter is selected within the route direction preceding it. The interval, and according to the interpolation interval Perform interpolation. Starting from the interpolation point that is farthest from the tail point, determine the safety of the connection between the interpolation point and the tail point in turn. If the safety constraint is met, replace the original tail route with the connection.

[0062] 5.5.2 Overall route simplification within the region After completing the initial and final local simplifications, the overall structure of the routes within the region is simplified to further remove redundant route points.

[0063] Let the simplified route points be: ; From the starting point of the route Initially, a simplified approach was adopted: verification segment by segment. Specifically, for the current starting point... Select route points sequentially from the end of the route forward. And determine the line segment Does it meet the feasibility requirements? If so, retain the route points. , and delete and All intermediate route points between them. Then, Starting from a new point, the above judgment process is repeated iteratively until the end of the route is reached.

[0064] Through the above iterative simplification process, the optimal navigation route is finally obtained within the region that has the fewest points, the route is continuous and feasible under safety constraints.

[0065] 6. The step of generating the final global route based on the optimal safe node path order includes the following steps: 6.1 Final Global Route Generation After obtaining the optimal safe node path from the starting point to the target point, the local route generation and safety constraint optimization processes in each safe navigation area are executed sequentially according to the safe navigation area corresponding to the safe node. The simplified routes obtained in each area are then sequentially spliced ​​together to finally form a global planned route from the starting point to the target point.

[0066] The global planning route obtained through the above steps is represented in a two-dimensional plane coordinate system as follows: ; The two-dimensional plane coordinates are described with a reference origin. To facilitate practical navigation control and navigation applications, the global navigation route in the form of two-dimensional plane coordinates is converted into latitude and longitude representation.

[0067] The two-dimensional plane coordinates of the reference origin are Earth's radius is Then any point The corresponding formula for calculating latitude and longitude coordinates is: ; ; Through the above coordinate transformation process, the latitude and longitude results of the global navigation route in the geographic coordinate system are obtained. This serves as the final output of the navigation route.

[0068] 6.2 Application Scenarios of Global Navigation Routes The global route planning method based on ship historical tracks can automatically generate executable navigation routes while meeting safe navigation constraints. It is applicable to a variety of practical application scenarios, such as safe navigation route planning in autonomous ship return scenarios and optimal navigation route planning between two points given a starting point and a target point.

[0069] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A global route planning method based on historical ship tracks, characterized in that, Includes the following steps: S1. Collect historical navigation point data of the ship, construct and dynamically update a two-dimensional grid environment model based on the historical navigation point data, and extract multiple safe navigation areas; S2. Set the latitude and longitude coordinates of the starting point and target point of this voyage mission; S3. Define each safe navigation area extracted in step S1 as a safe node, and construct a safe node network by combining the starting point and target point set in step S2. S4. On the safe node network constructed in step S3, the shortest path search algorithm is used to find the optimal safe node path from the starting node to the target node. S5. For each safe navigation area in the optimal safe node path obtained in step S4, construct a route planning model within a single safe navigation area. S6. Generate a global planned route from the starting point to the destination point based on the optimal safe node path sequence.

2. The method according to claim 1, characterized in that, Step S1 specifically includes the following steps: S1.1 Select a reference origin and convert the latitude and longitude coordinates of the reference origin into plane rectangular coordinates; convert the latitude and longitude coordinates of the ship's historical track points into local plane coordinates based on the reference origin; S1.2 Construct a grid map, set the grid step size and grid map coverage, map the local planar coordinates of historical track points to grid coordinates, and mark the corresponding grid cells as safe; when the distance between the ship's current position and the reference origin exceeds a preset threshold, clear the current grid map and reinitialize and record it with the current position as the new reference origin; S1.

3. Expand the neighborhood of the grid cells marked as safe, perform connectivity analysis on all expanded safe grids, and extract several connected regions formed by adjacent safe grids.

3. The method according to claim 2, characterized in that, Step S1.2 further includes track continuity assurance processing: after mapping track points to grid coordinates, with the current grid as the center, search for marked safe grids in the neighborhood, fill the connection between the current grid and the nearest safe grid, and mark the grids through which the connection passes as safe.

4. The method according to claim 1, characterized in that, Step S3, constructing the secure node network, specifically includes the following steps: S3.1 Perform geometric shrinkage processing on the extracted safe navigation area polygons. When the shape difference between the shrunken polygon and the original polygon is less than a preset threshold, the shrunken polygon is used as the corresponding safe navigation area representation. Each safe navigation area corresponds to a safe node, forming a safe area node set. S3.2 Construct a secure node network based on the secure nodes, determine the connection relationships between the nodes, calculate the connection weights between the secure nodes, and update the node network.

5. The method according to claim 4, characterized in that, In step S3.1, the Hausdorff distance metric is used to measure the geometric difference between the shrunken polygon and the original polygon.

6. The method according to claim 1, characterized in that, Step S4 specifically includes the following steps: S4.

1. Based on the secure node network, initialize the path cost matrix and path record matrix; S4.2 Select each node in the secure node network as a candidate relay node in turn; S4.3 For any pair of nodes, calculate the cost of the combined route formed through the candidate relay nodes; S4.4 Update the route cost and route record matrix; S4.

5. Based on the path record matrix, recursively backtrack from the starting node index to the target node index to generate the optimal safe node path.

7. The method according to claim 1, characterized in that, Step S5 specifically includes the following steps: S5.1 Determine the starting and ending points of the currently planned safe navigation area; S5.2 Determine whether the line segment connecting any two points satisfies the safety constraints; S5.

3. Based on the line connecting the first and last points and boundary constraints, perform route planning within the region, calculate the intersection of the line connecting the first and last points and the boundary of the safe navigation area, and calculate the route between adjacent intersection points based on the region boundary. S5.

4. Connect the local routes generated between all intersection pairs in sequence according to the order of the intersection pairs to form the final planned route of a single safe navigation area; S5.

5. For the initial planned route obtained within a single safe navigation area, perform local route simplification and overall route simplification based on the head and tail buffer zones to reduce redundant waypoints and lower the complexity of the route polyline.

8. The method according to claim 1, characterized in that, Step S6 also includes a coordinate inverse transformation process: converting the global navigation route points in the plane coordinate system into latitude and longitude coordinates to obtain the final output navigation route.

Images