A path planning method and system for an amphibious unmanned platform

Abstract

The present invention provides a path planning method and system for an amphibious unmanned platform, which relates to the technical field of path planning, including obtaining sea charts, land structured maps and unstructured satellite maps; constructing an amphibious navigation map according to the sea charts, land structured maps and unstructured satellite maps; determining a sea path according to the amphibious navigation map using an A* algorithm; determining a comprehensive path for land-water conversion and land areas according to the amphibious navigation map and the sea path; and splicing the sea path and the land path to obtain an amphibious planning path. By constructing an amphibious navigation map and determining a comprehensive path for land-water conversion and land areas, path planning can be completed for the sea, land and the junction of land and water at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of path planning, and in particular to a path planning method and system for an amphibious unmanned platform. Background Art

[0002] Map-based ground vehicle path planning and navigation is generally carried out on structured maps. The map road information used for vehicle planning is fixed and belongs to structured roads. For example, the road length and width information generally do not change, and the number of optional roads is generally small. Generally speaking, A*, Dijkstra or other path-finding algorithms can be used to select a shortest path, and the planning speed can be very fast. The planning and navigation of unmanned boats at sea based on nautical charts is carried out on nautical charts. The nautical charts used for unmanned boat path planning have basic nautical chart navigation information, such as islands, reefs, dangerous areas, etc. The path planning algorithm generally converts the nautical chart into a raster map and searches on the raster map, such as using the common A*, Dijkstra algorithm, etc. Amphibious unmanned platform Amphibious unmanned platform can travel in unstructured environments at sea and on land, and also needs to have the function of realizing water-land conversion at the junction of water and land. The existing unmanned boat planning and navigation is carried out on the nautical chart, and can only be planned on the sea surface based on the nautical chart, and cannot realize the land path planning function based on the map, nor can it realize the land-water conversion planning function; the ground vehicle path planning and navigation based on the map is carried out on the map, and can only be planned in the structured map, and cannot be planned on the sea surface based on the nautical chart, and the land-water conversion planning function is also impossible to realize. Therefore, a path planning technology for amphibious unmanned platforms is urgently needed. Summary of the invention

[0003] The purpose of the present invention is to provide a path planning method and system for an amphibious unmanned platform, which can simultaneously perform path planning on the sea surface, land and the interface between land and water.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A path planning method for an amphibious unmanned platform, comprising:

[0006] Obtain nautical charts, land structured maps and unstructured satellite maps;

[0007] constructing an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map;

[0008] Determine a sea path using an A* algorithm according to the amphibious navigation map;

[0009] Determine a comprehensive route for land-water transition and land area based on the amphibious navigation map and the sea route;

[0010] The sea path and the land path are spliced ​​to obtain an amphibious planning path.

[0011] Optionally, constructing an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map specifically includes:

[0012] Determining the waterway and the impassable area at sea in the nautical chart according to the unstructured satellite map;

[0013] Determining roads and land inaccessible areas in the land structured map according to the unstructured satellite map;

[0014] The unstructured satellite map is input into a land-water interface data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water interface area; the land-water interface data extraction model is obtained by training a neural network using the annotated unstructured satellite map;

[0015] An amphibious navigation map is constructed based on the coordinates, altitude and terrain passability of waterways, impassable areas at sea, roads and impassable areas on land, and each point at the junction of land and water.

[0016] Optionally, before acquiring the nautical chart, the land structured map and the unstructured satellite map, the method further includes:

[0017] Obtain unstructured satellite atlas;

[0018] The coordinates, altitude and terrain passability of each point at the land-water junction area in each unstructured satellite map in the unstructured satellite map set are marked to obtain a marked unstructured satellite map set;

[0019] With unstructured satellite atlas as input and labeled unstructured satellite atlas as expected output, the neural network is trained to obtain the land-water connection data extraction model.

[0020] Optionally, determining the integrated path of the land-water conversion and the land area according to the amphibious navigation map and the sea path specifically includes:

[0021] According to the sea route, determining a starting point and a destination point of the integrated route;

[0022] Determine the starting point of the integrated path as the parent node and add it to the integrated path node set;

[0023] Determine a plurality of extension nodes of the parent node using a kinematic equation and the amphibious navigation map;

[0024] Determining the actual moving cost, heuristic estimation cost, node height cost, node traversability cost and node acceleration of each of the extended nodes according to the boundary conditions;

[0025] Determine a weight coefficient group according to the motion state and boundary conditions of the amphibious unmanned platform;

[0026] Determining the costs of the plurality of extended nodes using a cost evaluation function according to the weight coefficient group, and the actual moving cost, heuristic estimation cost, node height cost, node traversability cost and node acceleration of each of the extended nodes;

[0027] Determine whether the distance between the extended node with the minimum cost and the target point is less than the distance threshold, and obtain the judgment result;

[0028] If the judgment result is no, the extended node with the minimum cost is used as the child node of the parent node, the child node and the parent node pointer of the child node are added to the integrated path node set, the child node is used as the parent node and the step of "using the kinematic equation and the amphibious navigation map to determine multiple extended nodes of the parent node" is returned;

[0029] If the judgment result is yes, the target point of the comprehensive path is used as the child node of the parent node, the child node and the parent node pointer of the child node are added to the comprehensive path node set, and the comprehensive path of the water-land conversion and the land area is determined according to the comprehensive path node set.

[0030] Optionally, determining the starting point and the target point of the comprehensive path according to the sea path specifically includes:

[0031] When the direction of the path planning is from sea to land, the end point of the sea path is determined as the starting point of the integrated path, and the target point of the amphibious planning path is determined as the target point of the integrated path;

[0032] When the direction of path planning is from land to sea, the starting point of the amphibious planning path is determined as the starting point of the comprehensive path, and the starting point of the sea path is determined as the target point of the amphibious planning path.

[0033] Optionally, the cost evaluation function is:

[0034] f(n)=w1g(n)+w2h(n)+w3e(n)+w4t(n)+w5a(n);

[0035] Among them, f(n) represents the cost of the nth extended node; g(n), h(n), e(n), t(n) and a(n) represent the actual moving cost, heuristic estimated cost, node height cost, node passable cost and node acceleration of the nth extended node respectively; w1, w2, w3, w4, w5 represent the actual moving cost weight coefficient, heuristic estimated cost weight coefficient, node height cost weight coefficient, node passable cost weight coefficient and node acceleration weight coefficient of the nth extended node respectively.

[0036] A path planning system for an amphibious unmanned platform, comprising:

[0037] A map acquisition module, used to acquire nautical charts, land structured maps and unstructured satellite maps;

[0038] An amphibious navigation map construction module, used to construct an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map;

[0039] A sea path determination module, used to determine a sea path using an A* algorithm according to the amphibious navigation map;

[0040] A comprehensive path determination module, used to determine the comprehensive path of the land-water conversion and the land area according to the amphibious navigation map and the sea path;

[0041] The amphibious planning path determination module is used to splice the sea path and the land path to obtain the amphibious planning path.

[0042] Optionally, the amphibious navigation map construction module specifically includes:

[0043] A marine data extraction unit, used to determine the waterway and the impassable area at sea in the nautical chart according to the unstructured satellite map;

[0044] A land data extraction unit, used to determine the roads and land inaccessible areas in the land structured map according to the unstructured satellite map;

[0045] A land-water connection data extraction unit is used to input the unstructured satellite map into a land-water connection data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water connection area; the land-water connection data extraction model is obtained by training a neural network using the annotated unstructured satellite map;

[0046] The amphibious navigation map construction unit is used to construct an amphibious navigation map based on the coordinates, altitude and terrain passability of each point in the waterway, impassable areas at sea, roads and impassable areas on land, and the water-land junction area.

[0047] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0048] The present invention provides a path planning method and system for an amphibious unmanned platform, which obtains a sea chart, a land structured map and an unstructured satellite map; constructs an amphibious navigation map according to the sea chart, the land structured map and the unstructured satellite map; determines a sea path according to the amphibious navigation map using an A* algorithm; determines a comprehensive path for land-water conversion and land areas according to the amphibious navigation map and the sea path; and splices the sea path and the land path to obtain an amphibious planning path. By constructing an amphibious navigation map and determining a comprehensive path for land-water conversion and land areas, path planning can be completed for the sea, land and the junction of land and water at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0050] Figure 1 A flowchart for constructing an amphibious navigation map in an embodiment of the present invention;

[0051] Figure 2 It is an amphibious navigation map in an embodiment of the present invention;

[0052] Figure 3 Simplified schematic diagram of an amphibious unmanned platform;

[0053] Figure 4 This is a schematic diagram of calculation of expansion nodes of an amphibious unmanned platform in an embodiment of the present invention;

[0054] Figure 5 A schematic diagram of a path search based on an amphibious navigation map in an embodiment of the present invention;

[0055] Figure 6 This is a flow chart of a path planning method for an amphibious unmanned platform in an embodiment of the present invention;

[0056] Figure 7 Schematic diagram of path planning for an amphibious unmanned platform in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] The purpose of the present invention is to provide a path planning method and system for an amphibious unmanned platform, which can simultaneously perform path planning on the sea surface, land and the interface between land and water.

[0059] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] like Figure 6 The present invention provides a path planning method for an amphibious unmanned platform, comprising:

[0061] Obtain nautical charts, land structured maps and unstructured satellite maps;

[0062] Construct amphibious navigation maps based on nautical charts, land structured maps and unstructured satellite maps;

[0063] According to the amphibious navigation map, the A* algorithm is used to determine the sea path;

[0064] Determine the integrated route for land-water transition and land areas based on amphibious navigation maps and sea routes;

[0065] The sea path and the land path are spliced ​​together to obtain the amphibious planning path.

[0066] Specifically, an amphibious navigation map is constructed based on nautical charts, land structured maps and unstructured satellite maps, including:

[0067] Determine navigation routes and impassable areas on the sea in nautical charts based on unstructured satellite maps;

[0068] Determine the roads and land impassable areas in the land structured map based on the unstructured satellite map;

[0069] The unstructured satellite map is input into the land-water interface data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water interface area; the land-water interface data extraction model is obtained by training the neural network using the annotated unstructured satellite map;

[0070] An amphibious navigation map is constructed based on the coordinates, altitude and terrain passability of waterways, impassable areas at sea, roads and impassable areas on land, and each point at the junction of land and water.

[0071] Furthermore, before obtaining the nautical chart, the land structured map and the unstructured satellite map, it also includes:

[0072] Obtain unstructured satellite atlas;

[0073] The coordinates, altitude and terrain passability of each point at the land-water junction area in each unstructured satellite map in the unstructured satellite map set are marked to obtain a marked unstructured satellite map set;

[0074] With unstructured satellite atlas as input and labeled unstructured satellite atlas as expected output, the neural network is trained to obtain the land-water connection data extraction model.

[0075] Preferably, determining the integrated path of the land-water conversion and the land area according to the amphibious navigation map and the sea path specifically includes:

[0076] According to the sea route, determine the starting point and target point of the integrated route;

[0077] Determine the starting point of the integrated path as the parent node and add it to the integrated path node set;

[0078] Using kinematic equations and amphibious navigation maps, multiple extension nodes of a parent node are determined;

[0079] Determine the actual moving cost, heuristic estimated cost, node height cost, node traversability cost and node acceleration of each extended node according to boundary conditions;

[0080] Determine a weight coefficient group according to the motion state and boundary conditions of the amphibious unmanned platform;

[0081] According to the weight coefficient group, as well as the actual moving cost, heuristic estimation cost, node height cost, node traversability cost and node acceleration of each extended node, the cost evaluation function is used to determine the costs of multiple extended nodes;

[0082] Determine whether the distance between the extended node with the minimum cost and the target point is less than the distance threshold, and obtain the judgment result;

[0083] If the judgment result is no, the expansion node with the minimum cost is used as the child node of the parent node, the child node and the parent node pointer of the child node are added to the integrated path node set, the child node is used as the parent node and the step "using the kinematic equation and the amphibious navigation map to determine multiple expansion nodes of the parent node" is returned;

[0084] If the judgment result is yes, the target point of the comprehensive path is used as the child node of the parent node, and the child node and the parent node pointer of the child node are added to the comprehensive path node set, and the comprehensive path of the water-land conversion and the land area is determined according to the comprehensive path node set.

[0085] Specifically, according to the sea route, the starting point and target point of the comprehensive route are determined as follows:

[0086] When the direction of the path planning is from sea to land, the end point of the sea path is determined as the starting point of the integrated path, and the target point of the amphibious planning path is determined as the target point of the integrated path;

[0087] When the direction of path planning is from land to sea, the starting point of the amphibious planning path is determined as the starting point of the comprehensive path, and the starting point of the sea path is determined as the target point of the amphibious planning path.

[0088] Among them, the cost evaluation function is:

[0089] f(n)=w1g(n)+w2h(n)+w3e(n)+w4t(n)+w5a(n);

[0090] Among them, f(n) represents the cost of the nth extended node; g(n), h(n), e(n), t(n) and a(n) represent the actual moving cost, heuristic estimated cost, node height cost, node passable cost and node acceleration of the nth extended node respectively; w1, w2, w3, w4, w5 represent the actual moving cost weight coefficient, heuristic estimated cost weight coefficient, node height cost weight coefficient, node passable cost weight coefficient and node acceleration weight coefficient of the nth extended node respectively.

[0091] The amphibious navigation map used by the amphibious unmanned platform path planning algorithm consists of three basic maps (electronic nautical charts, land structured maps, and unstructured high-definition satellite maps). These three basic maps are provided by special map software, such as Baidu Maps, Google Maps, and ORCA Master G2 (a new touch-screen electronic nautical chart navigation system). The three basic maps are rasterized and finally merged into a rasterized amphibious navigation map. The map construction process is as follows: Figure 1 As shown, the amphibious navigation map includes the position (x, y), altitude z, terrain traversable and structured road information (or sea surface information) route, and the navigation map becomes 5 dimensions.

[0092] The detailed process of establishing an amphibious navigation map is as follows: When on the sea, extract channel information. If it is a dangerous area or a place where the altitude is not the set value, the grid is considered impassable. This part of the raster map information is generated by electronic nautical charts and unstructured high-definition satellites; in the land structured map, extract road information and impassable areas. This part of the raster map information is generated by structured maps and unstructured high-definition satellites; after removing the land structured map area and the sea area, the remaining unstructured raster map information is generated by unstructured high-definition satellite maps. This part of the raster map is established based on the actual location, altitude, and terrain permeability (based on perception and deep learning, etc.). The amphibious navigation map uses different colors to represent different raster information. The amphibious navigation map is like Figure 2 shown.

[0093] The amphibious unmanned platform is mainly driven by four independently controllable track wheels and water propulsion. Figure 3 As shown. Among them: X G O G Y G represents the global coordinate system, X L O L Y L is the body coordinate system, η is the vector composed of its own position (x, y) and yaw angle ψ; v represents its own velocity vector, u, v and r represent the linear velocity along the x and y axes and the angular velocity around the z axis respectively, and the η and v vectors are expressed by the following formulas:

[0094]

[0095] The kinematic equation of the amphibious unmanned platform is expressed by the formula:

[0096]

[0097] The path planning algorithm for amphibious unmanned platforms is divided into two steps:

[0098] Step 1: At sea, the grid map generated by the electronic chart only needs to determine the passability of each grid. At this time, the height cost is calculated based on the sea level. If it is higher than the cost value, it is considered impassable. The marine path planning algorithm uses the ordinary A* algorithm to meet the needs. The generated smooth path oceanPath is as follows Figure 7 shown.

[0099] The second step is to build a rasterized amphibious navigation map based on the land structured map and unstructured high-definition satellite map when the land-water transition is in progress. When planning the land-water transition path, in addition to the five-dimensional information of the map, the amphibious unmanned platform status information must also be added to ensure the safety and reliability of vehicle operation. The amphibious unmanned platform status State includes position, attitude, speed, and whether reverse movement is allowed, that is:

[0100] State=[x,y,ψ,u,v,r,reverse] T

[0101] At this point, all the information that the path planning algorithm needs to consider in the amphibious navigation map is represented by nodes as follows:

[0102] Node=[x,y,z,traversable,route,ψ,u,v,r,reverse,parent] T

[0103] Among them, parent is a pointer to the parent node. In the amphibious navigation map, Node has 11 dimensions. The specific information of each Node is calculated based on the grid information of the amphibious navigation map and the kinematic relationship of the amphibious unmanned platform. That is, according to the initial parent node Node0, each variable in State0 is discretely sampled (the specific parameter value is obtained by the sensor), and the formula (1.1) is used to solve it to get the extended child node Node 11 ,Node 12 , ..., Node 1n State information, and then solve the node Node based on the State information and the amphibious navigation map 1n The remaining information, the obtained extended node has the kinematic State information and ground Figure 5 The extended node calculation diagram of the current node is as follows: Figure 4 shown.

[0104] The basis and boundary conditions for how the path planning algorithm selects the extended node as the current parent node to continue calculation are as follows:

[0105] ①Traversable means traversable calculation, and the cost of intraversable is infinite;

[0106] ② The vertical diagonal altitude z needs to be multiplied by the weight coefficient. The linear motion z remains unchanged. When turning left or right, z = (2~5) × z;

[0107] ③Reverse judgment: to allow reversing, the reversing weight coefficient needs to be multiplied;

[0108] ④ The change rate of altitude z between the parent node and the child node is less than the set threshold Δ z .

[0109] ⑤State posture and velocity sampling space meet the set threshold Δ state

[0110] In the amphibious navigation map, from the initial node Node0 to the nth extended child node Node n , the calculation formula of the extended node cost evaluation function f(n) is as follows:

[0111] f(n)=w1g(n)+w2h(n)+w3e(n)+w4t(n)+w5a(n) (1.2)

[0112] in:

[0113] w1 represents the weight coefficient of the actual moving cost g(n) from the starting point to the node n. w1 and g(n) are consistent with the description of the A* algorithm.

[0114] w2 represents the weight coefficient of the heuristic estimated cost h(n) from node n to end point G. w2 and h(n) are consistent with the description of A* algorithm;

[0115] w3 represents the weight coefficient of the node height cost e(n), which is related to z and reverse parameters (if reverse is TRUE, then w3 = w3 / 2), and e(n) represents the amphibious navigation map height cost value z (i.e. boundary condition ②);

[0116] w4 represents the weight coefficient of the node traversability cost t(n), t(n) represents the traversable value of the map traversability parameter. The larger w4 is, the stronger the traversability is (determine the specific value according to actual needs);

[0117] w5 represents the weight coefficient of the node acceleration a(n), which is determined by the rate of change of the velocity between the current state State and the previous State of the platform. The larger w5 is, the smoother the calculated path is (the specific value is determined according to actual needs).

[0118] Formula (1.2) is used to calculate the f(n) value of each extended node, and the node with the smallest f(n) in the extension is selected as the parent node to continue calculation and search until it reaches the vicinity of the target G. The path planning algorithm searches for the amphibious navigation map as shown in the following figure: Figure 5 shown. Figure 5The inaccessible area is set as unknown unkown, and different grid grayscales (or colors) represent different map information. The starting position S of the path planning is the end point G of the sea path, the intermediate position to be passed is the landing point L (known), and the target position is the end point position T specified on land. The optimal path landPath generated by the path planning algorithm is as follows Figure 7 As shown in FIG. , the final navigation path path of the amphibious unmanned platform is composed of the ocean path oceanPath and landPath.

[0119] In addition, the present invention also provides a path planning system for an amphibious unmanned platform, comprising:

[0120] A map acquisition module, used to acquire nautical charts, land structured maps and unstructured satellite maps;

[0121] An amphibious navigation map construction module is used to construct an amphibious navigation map based on nautical charts, land structured maps and unstructured satellite maps;

[0122] A sea path determination module, used to determine the sea path using the A* algorithm based on the amphibious navigation map;

[0123] An integrated path determination module is used to determine the integrated path of the water-land transition and the land area based on the amphibious navigation map and the sea path;

[0124] The amphibious planning path determination module is used to splice the sea path and the land path to obtain the amphibious planning path.

[0125] Among them, the amphibious navigation map construction module specifically includes:

[0126] A marine data extraction unit for determining the waterways and impassable areas on the sea in the nautical chart based on the unstructured satellite map;

[0127] A land data extraction unit, used for determining roads and land impassable areas in a land structured map based on an unstructured satellite map;

[0128] The land-water connection data extraction unit is used to input the unstructured satellite map into the land-water connection data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water connection area; the land-water connection data extraction model is obtained by training the neural network using the annotated unstructured satellite map;

[0129] The amphibious navigation map construction unit is used to construct an amphibious navigation map based on the coordinates, altitude and terrain passability of each point in the waterway, impassable areas at sea, roads and impassable areas on land, and the water-land junction area.

[0130] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0131] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A path planning method for an amphibious unmanned platform, characterized in that: include: Obtain nautical charts, land structured maps and unstructured satellite maps; constructing an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map; Determine a sea path using an A* algorithm according to the amphibious navigation map; Determine a comprehensive route for land-water transition and land area based on the amphibious navigation map and the sea route; splicing the sea path and the land path to obtain an amphibious planning path; Determining the integrated path of the land-water conversion and the land area according to the amphibious navigation map and the sea path specifically includes: According to the sea route, determining a starting point and a destination point of the integrated route; Determine the starting point of the integrated path as the parent node and add it to the integrated path node set; Determine a plurality of extension nodes of the parent node using a kinematic equation and the amphibious navigation map; Determining the actual moving cost, heuristic estimation cost, node height cost, node traversability cost and node acceleration of each of the extended nodes according to the boundary conditions; Determine a weight coefficient group according to the motion state and boundary conditions of the amphibious unmanned platform; Determining the costs of the plurality of extended nodes using a cost evaluation function according to the weight coefficient group, and the actual moving cost, heuristic estimation cost, node height cost, node traversability cost and node acceleration of each of the extended nodes; Determine whether the distance between the extended node with the minimum cost and the target point is less than the distance threshold, and obtain the judgment result; If the judgment result is no, the extended node with the minimum cost is used as the child node of the parent node, the child node and the parent node pointer of the child node are added to the integrated path node set, the child node is used as the parent node and the step of "using the kinematic equation and the amphibious navigation map to determine multiple extended nodes of the parent node" is returned; If the judgment result is yes, the target point of the comprehensive path is taken as the child node of the parent node, the child node and the parent node pointer of the child node are added to the comprehensive path node set, and the comprehensive path of the water-land conversion and the land area is determined according to the comprehensive path node set; Determining the starting point and the target point of the comprehensive path according to the sea path specifically includes: When the direction of the path planning is from sea to land, the end point of the sea path is determined as the starting point of the integrated path, and the target point of the amphibious planning path is determined as the target point of the integrated path; When the direction of the path planning is from land to sea, the starting point of the amphibious planning path is determined as the starting point of the comprehensive path, and the starting point of the sea path is determined as the target point of the amphibious planning path; The cost evaluation function is: f(n)=w1g(n)+w2h(n)+w3e(n)+w4t(n)+w5a(n); Among them, f(n) represents the cost of the nth extended node; g(n), h(n), e(n), t(n) and a(n) represent the actual moving cost, heuristic estimated cost, node height cost, node passable cost and node acceleration of the nth extended node respectively; w1, w2, w3, w4, w5 represent the actual moving cost weight coefficient, heuristic estimated cost weight coefficient, node height cost weight coefficient, node passable cost weight coefficient and node acceleration weight coefficient of the nth extended node respectively.

2. A path planning method for an amphibious unmanned platform according to claim 1, characterized in that: The step of constructing an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map specifically includes: Determining the waterway and the impassable area at sea in the nautical chart according to the unstructured satellite map; Determining roads and land inaccessible areas in the land structured map according to the unstructured satellite map; The unstructured satellite map is input into a land-water interface data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water interface area; the land-water interface data extraction model is obtained by training a neural network using the annotated unstructured satellite map; An amphibious navigation map is constructed based on the coordinates, altitude and terrain passability of waterways, impassable areas at sea, roads and impassable areas on land, and each point at the junction of land and water.

3. A path planning method for an amphibious unmanned platform according to claim 1, characterized in that: Before obtaining the nautical chart, the land structured map and the unstructured satellite map, the method further includes: Obtain unstructured satellite atlas; The coordinates, altitude and terrain passability of each point at the land-water junction area in each unstructured satellite map in the unstructured satellite map set are marked to obtain a marked unstructured satellite map set; With unstructured satellite atlas as input and labeled unstructured satellite atlas as expected output, the neural network is trained to obtain the land-water connection data extraction model.

4. A path planning system for an amphibious unmanned platform, characterized in that: include: A map acquisition module, used to acquire nautical charts, land structured maps and unstructured satellite maps; An amphibious navigation map construction module, used to construct an amphibious navigation map according to the nautical chart, the land structured map and the unstructured satellite map; A sea path determination module, used to determine a sea path using an A* algorithm according to the amphibious navigation map; The integrated path determination module is used to determine the integrated path of the land-water conversion and the land area according to the amphibious navigation map and the sea path; the integrated path determination module is also used to determine the starting point and the target point of the integrated path according to the sea path; determine the starting point of the integrated path as the parent node to be added to the integrated path node set; use the kinematic equation and the amphibious navigation map to determine multiple extended nodes of the parent node; determine the actual movement cost, heuristic estimation cost, node height cost, node passable cost and node acceleration of each of the extended nodes according to the boundary conditions; determine the weight coefficient group according to the motion state and boundary conditions of the amphibious unmanned platform; according to the weight The coefficient group, as well as the actual moving cost, heuristic estimation cost, node height cost, node passable cost and node acceleration of each of the extended nodes, use the cost evaluation function to determine the cost of multiple extended nodes; judge whether the distance between the extended node with the smallest cost and the target point is less than the distance threshold, and obtain a judgment result; if the judgment result is no, use the extended node with the smallest cost as the child node of the parent node, add the child node and the parent node pointer of the child node to the integrated path node set, use the child node as the parent node and return to the step "using the kinematic equation and the amphibious navigation map to determine multiple extended nodes of the parent node"; if the judgment result is yes, Then, the target point of the integrated path is taken as the child node of the parent node, the child node and the parent node pointer of the child node are added to the integrated path node set, and the integrated path of the land-water conversion and the land area is determined according to the integrated path node set; wherein, according to the sea path, determining the starting point and target point of the integrated path specifically includes: when the direction of path planning is from sea to land, determining the end point of the sea path as the starting point of the integrated path, and determining the target point of the amphibious planning path as the target point of the integrated path; when the direction of path planning is from land to sea, determining the starting point of the amphibious planning path as the starting point of the integrated path, and determining the starting point of the sea path as the starting point of the amphibious planning path. Target point; the cost evaluation function is: f(n)=w1g(n)+w2h(n)+w3e(n)+w4t(n)+w5a(n); wherein, f(n) represents the cost of the nth extended node; g(n), h(n), e(n), t(n) and a(n) represent the actual moving cost, heuristic estimated cost, node height cost, node passable cost and node acceleration of the nth extended node respectively; w1, w2, w3, w4 and w5 represent the actual moving cost weight coefficient, heuristic estimated cost weight coefficient, node height cost weight coefficient, node passable cost weight coefficient and node acceleration weight coefficient of the nth extended node respectively; The amphibious planning path determination module is used to splice the sea path and the land path to obtain the amphibious planning path.

5. A path planning system for an amphibious unmanned platform according to claim 4, characterized in that: The amphibious navigation map construction module specifically includes: A marine data extraction unit, used to determine the waterway and the impassable area at sea in the nautical chart according to the unstructured satellite map; A land data extraction unit, used to determine the roads and land inaccessible areas in the land structured map according to the unstructured satellite map; A land-water connection data extraction unit is used to input the unstructured satellite map into a land-water connection data extraction model to obtain the coordinates, altitude and terrain passability of each point in the land-water connection area; the land-water connection data extraction model is obtained by training a neural network using the annotated unstructured satellite map; The amphibious navigation map construction unit is used to construct an amphibious navigation map based on the coordinates, altitude and terrain passability of each point in the waterway, impassable areas at sea, roads and impassable areas on land, and the water-land junction area.

Images