A path planning method based on GeoSOT-3D grid modeling and improved A-star algorithm fusion

Abstract

The application discloses a path planning method based on GeoSOT-3D grid modeling and improved A-star algorithm fusion, and relates to the fields of computer data processing and flight control technology.The method comprises the following steps: three-dimensional environment modeling based on a GeoSOT 3D grid to obtain a three-dimensional environment model; improving an A-star algorithm to obtain an improved A-star algorithm; iterative search in the constructed three-dimensional environment model based on the improved A-star algorithm to find an optimal discrete path point set from a starting point to a target point; and interpolation smoothing processing of the optimal discrete path point set by using a Catmull-Rom spline curve to generate a continuous smooth path meeting the dynamics constraints of a spacecraft.The method realizes efficient, safe and smooth path planning of the spacecraft by fusing three-dimensional grid modeling, optimization of a neighborhood search strategy, a safe obstacle avoidance mechanism and path smoothing optimization.

Description

Technical Field

[0001] This invention relates to the fields of computer data processing and flight control technology, and in particular to a path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A-satellite algorithm. Background Technology

[0002] As an important component of strategic emerging economies and new productive forces, the economy has become a significant growth engine for optimizing and upgrading industrial structure, while path planning is a core element in ensuring flight safety and improving the efficiency of airspace resource utilization.

[0003] While the conventional A* algorithm is the mainstream among existing path planning algorithms, it suffers from significant bottlenecks in 3D environment applications: First, the 26-neighborhood search leads to a surge in expanded nodes, resulting in high computational overhead and insufficient real-time performance; second, it fails to consider aircraft size and safety distance, posing potential risks to path safety; third, the planned path is a discrete polyline, which does not match the aircraft's dynamic constraints; and fourth, traditional grid modeling struggles to achieve high-precision, multi-scale global environmental representation, limiting its adaptability. Furthermore, the complexity of flight environments (including terrain, buildings, towers, and other obstacles) places higher demands on the accuracy of environmental modeling and the robustness of the path planning algorithm.

[0004] Therefore, proposing a path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm to solve the difficulties of the existing technology is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A-Star algorithm. By integrating three-dimensional mesh modeling, optimized neighborhood search strategy, safe obstacle avoidance mechanism and path smoothing optimization technology, it realizes efficient, safe and smooth path planning for aircraft, which is suitable for flight path calculation of aircraft such as UAVs in economic scenarios.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A path planning method based on the fusion of GeoSOT-3D mesh modeling and an improved A* algorithm includes: S1. Model the three-dimensional environment based on GeoSOT3D mesh to obtain the three-dimensional environment model; S2. Improve the A* algorithm to obtain the improved A* algorithm; S3. Based on the improved A* algorithm, an iterative search is performed in the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point; S4. Use Catmull-Rom spline curves to interpolate and smooth the optimal discrete path point set to generate a continuous and smooth path that satisfies the aircraft dynamics constraints.

[0007] Optionally, in the above method, in S1, 3D environment modeling is performed based on GeoSOT3D mesh, specifically as follows: Using the CGCS2000 national geodetic coordinate system, the target airspace is recursively divided in three directions: longitude, latitude, and geodetic height, to construct a LOD model with a three-dimensional tile pyramid structure with multiple levels. Each grid cell in the LOD model is uniquely encoded, and the walkable state parameter of all grid cells is initialized to True. Based on the interactive operation of selecting the obstacle area and its height, the passability status parameter of the selected grid cell is updated to False, thus completing the integrated definition of obstacle spatial information and passability attributes and constructing a three-dimensional environment model.

[0008] Optionally, the above method involves uniquely encoding each grid cell in the LOD model. The grid encoding consists of ellipsoidal quaternary variable-length encoding and height-domain binary variable-length encoding, and includes the spatial location and hierarchical information of the grid.

[0009] Optionally, in S2, the above method can be improved by modifying the A* algorithm, specifically as follows: S201. Define the cost function: , in, Let n be the total cost of node n. The actual cost from the starting point to node n. This represents the estimated cost from node n to the target node. Euclidean distance is used for calculation The formula is: , For the current node, For the target node; S202. Based on the horizontal angle α and elevation angle β of the projection between the current node and the target node, the search range in the three-dimensional space is reduced from 26 neighborhoods to no more than 7 neighborhoods, and the node interval to be expanded is dynamically determined. S203. Detect whether there are obstacles in the six direct neighboring nodes of the current node in the positive and negative directions of the X-axis, Y-axis, and Z-axis. If there is an obstacle in a certain direction, then according to the preset obstacle avoidance rule table, exclude the child nodes located on the diagonal of the obstacle direction from the node interval to be expanded, and obtain the final node set to be expanded.

[0010] Optionally, in S3 of the above method, based on the improved A* algorithm, an iterative search is performed in the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point, specifically: Based on the cost function from step S2 and the final set of nodes to be expanded, an iterative search is performed in the constructed 3D environment model until the total cost from the starting point to the target point is found. The smallest set of discrete path points is the optimal set of discrete path points.

[0011] Optionally, in S4 of the above method, Catmull-Rom spline curves are used to interpolate and smooth the optimal discrete path point set to generate a continuous, smooth path that satisfies the aircraft dynamics constraints. Specifically: The optimal discrete path point set is optimized by interpolation using CatmullRom spline curves. Based on the CatmullRomSpline class of the Cesium engine, 200 sampling points are generated to form a continuous and smooth curve trajectory.

[0012] As can be seen from the above technical solution, compared with the prior art, the present invention provides a path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm, which has the following beneficial effects: The present invention, through a 7-neighborhood expansion search strategy, compared with the conventional A* algorithm, achieves the following advantages: The algorithm's 26-neighborhood search, in GeoSOT-3D mesh environments of levels 19, 20, and 21, reduces the average number of expanded search nodes by over 83% and the average processing time by 11% to 23%, solving the problems of high computational overhead and poor real-time performance caused by redundant node expansion in 3D environments. A safe obstacle avoidance mechanism is introduced, considering the actual size of the aircraft and the safe obstacle avoidance distance, avoiding paths that are too close to obstacles and eliminating the risk of flight collisions. Through Catmull-Rom spline curve optimization, discrete polygonal lines are transformed into continuous smooth trajectories, matching the dynamic constraints of the actual aircraft motion and improving flight stability. Based on the GeoSOT-3D mesh, global multi-scale, high-precision 3D environment modeling is achieved, supporting the construction of mesh scenes at different levels, solving the problem that traditional raster modeling is difficult to adapt to wide-area scenes. Secondary development is supported, and it can be connected to digital bases (such as Air Engine). The platform's airspace grid engine and airspace computing engine are adapted to meet the needs of multiple scenarios such as flight monitoring and drone operation. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0014] Figure 1 A flowchart illustrating a path planning method based on the fusion of GeoSOT-3D mesh modeling and an improved AlphaGo algorithm, provided by this invention; Figure 2 The flowchart of the path planning method based on GeoSOT-3D mesh modeling and improved AlphaGo algorithm provided by this invention describes the execution process of finding the optimal discrete path point set in a three-dimensional environment model through iterative search. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Reference Figure 1 As shown, this invention discloses a path planning method based on the fusion of GeoSOT-3D mesh modeling and an improved A* algorithm, comprising: S1. Model the three-dimensional environment based on GeoSOT3D mesh to obtain the three-dimensional environment model; S2. Improve the A* algorithm to obtain the improved A* algorithm; S3. Based on the improved A* algorithm, an iterative search is performed in the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point; S4. Use Catmull-Rom spline curves to interpolate and smooth the optimal discrete path point set to generate a continuous and smooth path that satisfies the aircraft dynamics constraints.

[0017] Furthermore, in S1, 3D environment modeling is performed based on GeoSOT3D meshes, specifically as follows: Using the CGCS2000 national geodetic coordinate system, the global space is recursively divided into three dimensions: longitude (-180°~180°), latitude (-90°~90°), and geodetic height. The division depth is 32 levels, covering an area from about 6302km underground to about 528680km above the ground, forming a multi-detailed LOD model with a three-dimensional tile pyramid structure. Each grid cell in the LOD model is uniquely encoded, and the walkable state parameter of all grid cells is initialized to True. Based on the interactive operation of selecting the obstacle area and its height, the passability status parameter of the selected grid cell is updated to False, thus completing the integrated definition of obstacle spatial information and passability attributes and constructing a three-dimensional environment model.

[0018] Furthermore, before path planning, the "accessibility status" parameter of all grid cells is initialized to True by default, indicating that there are no obstacles. Users can select a specific area on the 3D map and set its height range through the interactive interface. The system will automatically calculate and lock all grid cells covered by that area, update their "accessibility status" parameter to False, and highlight them in red on the interface. In this way, the spatial information and access attributes of obstacles such as buildings, towers, and no-fly zones are fused and defined, providing a precise three-dimensional environmental carrier for subsequent path search. It supports the rapid construction of three-dimensional map grids at any location in the world for specified areas and specified levels (such as level 19, level 20, and level 21) to adapt to the accuracy requirements of different flight scenarios. Furthermore, each grid cell in the LOD model is uniquely encoded. This encoding consists of an ellipsoidal quaternary one-dimensional variable-length encoding (maximum 32 bits) and a height-domain binary one-dimensional variable-length encoding (maximum 132 bits). The encoding itself contains the spatial location and hierarchical information of the grid.

[0019] Furthermore, in S2, the A* algorithm is improved as follows: S201. Define the cost function: , in, Let n be the total cost of node n. The actual cost from the starting point to node n. This represents the estimated cost from node n to the target node. Euclidean distance is used for calculation The formula is: , For the current node, For the target node; S202. Based on the horizontal angle α and elevation angle β of the projection between the current node and the target node, the search range in the three-dimensional space is reduced from 26 neighborhoods to no more than 7 neighborhoods, reducing the redundancy of the expanded nodes, reducing the computational overhead, and dynamically determining the interval of the node to be expanded. Traditional 3D A* algorithm typically considers 26 neighboring nodes of the current node (up, down, left, right, front, back, and various diagonal directions), leading to the expansion of a large number of invalid nodes. This invention reduces the search range by introducing directional constraints. Specifically, it calculates the directional relationship between the current node and the target node. In this invention, the horizontal angle α and the elevation angle β of the projections of the current node and the target node are used to quantize this direction. Before expanding the child nodes of the current node, the horizontal angle α and the elevation angle β of the projections of the current node and the target node are first calculated. Based on the values ​​of α and β, the expansion interval is determined by looking up a table. Table 1 shows the corresponding set of expanded nodes under different α and β intervals. Table 1. Sets of extended nodes corresponding to different α and β intervals.

[0020] In this way, the original 26-neighborhood search is dynamically reduced to a maximum of 7 neighborhoods, and the average number of expanded nodes can be reduced by about 83%, which significantly improves search efficiency. S203. Detect whether there are obstacles in the six direct neighboring nodes of the current node in the positive and negative X-axis, positive and negative Y-axis, and positive and negative Z-axis directions. If there is an obstacle in a certain direction, then according to the preset obstacle avoidance rule table, exclude the child nodes located on the diagonal of the obstacle direction from the node interval to be expanded, and obtain the final set of nodes to be expanded. The specific obstacle avoidance rule table is shown in Table 2. Table 2 Obstacle Avoidance Rules

[0021] Furthermore, in S3, based on the improved A* algorithm, an iterative search is performed within the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point, specifically: Based on the cost function from step S2 and the final set of nodes to be expanded, an iterative search is performed in the constructed 3D environment model until the total cost from the starting point to the target point is found. The smallest set of discrete path points is the optimal set of discrete path points. Furthermore, refer to Figure 2 As shown, an iterative search is performed within the constructed 3D environment model, specifically as follows: S301. Create an Open list (to store nodes to be processed) and a Closed list (to store nodes that have been processed), and add the starting node to the Open list; S302. Check if the Open list is empty. If it is empty, path planning fails and an empty path is returned. S303. Retrieve the total cost from the Open list. f(n) The smallest node is selected as the current processing node and moved to the Closed list; S304. Determine if the current node is the target node; if yes, backtrack the parent node pointer, organize the preliminary discrete path point set, and proceed to step S4; if no, proceed to step S305. S305. Calculate the horizontal angle α and elevation angle β between the current node and the target node, and determine the extended node interval based on α and β; S306. Detect whether there are obstacles in the six directions of the current node: up, down, left, right, front, and back. Based on the detection results and the preset obstacle avoidance rules, exclude nodes that cannot be expanded due to safety factors from the candidate set. S307. For each remaining child node to be expanded, determine whether it is already in the Open list: if it already exists, check and update its minimum cost value. g(n) and its parent node; if it does not exist, then calculate its parent node. f(n) , g(n) , h(n) The value is then added to the Open list; S308. Repeat steps S302 to S307 until the target node is found or the Open list is empty.

[0022] Furthermore, in S4, Catmull-Rom spline curves are used to interpolate and smooth the optimal discrete path point set, generating a continuous and smooth path that satisfies the aircraft dynamics constraints. Specifically: The path obtained through the above steps is a broken line formed by connecting the center points of a discrete grid. This path has "hard angles" and does not meet the dynamic constraints of the aircraft. To address this issue, this invention employs Catmull-Rom spline curves for interpolation optimization of discrete path points. Based on the CatmullRomSpline class provided by the Cesium engine, the key discrete points output by the improved A* algorithm are used as control points to generate 200 uniformly distributed sampling points. These sampling points are then connected to form a smooth curved trajectory. This curved trajectory strictly passes through all original path points and has continuous tangents at each point, ensuring the smoothness of the path and effectively reducing aircraft vibration during flight, thus improving flight quality.

[0023] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0024] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A path planning method based on the fusion of GeoSOT-3D mesh modeling and an improved AlphaGo algorithm, characterized in that, include: S1. Model the three-dimensional environment based on GeoSOT3D mesh to obtain the three-dimensional environment model; S2. Improve the A* algorithm to obtain the improved A* algorithm; S3. Based on the improved A* algorithm, an iterative search is performed in the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point; S4. Use Catmull-Rom spline curves to interpolate and smooth the optimal discrete path point set to generate a continuous and smooth path that satisfies the aircraft dynamics constraints.

2. The path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm as described in claim 1, characterized in that, In S1, 3D environment modeling is performed based on GeoSOT3D meshes, specifically as follows: Using the CGCS2000 national geodetic coordinate system, the target airspace is recursively divided in three directions: longitude, latitude, and geodetic height, to construct a LOD model with a three-dimensional tile pyramid structure with multiple levels. Each grid cell in the LOD model is uniquely encoded, and the walkable state parameter of all grid cells is initialized to True. Based on the interactive operation of selecting the obstacle area and its height, the passability status parameter of the selected grid cell is updated to False, thus completing the integrated definition of obstacle spatial information and passability attributes and constructing a three-dimensional environment model.

3. The path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm as described in claim 2, characterized in that, Each grid cell in the LOD model is uniquely encoded. The grid encoding consists of ellipsoidal quaternary variable-length encoding and height-domain binary variable-length encoding, which includes the spatial location and hierarchical information of the grid.

4. The path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm as described in claim 3, characterized in that, In S2, the A* algorithm is improved as follows: S201. Define the cost function: ， in, Let n be the total cost of node n. The actual cost from the starting point to node n. This represents the estimated cost from node n to the target node. Euclidean distance is used for calculation The formula is: ， For the current node, For the target node; S202. Based on the horizontal angle α and elevation angle β of the projection between the current node and the target node, the search range in the three-dimensional space is reduced from 26 neighborhoods to no more than 7 neighborhoods, and the node interval to be expanded is dynamically determined. S203. Detect whether there are obstacles in the six direct neighboring nodes of the current node in the positive and negative directions of the X-axis, Y-axis, and Z-axis. If there is an obstacle in a certain direction, then according to the preset obstacle avoidance rule table, exclude the child nodes located on the diagonal of the obstacle direction from the node interval to be expanded, and obtain the final node set to be expanded.

5. The path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm as described in claim 4, characterized in that, In S3, based on the improved A* algorithm, an iterative search is performed in the constructed 3D environment model to find the optimal set of discrete path points from the starting point to the target point, specifically: Based on the cost function from step S2 and the final set of nodes to be expanded, an iterative search is performed in the constructed 3D environment model until the total cost from the starting point to the target point is found. The smallest set of discrete path points is the optimal set of discrete path points.

6. The path planning method based on the fusion of GeoSOT-3D mesh modeling and the improved A* algorithm as described in claim 5, characterized in that, In S4, Catmull-Rom spline curves are used to interpolate and smooth the optimal discrete path point set, generating a continuous and smooth path that satisfies the aircraft dynamics constraints. Specifically: The optimal discrete path point set is optimized by interpolation using CatmullRom spline curves. Based on the CatmullRomSpline class of the Cesium engine, 200 sampling points are generated to form a continuous and smooth curve trajectory.

Images