A Mountain Path Planning Method Based on an Improved A* Algorithm

Abstract

This invention discloses a mountain path planning method based on an improved A* algorithm, comprising: acquiring lidar or image data of the target area using a UAV and generating point cloud data to construct a digital elevation model containing elevation and slope information; subsequently, performing coordinate unification, noise reduction, and downsampling processing on the point cloud data, and combining the start and end points for region clipping and rasterization to construct a local raster map; employing the improved A* algorithm during path search, introducing a slope correction mechanism and an equivalent distance cost model, comprehensively incorporating distance, slope, and elevation factors into the path cost function, and dynamically adjusting the path cost based on the slope influence coefficient, while simultaneously constraining high-slope and impassable areas. This method can effectively reduce the search space and computational complexity, improve the efficiency and accuracy of path planning in complex mountainous environments, and is applicable to scenarios such as power transmission line planning and mountain engineering path design.

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, specifically relating to a mountain path planning method based on an improved A* algorithm. Background Technology

[0002] In real-world applications, the impact of terrain complexity on path planning cannot be ignored. Especially in complex mountainous or urban environments, terrain undulations, obstacle distribution, and environmental factors (such as vegetation cover and geological conditions) all significantly influence path selection. Traditional two-dimensional path planning algorithms, such as Dijkstra's algorithm and A* algorithm, have been widely used in planar environments. However, in three-dimensional complex terrain scenarios, their path planning results often fail to meet practical engineering needs because they cannot fully represent elevation changes and terrain constraints. Therefore, adapting to three-dimensional complex terrain conditions and improving the applicability of path planning remains a crucial challenge in the field.

[0003] As power infrastructure construction continues to expand into mountainous areas, transmission line construction projects face numerous challenges. The complex terrain of mountainous areas makes traditional route planning methods, which rely on manual on-site surveys, inefficient and difficult to obtain comprehensive continuous terrain information. This can easily lead to the omission of unfavorable factors such as steep slopes, complex valleys, and areas with geological risks, thereby affecting the reliability of route planning and the safety of the project.

[0004] Unmanned aerial vehicles (UAVs) are widely used in topographic mapping due to their flexibility, maneuverability, and efficient data acquisition capabilities. However, in actual data acquisition, point cloud or image data obtained by UAVs or remote sensing methods often suffer from high data redundancy, significant noise interference, and inconsistent data structures. Without effective preprocessing, these issues will affect the accuracy and computational efficiency of subsequent terrain modeling and path analysis.

[0005] In the prior art, for example, CN117571012A discloses a global path planning method for unmanned vehicles in off-road environments. This method constructs a digital elevation model through satellite elevation data, obtains ground type information by combining remote sensing images, and introduces factors such as slope, surface undulation degree and ground type to construct a raster map, and performs path search based on the map.

[0006] However, the above-mentioned existing technologies still have the following shortcomings: the method mainly relies on satellite elevation data and remote sensing images, which are limited in terms of data acquisition accuracy and spatial resolution, making it difficult to adapt to the needs of local high-precision terrain modeling, especially in complex mountainous areas or areas with significant micro-topographic changes, where its terrain representation ability is insufficient. Furthermore, in complex terrain scenarios, this method has limited ability to handle data redundancy and control computational complexity after multi-source data fusion, which may lead to a decrease in computational efficiency in large-scale raster environments.

[0007] Therefore, existing technologies still have certain limitations in terms of high-precision terrain representation, comprehensive modeling of multi-factor path costs, and path search efficiency in complex 3D environments. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a mountain path planning method based on an improved A* algorithm.

[0009] The objective of this invention can be achieved through the following technical solutions: This invention provides a mountain path planning method based on an improved A* algorithm, comprising the following steps: Acquire drone data of the target area and generate point cloud data. Construct a digital elevation model based on the point cloud data to obtain a mountain model containing elevation and slope information. The point cloud data is preprocessed, and the mountain model is cropped according to the starting and ending positions to obtain the target area data. The target area data is then divided into grids to construct a grid map. Based on the grid map, an improved A* algorithm is used for path search. During the path search process, a path cost is constructed, which includes path distance, slope, and elevation factors. During the path search process, the path cost is corrected based on the slope, and areas exceeding preset conditions are restricted to obtain the path planning result from the starting point to the ending point.

[0010] Furthermore, the process of acquiring UAV data of the target area and generating point cloud data, and constructing a digital elevation model based on the point cloud data to obtain a mountain model containing elevation and slope information, specifically includes: The target area is scanned using lidar sensors or image acquisition equipment carried by the drone to obtain raw remote sensing data. The original remote sensing data is subjected to coordinate registration and spatial alignment to unify the coordinate system and obtain initial point cloud data; The initial point cloud data is filtered to remove noise points and outliers, and ground point data is extracted using a classification algorithm. A digital elevation model is constructed based on the ground point data, and terrain slope information is calculated by the elevation difference between adjacent points, wherein the slope at any location is determined based on the rate of change of elevation between that location and its neighboring points. The elevation and slope information in the digital elevation model are fused to generate a mountain model for route planning.

[0011] Furthermore, the slope is calculated using the following formula: in, This is the tangent of the slope; This indicates the location in the digital elevation model. The elevation value at that location, This indicates the spatial resolution between adjacent grid cells.

[0012] Furthermore, the preprocessing of the point cloud data and the cropping of the mountain model based on the starting and ending positions to obtain the target area data specifically include: The point cloud data is subjected to coordinate offset correction to unify the point cloud data to a preset reference coordinate system, and the point cloud coordinates are discretized to map continuous coordinates to discrete grid coordinates. The point cloud data is downsampled by dividing the space using a voxel grid filtering method and selecting representative points within each voxel unit to reduce the amount of point cloud data. The side length of the voxel unit is set to a preset resolution parameter. Based on the input start and end positions, a rectangular region encompassing the start and end positions is determined in the mountain model. The boundary range of the rectangular region satisfies: in, and These represent the planar coordinates of the starting and ending positions, respectively. and This is the preset extended range parameter; The point cloud data is cropped based on the rectangular region to obtain the target region data, while retaining the elevation and slope information within the target region.

[0013] Furthermore, the improved A* algorithm includes the following steps: Define a start grid and an end grid in the grid map, create an open list and a closed list, and add the start grid to the open list; Select the grid with the minimum evaluation function value from the open list as the current grid and add the current grid to the closed list. Then, traverse the adjacent grids of the current grid. The adjacent grids of the current grid include the grids that are adjacent to the current grid in the row and column coordinate directions, and the grids that are adjacent to the current grid in the diagonal direction. The accessibility of the adjacent grid cells is determined. If the adjacent grid cell is not visited and the access conditions are met, the adjacent grid cell is added to the open list, and its evaluation function value and parent node information are updated. Repeat the adjacent grid expansion process until the endpoint grid is added to the closed list to obtain the optimal path.

[0014] Furthermore, the evaluation function is defined by the following formula: in, Indicates the current grid The evaluation function value; This indicates the distance from the starting grid cell to the current grid cell. The cumulative path cost, Indicates starting from the current grid The estimated cost to the endpoint grid.

[0015] Furthermore, the estimated cost The formula is: in, and These represent the current grid cells. Column coordinates and row coordinates, and These represent the column and row coordinates of the endpoint grid, respectively. Indicates the current grid Elevation value, Indicates the elevation value of the endpoint grid. Indicates the spatial resolution of the raster.

[0016] Furthermore, the cumulative path cost The formula is: in, Indicates the first in the path Spatial distance between adjacent grid cells Indicates the first The tangent of the slope corresponding to the path segment. This indicates the vehicle's speed on a flat road, specifically the speed when the gradient is 0. This represents the slope influence coefficient.

[0017] Furthermore, the spatial distance between adjacent grid cells It is obtained through three-dimensional Euclidean distance calculation, and the calculation formula is as follows: in, , and These represent the first and second parts of the path, respectively. The column coordinates, row coordinates, and elevation values ​​of each grid cell.

[0018] Furthermore, the slope influence coefficient The determination is based on vehicle characteristics, road conditions, and gradient range, specifically including: The slope influence coefficient based on vehicle characteristics The basic settings include vehicle characteristics such as vehicle mass, power output capability, and drive mode. A higher vehicle mass or lower power output capability corresponds to a higher gradient influence coefficient. The larger the value; The influence coefficient of slope based on road conditions The correction is made, and the road surface conditions include the road surface friction coefficient and the road surface smoothness. When the road surface friction coefficient is low or the road surface smoothness is poor, the slope influence coefficient is increased. The possible values ​​of ; The slope influence coefficient based on the slope range Adaptive adjustments are made; when the slope is within a relatively small, pre-set range, a lower slope influence coefficient is adopted. When the slope is within a second preset large range, the slope influence coefficient is increased. .

[0019] Compared with the prior art, the present invention has the following advantages: (1) In existing technologies, when relying on satellite elevation data and remote sensing images for terrain modeling, the limited resolution of the data sources makes it difficult to accurately reflect the micro-topographical changes in complex mountainous environments, resulting in insufficient expression of detailed information such as local slopes and valleys, thus affecting the accuracy of path planning. This invention achieves high-precision three-dimensional terrain reconstruction of the target area by using UAVs to collect point cloud data and constructing a digital elevation model, and combines slope information for terrain expression, thereby more accurately reflecting the spatial undulation characteristics of complex mountainous areas and improving the accuracy of terrain modeling and the reliability of path planning.

[0020] (2) Existing technologies, when constructing path cost models, typically only consider single or simple superimposed factors such as slope, surface undulation, or ground type, lacking a comprehensive characterization of multiple terrain influencing factors, resulting in a discrepancy between path cost and actual travel cost. This invention constructs a path cost model by introducing multiple factors such as path distance, slope, and elevation on a grid map basis, and conducts a comprehensive evaluation during the path search process, thereby achieving a more comprehensive expression of travel costs in complex terrain and improving the rationality and engineering applicability of path selection.

[0021] (3) In the path evaluation process of complex three-dimensional terrain environments, the existing technology does not adequately consider the dynamic impact of elevation changes, and the path search process is prone to the problem of large deviations between the local optimal path and the actual optimal path. This invention introduces a cost update mechanism based on slope correction into the A* algorithm and dynamically adjusts the path cost in combination with elevation information, so that the path search process can more fully reflect the impact of terrain undulations on the travel cost, thereby improving the global optimality and stability of the path planning results.

[0022] (4) In the process of large-scale terrain raster modeling and path search, the existing technology usually performs calculations directly based on global data, resulting in a large search space and a lot of redundant calculations, which affects the computational efficiency. The present invention preprocesses and prunes the point cloud data, retaining only the key areas containing the start and end points, and performs raster division and path search within these areas, thereby effectively reducing the search space, reducing computational complexity, and improving the computational efficiency and real-time performance of path planning.

[0023] (5) Existing technologies lack effective constraint mechanisms for different terrain risk areas during the route planning process, which can easily lead to routes traversing steep slopes or complex terrain areas, reducing route safety. This invention introduces a slope-based cost correction and inaccessible area restriction mechanism during the route search process to penalize or shield high-slope areas, thereby effectively avoiding dangerous areas and improving route safety and feasibility.

[0024] (6) Existing technologies typically use fixed or empirical values ​​for parameter settings, which are difficult to adapt to different vehicle types and terrain conditions. This invention introduces a slope influence coefficient and adaptively adjusts it in combination with vehicle characteristics, road conditions, and slope range, enabling the path cost model to dynamically change according to the actual application scenario, thereby improving the model's adaptability and engineering versatility. Attached Figure Description

[0025] Figure 1 This is a flowchart of the mountain path planning method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the point cloud data preprocessing interface and its results according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the path planning interface and its three-dimensional terrain map results according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the path planning coordinate output results according to an embodiment of the present invention. Detailed Implementation

[0026] 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, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] Example 1: This embodiment provides a mountain path planning method based on an improved A* algorithm, such as... Figure 1 As shown, it includes the following steps: Step S1: Acquire UAV data of the target area and generate point cloud data. Based on the point cloud data, construct a digital elevation model to obtain a mountain model containing elevation and slope information. In one specific embodiment, UAV data of the target area is acquired and point cloud data is generated. A digital elevation model is then constructed based on the point cloud data to form a mountain model containing elevation and slope information. The core objective of this process is to replace traditional two-dimensional or low-resolution elevation data with high-precision three-dimensional terrain reconstruction, thereby improving the terrain representation capability of complex mountain environments and the quality of basic data for path planning.

[0028] During the data acquisition phase, the target area is scanned using lidar sensors or image acquisition equipment mounted on the UAV to obtain raw remote sensing data. LiDAR or image data is used because it has high spatial resolution and surface penetration capabilities, effectively acquiring true surface information under complex vegetation cover or undulating terrain conditions, thereby reducing the problem of insufficient accuracy of traditional satellite data in local areas.

[0029] The acquired raw remote sensing data undergoes coordinate registration and spatial alignment to unify the spatial reference system of different data sources, forming initial point cloud data. This step aims to eliminate spatial deviations between multi-source data, ensuring spatial consistency in subsequent terrain modeling and avoiding terrain misalignment or modeling errors caused by inconsistent coordinates.

[0030] Furthermore, the initial point cloud data is filtered to remove noise and outliers, and ground point data is extracted using a classification algorithm. This step is designed based on the characteristic that actual point cloud data typically contains vegetation, buildings, and noise interference points. By cleaning the data, the accuracy of surface modeling is improved, enabling the subsequent digital elevation model to accurately reflect the terrain undulations.

[0031] After obtaining ground point data, a digital elevation model is constructed based on this data, and terrain slope information is calculated using the elevation difference between adjacent points. The slope at any location is determined by the rate of change of elevation between that location and its neighboring points. Essentially, this approach uses local spatial gradients to characterize the trend of terrain change, thereby providing terrain constraints for path cost assessment.

[0032] In one embodiment, the slope is calculated using the following expression: in, This is the tangent of the slope; This indicates the location in the digital elevation model. The elevation value at that location, This indicates the spatial resolution between adjacent grid cells.

[0033] Furthermore, the elevation and slope information from the digital elevation model are fused to generate a mountain model for route planning. The significance of this fusion process lies in unifying the expression of absolute height information and terrain change rate information, enabling route planning to consider not only spatial location differences but also the difficulty of terrain accessibility. This provides more complete terrain-based data support for subsequent cost function calculations based on the A* algorithm.

[0034] Through the above processing flow, the transformation from raw UAV data to structured mountain models can be realized, expanding the terrain representation from a single elevation description to a joint elevation and slope description, thereby significantly improving the accuracy and stability of the basic data for path planning in complex mountainous environments, and providing more reasonable terrain constraints for subsequent path cost assessment.

[0035] Step S2: Preprocess the point cloud data and crop the mountain model according to the starting and ending positions to obtain the target area data. Divide the target area data into grids and construct a grid map. In one specific embodiment, the point cloud data is preprocessed, and the mountain model is cropped based on the starting and ending points to obtain target area data. Then, a raster map is constructed based on this target area data. The purpose of this step is to transform global mountain data into a local computational region through data dimensionality reduction and spatial constraints, thereby reducing the computational complexity of subsequent path search and improving path planning efficiency.

[0036] In the data preprocessing stage, coordinate offset correction is performed on the point cloud data acquired based on the mountain model to unify the point cloud data to a preset reference coordinate system, thereby eliminating spatial deviation problems caused by different data sources or acquisition errors. Based on the unified coordinate system, the point cloud coordinates are discretized, mapping continuous spatial coordinates to discrete grid coordinates for subsequent rasterization processing and A* algorithm search. This processing method can transform irregular spatial data into a structured data form, improving the algorithm's processability.

[0037] Furthermore, the point cloud data is downsampled, and a voxel grid filtering method is used to divide the spatial region. A representative point is selected within each voxel cell as the spatial representation of that cell, thereby compressing and simplifying the point cloud data. The side length of the voxel cell is set to a preset resolution parameter, which controls the balance between data accuracy and computational efficiency. Smaller voxel sizes result in higher accuracy but higher computational cost, while larger voxel sizes result in higher computational efficiency but lower detail representation. This design aims to reduce data redundancy while ensuring terrain representation accuracy and improving the efficiency of subsequent path planning.

[0038] During the region trimming stage, a rectangular region encompassing the start and end points is determined within the mountain model based on the input start and end points. This limits the spatial scope of the path search, thus avoiding the computational burden of a global search. The boundary of the rectangular region is set as follows: in, and These represent the planar coordinates of the starting and ending positions, respectively. and This is the preset extended range parameter; After defining the rectangular region, the point cloud data is cropped based on this region, retaining only the data points within the region, and simultaneously preserving the corresponding elevation and slope information. This processing method can significantly reduce interference from irrelevant data while ensuring the integrity of the path search, allowing subsequent raster map construction to focus more on the effective calculation area.

[0039] In the rasterization stage, the cropped target area data is mapped onto a regular raster space to construct a raster map. The raster scale is used to uniformly express spatial relationships, ensuring that each raster cell contains both spatial location and terrain information. Through rasterization, the continuous terrain space is transformed into a discrete search space, providing a standardized input structure for subsequent A* path search.

[0040] Through the above processing, the transformation from global point cloud data to local raster search space is realized, which effectively reduces the data scale and search complexity while ensuring path reachability, thereby improving the computational efficiency and stability of the overall path planning.

[0041] Step S3: Based on the grid map, the improved A* algorithm is used for path search. During the path search process, the path cost is constructed, which includes path distance, slope and elevation factors. In one specific implementation, a modified A* algorithm is used for path search based on a raster map, and path cost is constructed during the path search process to achieve a comprehensive evaluation of the merits of paths in complex mountainous environments. The path cost includes at least path distance, slope, and elevation factors, so that the path selection result can simultaneously reflect spatial distance cost and terrain accessibility.

[0042] In this path search process, a search space is first constructed based on the target area raster map obtained in step S2, and an improved A* algorithm is introduced into this search space for path optimization. The A* algorithm improves path search efficiency by combining actual path costs with heuristically estimated costs, reducing the expansion of invalid nodes while ensuring search effectiveness. In mountain path planning, this algorithm enables the search for the globally optimal path from the starting point to the destination under complex terrain constraints.

[0043] In the path cost modeling process, distance, slope, and elevation factors are used together as the basis for path evaluation. Distance reflects spatial movement costs, slope reflects the impact of terrain inclination on accessibility, and elevation reflects the constraints of terrain undulation on path selection. By involving multiple factors in cost calculation, the path evaluation results more closely reflect actual mountainous terrain conditions, thus avoiding low-feasibility paths based solely on distance.

[0044] In constructing the heuristic function, a multi-factor weighting approach is introduced, incorporating distance, slope, elevation, and environmental impact factors into the heuristic estimation model. Corresponding weight coefficients are set based on the degree of influence of different factors in route planning, thereby achieving an adaptive expression of route costs under different terrain conditions. The reason for introducing the weighting mechanism is that the degree of influence of various factors on travel costs varies in different mountainous environments. For example, slope has a more significant impact in steep areas, while distance dominates in flat areas. Adjusting the weights can improve the model's applicability and robustness.

[0045] For handling special terrain features, steep slopes and cliffs in mountainous areas are modeled and processed separately. When the slope corresponding to a grid exceeds a preset threshold, the path cost of that grid is increased, thereby reducing the likelihood of a path traversing that area. When a grid's corresponding area is determined to be impassable terrain, it is directly excluded from the expandable nodes during the path search process. The purpose of this approach is to enhance the safety constraints in the path planning process, prevent paths from traversing high-risk areas, and improve path feasibility.

[0046] In the process of combining heuristic functions with the A* search framework, node expansion is based on an open and closed list mechanism. By continuously selecting the grid with the smallest evaluation function value as the current expansion node, and traversing and updating its neighboring grids, the target grid is gradually approached. The selection of adjacent grids includes horizontal, vertical, and diagonal adjacency relationships, and combines them with elevation information to form a three-dimensional neighborhood relationship, enabling the path search to adapt to the spatial continuity of real terrain.

[0047] The evaluation function takes the following form: in, Indicates the current grid The evaluation function value; Indicates the distance from the starting grid cell to the current grid cell. The cumulative path cost, Indicates starting from the current grid The estimated cost to reach the endpoint grid.

[0048] Estimated cost The formula is: in, and These represent the current grid cells. Column coordinates and row coordinates, and These represent the column and row coordinates of the endpoint grid, respectively. Indicates the current grid Elevation value, Indicates the elevation value of the endpoint grid. This represents the spatial resolution of the raster. It can simultaneously consider horizontal distance and vertical height difference, thus more realistically reflecting the actual geometric distance in three-dimensional space, improving the rationality of heuristic estimation and the efficiency of path search.

[0049] Step S4: During the path search process, the path cost is adjusted according to the slope, and areas exceeding the preset conditions are restricted to obtain the path planning result from the starting point to the ending point. In one embodiment, existing route planning methods typically use geometric distance or Euclidean distance directly to evaluate the merits of a route during the cost estimation process. However, this approach fails to fully consider the impact of slope on actual travel speed in mountainous environments, resulting in a discrepancy between the route evaluation results and the actual travel costs in areas with steep slopes. This can lead to problems such as low travel efficiency or inaccessibility of the planned route in practical applications.

[0050] To address the aforementioned issues, this implementation method introduces a slope impact factor into the path cost assessment and modifies the traditional distance cost based on the concept of equivalent horizontal distance. This transforms the impact of terrain slope on travel speed into a unified cost expression form, making the path evaluation results more consistent with actual travel costs.

[0051] Specifically, assuming the vehicle is on a slope of... When driving on a slope, its speed is expressed as: in, It is the speed at which the vehicle travels on a slope; It is the speed of driving on a flat road (the speed when the gradient is 0). It is a constant representing the degree to which the gradient affects driving speed; It is the tangent of the terrain slope. The velocity model is based on the objective law that the greater the slope, the greater the resistance and the lower the speed during actual mountain travel. The influence of slope is introduced into the velocity expression through a linear decay form, so that the model has good computational stability and engineering applicability.

[0052] Based on this, the relationship between unit slope distance and equivalent horizontal distance can be obtained as follows: Where S represents the actual distance traveled on the slope, and S0 represents the equivalent horizontal ground distance. The purpose of this relationship is to uniformly convert slope paths into a horizontally comparable scale, enabling paths under different slope conditions to be compared under a unified evaluation system, thereby avoiding the evaluation distortion caused by using only geometric distance.

[0053] Based on the above equivalence relationship, the traditional cumulative cost function is modified to obtain the improved cumulative cost expression: in, Indicates the first in the path Spatial distance between adjacent grid cells Indicates the first The tangent of the slope corresponding to the path segment. This indicates the vehicle's speed on a flat road, specifically the speed when the gradient is 0. This represents the slope impact coefficient. The actual distance of each path segment is dynamically recalculated according to its corresponding slope, so that path segments with steeper slopes account for a higher proportion of the cost. This automatically reduces the priority of high-slope areas during the path search process, improving the overall path accessibility and engineering rationality.

[0054] Spatial distance between adjacent grids It is obtained through three-dimensional Euclidean distance calculation, and the calculation formula is as follows: in, , and These represent the first and second parts of the path, respectively. The algorithm calculates the column coordinates, row coordinates, and elevation values ​​of each grid cell. It also considers horizontal movement distance and elevation changes to ensure the path length has a consistent physical meaning in three-dimensional space. This improves the accuracy and consistency of path cost calculation and provides a reliable basic input for slope-corrected cost models.

[0055] In one embodiment, the slope influence coefficient The slope is used to characterize the intensity of the impact of the slope on the vehicle or the speed of travel. Its value is not a fixed constant, but is dynamically or semi-empirically set according to the actual application scenario, so as to ensure that the path cost model can be adapted to different engineering conditions and driving objects, thereby improving the engineering applicability and robustness of the path planning results.

[0056] Specifically, the slope influence coefficient The determination of the value takes into account vehicle characteristics, road conditions, slope range, and empirical data or experimental results.

[0057] Regarding vehicle characteristics, different types of vehicles exhibit significant differences in their power response in mountainous environments. For example, heavy vehicles, due to their larger mass or relatively insufficient power output, experience a more pronounced speed decrease as the gradient increases, while off-road vehicles, possessing stronger power output and off-road capabilities, are relatively less sensitive to changes in gradient. Therefore, when a vehicle has a large mass, low engine power, or weak driving performance, the corresponding... The value is relatively large to enhance the penalty effect of slope on path cost, making the planned path more in line with actual traffic capacity constraints.

[0058] Regarding road surface conditions, the coefficient of friction and the smoothness of the road surface directly affect a vehicle's actual driving ability under incline conditions. When the road surface is slippery, loose, or uneven, the vehicle's adhesion on the slope decreases, and the speed drop is more significant. In this case, it is necessary to increase... The value is used to enhance the impact of slope on cost; conversely, under conditions of relatively good road surface conditions and a high coefficient of friction, The value is relatively low to avoid excessively penalizing path costs, thus maintaining the rationality of path search.

[0059] Regarding the gradient range, the impact of different gradient intervals on vehicle speed exhibits a non-linear trend. Within a shallow gradient range, vehicle speed changes are relatively small, and the gradient's influence is relatively weak. Take the smaller value; within a moderate slope range, the speed begins to decrease significantly. Take a moderate value; within a larger gradient range, the speed decreases significantly, even approaching impassable conditions, therefore... A larger value is chosen to strengthen the penalty effect on high-slope areas, thereby automatically avoiding high-risk areas during path search.

[0060] In terms of empirical data and experimental calibration, The specific values ​​were obtained through actual measurements of the driving speeds of different vehicle types under varying gradient conditions or through statistical analysis of historical data. By establishing a model relating gradient to speed, the data was fitted and analyzed to obtain values ​​that closely match actual operating conditions. Value range. This method avoids deviations caused by purely theoretical settings, making the model more closely resemble real-world engineering applications.

[0061] By comprehensively determining the above-mentioned multi-factor method, the slope influence coefficient is... It can adaptively adjust according to different application scenarios, so that the path cost model can maintain good applicability and stability under different terrain conditions and different vehicle conditions, and further improve the rationality and engineering feasibility of the path planning results based on slope correction.

[0062] Example 2: The following is a verification example of the present invention, applied to mountain path planning based on the improved A* algorithm, to verify that the present invention can achieve path planning in complex mountain environments.

[0063] Table 1 Simulation Parameters Simulation parameters are shown in Table 1. The interface of the point cloud data preprocessing system for mountain path planning based on the A* algorithm is shown in Table 1. Figure 2 As shown, the start and end coordinates input here are the coordinates provided by the actual user. After coordinate offset processing, the corresponding coordinate positions used in the subsequent A* algorithm path planning are obtained, namely, the start (308, 328, 22) and the end (168, 442, 52). Figure 2 As can be seen, after downsampling and region clipping, the point cloud data was reduced from 54,277,817 to 123,974. Simulation results show that the method used in this invention can effectively reduce the path planning time of the A* algorithm and improve path planning efficiency. The path planning results are as follows: Figure 3 and Figure 4 As shown. According to Figure 3 , Figure 4 The simulation results show that the present invention has successfully implemented the path planning function in complex mountainous environments and displays the corresponding coordinate points found, which provides convenience for path planning.

[0064] This invention focuses on path planning in complex mountainous environments. By introducing UAV LiDAR technology, it verifies that the proposed method can achieve efficient path planning in such conditions. Through preprocessing of point cloud data, the proposed method significantly reduces the amount of data required by the A* algorithm, thereby effectively reducing the time required for path planning.

[0065] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0066] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A mountain path planning method based on an improved A* algorithm, characterized in that, Includes the following steps: Acquire drone data of the target area and generate point cloud data. Construct a digital elevation model based on the point cloud data to obtain a mountain model containing elevation and slope information. The point cloud data is preprocessed, and the mountain model is cropped according to the starting and ending positions to obtain the target area data. The target area data is then divided into grids to construct a grid map. Based on the grid map, an improved A* algorithm is used for path search. During the path search process, a path cost is constructed, which includes path distance, slope, and elevation factors. During the path search process, the path cost is corrected based on the slope, and areas exceeding preset conditions are restricted to obtain the path planning result from the starting point to the ending point.

2. The mountain path planning method based on the improved A* algorithm according to claim 1, characterized in that, The process of acquiring UAV data of the target area and generating point cloud data, and constructing a digital elevation model based on the point cloud data to obtain a mountain model containing elevation and slope information, specifically includes: The target area is scanned using lidar sensors or image acquisition equipment carried by the drone to obtain raw remote sensing data. The original remote sensing data is subjected to coordinate registration and spatial alignment to unify the coordinate system and obtain initial point cloud data; The initial point cloud data is filtered to remove noise points and outliers, and ground point data is extracted using a classification algorithm. A digital elevation model is constructed based on the ground point data, and terrain slope information is calculated by the elevation difference between adjacent points, wherein the slope at any location is determined based on the rate of change of elevation between that location and its neighboring points. The elevation and slope information in the digital elevation model are fused to generate a mountain model for route planning.

3. The mountain path planning method based on the improved A* algorithm according to claim 2, characterized in that, The slope is calculated using the following formula: in, This is the tangent of the slope; This indicates the location in the digital elevation model. The elevation value at that location, This indicates the spatial resolution between adjacent grid cells.

4. The mountain path planning method based on the improved A* algorithm according to claim 1, characterized in that, The preprocessing of the point cloud data and the cropping of the mountain model based on the starting and ending points to obtain the target area data specifically include: The point cloud data is subjected to coordinate offset correction to unify the point cloud data to a preset reference coordinate system, and the point cloud coordinates are discretized to map continuous coordinates to discrete grid coordinates. The point cloud data is downsampled by dividing the space using a voxel grid filtering method and selecting representative points within each voxel unit to reduce the amount of point cloud data. The side length of the voxel unit is set to a preset resolution parameter. Based on the input start and end positions, a rectangular region encompassing the start and end positions is determined in the mountain model. The boundary range of the rectangular region satisfies: in, and These represent the planar coordinates of the starting and ending positions, respectively. and This is the preset extended range parameter; The point cloud data is cropped based on the rectangular region to obtain the target region data, while retaining the elevation and slope information within the target region.

5. A mountain path planning method based on an improved A* algorithm according to claim 1, characterized in that, The improved A* algorithm includes the following steps: Define a start grid and an end grid in the grid map, create an open list and a closed list, and add the start grid to the open list; Select the grid with the minimum evaluation function value from the open list as the current grid and add the current grid to the closed list. Then, traverse the adjacent grids of the current grid. The adjacent grids of the current grid include the grids that are adjacent to the current grid in the row and column coordinate directions, and the grids that are adjacent to the current grid in the diagonal direction. The accessibility of the adjacent grid cells is determined. If the adjacent grid cell is not visited and the access conditions are met, the adjacent grid cell is added to the open list, and its evaluation function value and parent node information are updated. Repeat the adjacent grid expansion process until the endpoint grid is added to the closed list to obtain the optimal path.

6. A mountain path planning method based on an improved A* algorithm according to claim 5, characterized in that, The evaluation function is defined as follows: in, Indicates the current grid The evaluation function value; This indicates the distance from the starting grid cell to the current grid cell. The cumulative path cost, Indicates starting from the current grid The estimated cost to the endpoint grid.

7. A mountain path planning method based on an improved A* algorithm according to claim 6, characterized in that, The estimated cost The formula is: in, and These represent the current grid cells. Column and row coordinates, and These represent the column and row coordinates of the endpoint grid, respectively. Indicates the current grid Elevation value, Indicates the elevation value of the endpoint grid. Indicates the spatial resolution of the raster.

8. A mountain path planning method based on an improved A* algorithm according to claim 6, characterized in that, The cumulative path cost The formula is: in, Indicates the first in the path Spatial distance between adjacent grid cells Indicates the first The tangent of the slope corresponding to the path segment. This indicates the vehicle's speed on a flat road, specifically the speed when the gradient is 0. This represents the slope influence coefficient.

9. A mountain path planning method based on an improved A* algorithm according to claim 8, characterized in that, The spatial distance between adjacent grids It is obtained through three-dimensional Euclidean distance calculation, and the calculation formula is as follows: in, , and These represent the first and second parts of the path, respectively. The column coordinates, row coordinates, and elevation values ​​of each grid cell.

10. A mountain path planning method based on an improved A* algorithm according to claim 8, characterized in that, The slope influence coefficient The determination is based on vehicle characteristics, road conditions, and gradient range, specifically including: The slope influence coefficient based on vehicle characteristics The basic settings include vehicle mass, power output capability, and drive mode. A higher vehicle mass or lower power output capability corresponds to a higher gradient influence coefficient. The larger the value; The influence coefficient of slope based on road conditions The correction is made, and the road surface conditions include the road surface friction coefficient and the road surface smoothness. When the road surface friction coefficient is low or the road surface smoothness is poor, the slope influence coefficient is increased. The value of ; The slope influence coefficient based on the slope range Adaptive adjustments are made; when the slope is within a relatively small, pre-set range, a lower slope influence coefficient is adopted. When the slope is within a second preset large range, the slope influence coefficient is increased. .

Images