Probe vehicle path planning method and device, storage medium and electronic equipment

Abstract

The application discloses a kind of detection vehicle path planning method and device, storage medium and electronic equipment.It relates to path planning technical field, the method includes: based on the traffic capacity threshold of detection vehicle is traversed to the digital elevation map of target area, obtain target obstacle grid chart, wherein the grid in the target obstacle grid chart includes passable grid and impassable grid;The distance between each passable grid in the target obstacle grid chart and the nearest obstacle point is calculated, to obtain distance distribution chart;Based on the distance distribution chart, the obstacle density corresponding to each passable grid is calculated respectively;Based on the distance distribution chart and the obstacle density, obtain target cost function, and based on the target cost function, path solving is carried out, to obtain the planning path of the detection vehicle in the target area.Through the present application, the problem of poor detection vehicle path planning in the related art is solved.

Description

Technical Field

[0001] This application relates to the field of path planning technology, and more specifically, to a path planning method and apparatus for a probe vehicle, a storage medium, and electronic equipment. Background Technology

[0002] Path planning technology is a core supporting technology for surface exploration by lunar rovers, Mars rovers, and other extraterrestrial surface probes. Its development level directly restricts the ability to conduct wide-area, multi-point scientific sampling and construct research stations. However, existing technologies have significant shortcomings in engineering applications.

[0003] Existing continuous cost models assume that continuous changes in terrain parameters correspond to continuous changes in safety risks, neglecting that within the vehicular mobility range, the distance between the path and obstacles is the core factor determining the collision probability. This cognitive bias leads to generated paths often closely following obstacle boundaries, violating the safety intuition of ground operators, and requiring complex smoothing processes to eliminate inflection points, severely limiting real-time planning and rapid replanning capabilities. Furthermore, existing safety-first A... The algorithm's global static parameter configuration mode cannot adapt to the extremely strong spatial heterogeneity of the lunar surface, resulting in drastic performance fluctuations under different terrains. This not only increases the cost for ground engineers to manually adjust parameters, but also severely restricts the rover's emergency response capability and long-distance wide-area autonomous exploration level in complex and ever-changing environments.

[0004] There is currently no effective solution to the technical problem of poor path planning for probes in related technologies. Summary of the Invention

[0005] The main objective of this application is to provide a method and apparatus for path planning of a probe vehicle, a storage medium and an electronic device, in order to solve the technical problem of poor path planning for probe vehicles in related technologies.

[0006] To achieve the above objectives, according to one aspect of this application, a path planning method for a probe vehicle is provided. The method includes: traversing a digital elevation map of a target area based on a passability threshold of the probe vehicle to obtain a target obstacle grid map, wherein the grids in the target obstacle grid map include passable grids and impassable grids; calculating the distance between each passable grid in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map; calculating the obstacle density corresponding to each passable grid based on the distance distribution map; obtaining a target cost function based on the distance distribution map and the obstacle density; and solving for the path based on the target cost function to obtain the planned path of the probe vehicle in the target area.

[0007] Furthermore, based on the accessibility threshold of the detection vehicle, the digital elevation map of the target area is traversed to obtain the target obstacle raster map, including: identifying terrain obstacles on the digital elevation map of the target area and obtaining the identification results; traversing and dividing the digital elevation map according to the accessibility threshold and the identification results to obtain the initial obstacle raster map; and expanding the obstacle raster in the initial obstacle raster map through dilation operation to obtain the target obstacle raster map.

[0008] Furthermore, based on the distance distribution map, the obstacle density corresponding to each passable grid is calculated, including: obtaining the target radius, and obtaining the neighborhood space based on the target radius with each passable grid as the center; filtering the grids in the neighborhood space to obtain the filtered target grids; obtaining the target distance value corresponding to the filtered target grids from the distance distribution map; and calculating the obstacle density based on the target distance value.

[0009] Furthermore, based on the distance distribution map and obstacle density, the target cost function is obtained, including: calculating the dynamic weight value based on the obstacle density corresponding to the passable grid; calculating the cumulative cost function based on the dynamic weight value and the distance distribution map; calculating the heuristic cost function based on the distance distribution map; and obtaining the target cost function based on the cumulative cost function and the heuristic cost function.

[0010] Furthermore, dynamic weight values ​​are calculated based on the obstacle density corresponding to the passable grid, including: obtaining the adjustment range and adjustment coefficient of the dynamic weight; and calculating the dynamic weight values ​​based on the adjustment range, adjustment coefficient, and obstacle density.

[0011] Furthermore, based on the dynamic weight values ​​and the distance distribution map, the cumulative cost function is calculated, including: calculating the path length cost based on the grid resolution and length weight; calculating the safety cost based on the distance values ​​in the distance distribution map; and obtaining the cumulative cost function based on the dynamic weight values, the safety cost, and the path length cost.

[0012] Furthermore, obtaining the target radius includes: obtaining the maximum envelope radius of the probe vehicle, the preset safe distance threshold, and the reserved sensing margin; determining the upper limit constraint value and the lower limit constraint value based on the safe distance threshold; and obtaining the target radius based on the maximum envelope radius, the safe distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0013] Furthermore, the method also includes: determining whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; if there is a target passable grid whose distance value is less than the safe distance threshold, then setting the target passable grid as a target state.

[0014] In order to achieve the above object, according to another aspect of the present application, a path planning device for a survey vehicle is provided, comprising: a traversal unit configured to traverse a digital elevation map of a target area based on a threshold of passability of the survey vehicle to obtain a target obstacle grid map, wherein a grid in the target obstacle grid map comprises a passable grid and a non-passable grid; a first calculation unit configured to calculate a distance between each passable grid in the target obstacle grid map and a nearest obstacle point to obtain a distance distribution map; a second calculation unit configured to calculate an obstacle density corresponding to each passable grid based on the distance distribution map; and a solving unit configured to obtain a target cost function based on the distance distribution map and the obstacle density, and to solve a path based on the target cost function to obtain a planned path of the survey vehicle in the target area.

[0015] Further, the traversal unit comprises: an identification module configured to identify terrain obstacles of the digital elevation map of the target area to obtain an identification result; a division module configured to traverse and divide the digital elevation map according to the threshold of passability and the identification result to obtain an initial obstacle grid map; and an expansion module configured to expand the obstacle grids in the initial obstacle grid map by a dilation operation to obtain the target obstacle grid map.

[0016] Further, the second calculation unit comprises: a first acquisition module configured to acquire a target radius and obtain a neighborhood space based on the target radius with each passable grid as a center; a screening module configured to screen the grids in the neighborhood space to obtain a screened target grid; a second acquisition module configured to acquire a target distance value corresponding to the screened target grid from the distance distribution map; and a first calculation module configured to calculate the target distance value to obtain the obstacle density.

[0017] Further, the solving unit comprises: a second calculation module configured to calculate a dynamic weight value based on the obstacle density corresponding to the passable grid; a third calculation module configured to calculate a cumulative cost function based on the dynamic weight value and the distance distribution map; a fourth calculation module configured to calculate a heuristic cost function based on the distance distribution map; and a determination module configured to obtain the target cost function based on the cumulative cost function and the heuristic cost function.

[0018] Further, the second calculation module comprises: a first acquisition submodule configured to acquire an adjustment interval of the dynamic weight and an adjustment coefficient; and a first calculation submodule configured to calculate the dynamic weight value based on the adjustment interval, the adjustment coefficient and the obstacle density.

[0019] Furthermore, the third calculation module includes: a second calculation submodule, used to calculate the path length cost based on the grid resolution and length weight; a third calculation submodule, used to calculate the safety cost based on the distance value in the distance distribution map; and a first determination submodule, used to obtain the cumulative cost function based on the dynamic weight value, the safety cost, and the path length cost.

[0020] Furthermore, the first acquisition module includes: a second acquisition submodule, used to acquire the maximum envelope radius of the probe vehicle, a preset safe distance threshold, and a reserved sensing margin; a second determination submodule, used to determine the upper limit constraint value and the lower limit constraint value based on the safe distance threshold; and a third determination submodule, used to obtain the target radius based on the maximum envelope radius, the safe distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0021] Furthermore, the device also includes: a judgment unit, used to judge whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; and a setting unit, used to set the target passable grid as a target state if the distance value corresponding to the target passable grid is less than the safe distance threshold.

[0022] To achieve the above objectives, according to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to execute the above-described probe path planning method.

[0023] To achieve the above objectives, according to another aspect of this application, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the above-described probe path planning method during runtime.

[0024] In this embodiment, the following steps are employed: traversing the digital elevation map of the target area based on the accessibility threshold of the probe vehicle to obtain a target obstacle grid map, wherein the grids in the target obstacle grid map include passable grids and impassable grids; calculating the distance between each passable grid in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map; calculating the obstacle density corresponding to each passable grid based on the distance distribution map; obtaining the target cost function based on the distance distribution map and the obstacle density, and solving the path based on the target cost function to obtain the planned path of the probe vehicle in the target area, thus solving the technical problem of poor path planning for probe vehicles in related technologies.

[0025] In this application, the digital elevation map of the target area is traversed based on the mobility threshold of the probe vehicle, simplifying the complex terrain into a target obstacle grid map containing both traversable and impassable grids. This addresses the problem in existing continuous cost models where the impact of terrain parameter differences on mobility safety is over-quantified, leading to planning results that do not conform to engineering intuition. Then, the distance between each traversable grid in the target obstacle grid map and the nearest obstacle point is calculated to obtain a distance distribution map. Based on this distance distribution map, the obstacle density corresponding to each traversable grid is calculated, enabling path planning to directly use the distance to the obstacle as the core safety indicator. This solves the problem in related technologies where the algorithm-generated path closely follows the obstacle boundary, resulting in insufficient safety. Finally, a target cost function is constructed based on the distance distribution map and obstacle density, and path solving is performed based on this target cost function. This effectively solves the technical problem of poor probe vehicle path planning, achieving the effect of improving the safety, adaptability, and rationality of path planning. Attached Figure Description

[0026] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0027] Figure 1 This is a hardware structure block diagram of a computer terminal for implementing a path planning method for a probe vehicle, according to an embodiment of this application.

[0028] Figure 2 This is a flowchart of the probe path planning method provided according to the embodiments of this application;

[0029] Figure 3 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 1 ;

[0030] Figure 4 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 2 ;

[0031] Figure 5 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 3 ;

[0032] Figure 6 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 4 ;

[0033] Figure 7 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 5 ;

[0034] Figure 8 This is a schematic diagram of the probe path planning method provided in the embodiments of this application. Figure 6 ;

[0035] Figure 9 This is a schematic diagram of a probe path planning device provided according to an embodiment of this application;

[0036] Figure 10 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0037] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0038] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0039] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0040] A The A-Star Algorithm is a heuristic search algorithm widely used in graph search and graph traversal. It efficiently finds the optimal path between two points by combining the actual cost (g value) from the starting point to the current node and the estimated cost (h value, i.e., the heuristic function) from the current node to the target node to evaluate node priorities.

[0041] Example 1

[0042] According to an embodiment of this application, a method embodiment for path planning of a probe vehicle is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0043] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing a path planning method for a probe vehicle is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0044] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0045] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the probe path planning method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the above-mentioned probe path planning method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0046] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0047] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0048] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for planning the path of the probe vehicle is shown. Figure 2 This is a flowchart of a probe path planning method according to Embodiment 1 of this application. The probe path planning method includes:

[0049] Step S201: Based on the accessibility threshold of the detection vehicle, the digital elevation map of the target area is traversed to obtain the target obstacle grid map, wherein the grid in the target obstacle grid map includes passable grids and impassable grids.

[0050] Optionally, a high-precision digital elevation map of the target area can be acquired. This digital elevation map can be composed of regularly arranged grid cells, each storing a corresponding altitude value. A traffic capacity threshold model is then established by combining the physical parameters and motion performance indicators of the exploration vehicle itself. This model can include key parameters such as the exploration vehicle's maximum permissible climbing angle, maximum obstacle-crossing height, and minimum turning radius.

[0051] The digital elevation data is traversed and analyzed grid by grid. For each grid cell, it is compared with a preset accessibility threshold. If the maximum slope of a grid cell exceeds the maximum climbing angle allowed by the vehicle, or if the terrain undulation height within the grid cell exceeds the maximum obstacle-crossing height of the vehicle, the grid cell is determined to be an impassable area and marked as an impassable grid cell in the target obstacle grid map. Conversely, if all terrain feature parameters are within the vehicle's accessibility range, the grid cell is determined to be an accessible area and marked as an accessible grid cell.

[0052] In some embodiments, the target obstacle grid map can be a binary obstacle grid map, in which the value of passable grids is set to a first value (e.g., 0), and the value of impassable grids is set to a second value (e.g., 1). This binarization process eliminates the complex weighting of minor terrain differences within passable areas in traditional methods, simplifying continuous terrain costs into binary logical judgments, thereby reducing the computational complexity of subsequent path planning algorithms and ensuring the consistency between the terrain model and the actual passability of the exploration vehicle.

[0053] Step S202: Calculate the distance between each passable grid cell in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map.

[0054] Optionally, a set of obstacle points in the target obstacle grid map is determined. An obstacle point can be the center point of a grid cell marked as impassable. Then, a fast sweep method can be used to batch process the target obstacle grid map to calculate the distance from each traversable grid cell to the nearest obstacle point. The fast sweep method traverses the entire grid map through multiple linear scans and may specifically include the following steps:

[0055] The first scan starts from the top left corner of the target obstacle grid and traverses each grid cell to the right and down in row-first order. For each passable grid cell, the shortest distance estimate for the current cell is updated based on the distance values ​​of its adjacent grid cells (above, left, and top-left). The second scan starts from the bottom right corner of the target obstacle grid and traverses each grid cell to the left and up in row-first order. Similarly, the shortest distance estimate for the current cell is updated based on the distance values ​​of its adjacent grid cells (below, right, and bottom-right). The third and fourth scans perform similar traversal updates along the diagonal direction to further refine the distance estimates and ensure the accuracy of the Euclidean distance calculation.

[0056] Through the above multiple scans, the Euclidean distance from each passable grid cell to the nearest obstacle point is obtained, thus generating a distance distribution map.

[0057] In some embodiments, to reduce computational complexity and improve real-time performance, the distance distribution map is calculated offline before path planning begins. After pre-calculation, during path search, the obstacle distance to any passable grid can be retrieved simply by looking up a table, eliminating the need for real-time calculation and thus improving the algorithm's efficiency.

[0058] Furthermore, each value in the distance distribution map represents the safety margin at that grid point. The larger the distance value, the farther the location is from the obstacle, and the higher the safety; the smaller the distance value, the closer the location is to the obstacle, and the higher the potential risk.

[0059] In some embodiments, full processing is performed based on the previously generated binary barrier raster map. Figure 1 In a single batch processing step, the Euclidean distance from all passable grid cells to the nearest obstacle point can be calculated through four orthogonal linear scans. The resulting distance distribution map is shown below. Figure 3 As shown, its overall time complexity is only O(N) (where N is the total number of grid cells), which is far superior to the O(NlogN) complexity of traditional distance transformation algorithms. It is especially suitable for the computational needs of large-scale grid maps with a scale of hundreds of meters × hundreds of meters in the exploration mission of the probe vehicle. After the pre-calculation is completed, each passable grid cell will store its accurate Euclidean distance value to the nearest obstacle point. In the subsequent path search process, the obstacle distance query of any node can be completed by simply looking up the table in memory, realizing efficient access with a constant time complexity of O(1).

[0060] Step S203: Based on the distance distribution map, calculate the obstacle density corresponding to each passable grid cell.

[0061] Optionally, using a distance distribution map, the obstacle density corresponding to each passable grid cell can be calculated. For example, a neighborhood region can be extracted from the distance distribution map centered on the current passable grid cell. This neighborhood region serves as a sampling space for assessing the local congestion level. All grid cells within this neighborhood region are traversed, and the number of impassable grid cells is determined. Based on the number of impassable grid cells, the obstacle density corresponding to each passable grid cell can be calculated.

[0062] In some embodiments, each grid cell in the neighborhood may be weighted according to its distance value in the distance distribution map (e.g., the weight is inversely proportional to the distance, or only the barrier grid is weighted according to the distance) to quantify the contribution of the grid to the local environmental congestion.

[0063] Finally, an obstacle density distribution map is output. The calculated obstacle density value for each passable grid cell is stored in the corresponding grid attribute to form a complete obstacle density distribution map.

[0064] Step S204: Based on the distance distribution map and obstacle density, the target cost function is obtained, and the path is solved based on the target cost function to obtain the planned path of the probe vehicle in the target area.

[0065] Optionally, based on A The basic framework of a search algorithm is to define a target cost function, which typically consists of two parts: a cumulative cost and a heuristic cost. For example, the target cost function is: Among them, the cost of inspiration : Use the Euclidean distance from node n to the endpoint, satisfying A The algorithm's admissibility condition guarantees that the final path is globally optimal; cumulative cost. This combines security penalties with path length penalties. Cumulative cost. Calculations are performed based on the distance distribution map and obstacle density.

[0066] For example, the Euclidean distance value corresponding to the current search node n is read from the distance distribution map. This value represents the straight-line distance from node n to the nearest obstacle. A basic safety cost term is constructed, which is inversely proportional to the Euclidean distance value; that is, the smaller the distance value, the larger the basic safety cost term; and vice versa. This relationship reflects that the closer a node is to an obstacle, the higher the risk of collision, and therefore the higher the cost. Then, a safety penalty term is calculated based on the obstacle density. Finally, the cumulative cost is obtained based on the basic safety cost term and the safety penalty term.

[0067] Finally, pathfinding is performed. Starting with the origin as the initial node and the destination as the target node, the search algorithm is initiated using the constructed target cost function. During the search, candidate nodes are expanded sequentially, and the total cost f(n) for each candidate node is calculated. The node with the lowest total cost is prioritized for expansion until the destination is reached or the maximum number of iterations is achieved. The parent node pointers recorded during the search are backtracked to generate a discrete path point sequence from the origin to the destination. This discrete path point sequence is then smoothed to obtain the final planned path of the probe in the target area.

[0068] In summary, this method traverses the digital elevation map of the target area based on the vehicle's accessibility threshold, simplifying complex terrain into a target obstacle grid map containing both passable and impassable grids. This addresses the problem in existing continuous cost models where the impact of terrain parameter differences on mobility safety is over-quantified, leading to planning results that do not conform to engineering intuition. Then, the distance between each passable grid in the target obstacle grid map and the nearest obstacle point is calculated to obtain a distance distribution map. Based on this distance distribution map, the obstacle density corresponding to each passable grid is calculated, allowing path planning to directly use the distance to obstacles as the core safety indicator. This solves the problem in related technologies where algorithm-generated paths closely follow obstacle boundaries, resulting in insufficient safety. Finally, a target cost function is constructed based on the distance distribution map and obstacle density, and path solving is performed based on this target cost function. This effectively solves the technical problem of poor path planning for the vehicle, improving the safety, adaptability, and rationality of path planning.

[0069] Optionally, in the path planning method for the detection vehicle provided in this application embodiment, the digital elevation map of the target area is traversed based on the traffic capacity threshold of the detection vehicle to obtain a target obstacle grid map, including: performing terrain obstacle identification on the digital elevation map of the target area to obtain identification results; traversing and dividing the digital elevation map according to the traffic capacity threshold and the identification results to obtain an initial obstacle grid map; and expanding the obstacle grids in the initial obstacle grid map through dilation operations to obtain the target obstacle grid map.

[0070] In some embodiments, the digital elevation map of the target area is traversed according to the traffic capacity threshold of the detection vehicle to obtain a target obstacle grid map, specifically including the following steps:

[0071] First, terrain obstacle identification is performed on the digital elevation map of the target area to obtain the identification results. For example, each raster cell in the digital elevation map is traversed, and the terrain feature value of the current raster is calculated to obtain the above-mentioned identification results. The terrain feature value includes, but is not limited to, the slope, aspect, and surface roughness of the current raster.

[0072] Then, based on the accessibility threshold and the recognition results, the digital elevation map is traversed and divided. The calculated terrain feature values ​​are compared with the maximum climbing angle, maximum obstacle crossing height, and maximum surface roughness tolerance value preset by the detection vehicle. If any terrain feature value exceeds the corresponding threshold, the grid is determined to be an impassable area; if all terrain feature values ​​do not exceed the corresponding threshold, the grid is determined to be a passable area, thus obtaining the initial obstacle grid map.

[0073] Based on the initial obstacle grid map, various uncertainty error sources throughout the entire mission cycle of the exploration vehicle are quantified, including visual fusion positioning errors, wheel drive system motion control execution errors, DEM terrain modeling resolution errors, and engineering-reserved redundancy safety margins. Through dilation operations in mathematical morphology, all impassable grids are proportionally expanded, transforming the originally sharp, hard obstacle boundaries into soft boundaries with safety buffers, effectively mitigating the risk of accidental collisions that may result from the superposition of various system errors. In other words, the obstacle grids in the initial obstacle grid map are expanded through dilation operations to obtain the target obstacle grid map, such as... Figure 4 As shown, the complex multi-dimensional and nonlinear cost calculation problem in the traditional continuous cost model is simplified into a single binary logic judgment problem. This not only significantly reduces the computational complexity of subsequent Euclidean distance field pre-calculation, node cost evaluation and global path search from the source, but also eliminates the problem of dense path inflection points caused by the irregularity of the traditional cost graph, reduces the computational burden of subsequent path smoothing processing, and lays a solid data foundation for achieving high real-time global safe path planning.

[0074] By combining terrain obstacle recognition, binarization, and morphological dilation, continuous terrain elevation data can be transformed into discrete binary obstacle models. This eliminates minor terrain differences with limited safety impact within passable areas, simplifying subsequent computational complexity. Simultaneously, the safety buffer zone constructed using dilation can accommodate positioning errors, control errors, and terrain modeling errors of the exploration vehicle, transforming rigid physical obstacle boundaries into soft boundaries with safety redundancy. This effectively reduces the risk of accidental collisions caused by perception noise or execution deviations.

[0075] Optionally, in the path planning method for the detection vehicle provided in this application embodiment, the obstacle density corresponding to each passable grid is calculated based on the distance distribution map, including: obtaining the target radius, and obtaining the neighborhood space based on the target radius with each passable grid as the center; filtering the grids in the neighborhood space to obtain the filtered target grids; obtaining the target distance value corresponding to the filtered target grids from the distance distribution map; and calculating the obstacle density based on the target distance value.

[0076] In some embodiments, calculating the obstacle density corresponding to each passable grid cell specifically includes the following steps:

[0077] First, the target radius is obtained, and a neighborhood space is derived based on this radius, centered on each passable grid cell. The target radius can be a globally fixed constant, its value determined by the rover's body envelope radius, a preset safety distance threshold, and a reserved sensing margin. By setting a fixed target radius, a local sensing field of view with a uniform scale, i.e., the neighborhood space, is defined for each passable grid cell. This neighborhood space is a circular area, ensuring that the sampling range of local environmental features remains consistent across different locations and terrain scenarios, thereby guaranteeing the consistency of the benchmark for subsequent calculations.

[0078] Next, the grids in the neighborhood space are filtered to obtain the target grids. Within the defined neighborhood space, not all grids participate in the obstacle density calculation. First, grids belonging to impassable obstacle areas are removed because obstacle density reflects the environmental complexity surrounding traversable areas, not the attributes of the obstacles themselves. Second, high-risk grids that are less than the safe distance threshold from the nearest obstacle are removed; these areas are considered absolute danger zones and are not included in the regular environmental complexity assessment. The remaining valid traversable grids are the filtered target grids, representing the normal traversable areas around the current node that have reference value.

[0079] Then, the target distance values ​​corresponding to the filtered target grids are obtained from the distance distribution map. The distance distribution map pre-calculates the Euclidean distance from each passable grid to the nearest obstacle. For each filtered target grid, its corresponding Euclidean distance value is queried and read from the distance distribution table. This target distance value reflects the physical distance between the target grid and the obstacle, and is the basic data for quantifying the degree of danger in the local environment.

[0080] Finally, the obstacle density is calculated by analyzing the target distance values. The target distance values ​​of all filtered target grids are statistically analyzed, and their arithmetic mean is calculated to obtain the neighborhood average obstacle distance. To eliminate the influence of different units and unify the numerical range, the neighborhood average obstacle distance is normalized and mapped to the interval [0, 1], ultimately yielding the obstacle density of the currently passable grid. The closer the obstacle density value is to 0, the farther the effective grids in the neighborhood are from the obstacles, indicating a relatively open and safe environment; the closer the obstacle density value is to 1, the closer the effective grids in the neighborhood are to the obstacles, indicating a more crowded environment and higher risk.

[0081] In some embodiments, a circular local neighborhood space is extracted centered on each grid point based on a predetermined global fixed sensing radius R, forming a local sensing field of view. To avoid invalid computation and noise interference, nodes within the neighborhood are first screened: obstacle nodes that belong to impassable areas and nodes whose distance from obstacles is less than a safety threshold are removed. For high-risk nodes, only valid and passable nodes are retained for calculation. Then, the Euclidean distance field values ​​of all valid nodes are counted, and their arithmetic mean is calculated to obtain the neighborhood average obstacle distance Davg(n). This is then mapped to the [0, 1] interval using a normalization formula to finally obtain the local obstacle density ρ(n) of the current node. Here, ρ(n) approaching 0 indicates that there are almost no obstacles in the neighborhood and the environment is inherently safe; ρ(n) approaching 1 indicates that the obstacles in the neighborhood are highly dense and the collision risk is extremely high, thus achieving accurate and continuous quantification of the local environmental hazard level. In some embodiments, R is a globally fixed constant that does not change dynamically with node position, scene type, or task stage. It is the core parameter that balances the rationality of environmental perception, the realism of density quantification, and the computational efficiency of the algorithm.

[0082] In some embodiments, the target radius can be fine-tuned based on the operating mode of the probe. For example, when the probe is performing high-precision fixed-point sampling, the target radius can be appropriately reduced to more sensitively capture the influence of small obstacles; while during long-distance patrol operations, the target radius can be appropriately increased to predict distant obstacle groups in advance and achieve earlier safety warnings. However, regardless of the mode, the target radius should remain fixed within a single planning task to ensure the stability of the algorithm.

[0083] In some embodiments, a weighting mechanism can be introduced into the filtered target grid. Target grids closer to the current node have a greater impact on the obstacle density of the current node and can be assigned higher weights; target grids farther away have a smaller impact and are assigned lower weights. Calculating the average obstacle distance in the neighborhood using a weighted average method allows the obstacle density to better reflect the environmental characteristics of the immediate vicinity, further improving the accuracy of local perception.

[0084] By constructing a unified local perception field of view by setting a target radius, and by screening impassable areas and high-risk areas, the interference of invalid data and extreme outliers can be eliminated, making the calculation results more focused on the environmental characteristics of normal traversable areas. By statistically analyzing the target distance values ​​of effective grids and normalizing them, the obstacle density is obtained, realizing the process of transforming the abstract local terrain complexity into a quantitative value.

[0085] Optionally, in the path planning method for the probe vehicle provided in this application embodiment, obtaining the target radius includes: obtaining the maximum envelope radius of the probe vehicle, a preset safety distance threshold, and a reserved sensing margin; determining an upper limit constraint value and a lower limit constraint value based on the safety distance threshold; and obtaining the target radius based on the maximum envelope radius, the safety distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0086] In some embodiments, obtaining the target radius specifically includes the following steps:

[0087] First, the maximum envelope radius of the probe vehicle, the preset safe distance threshold, and the reserved sensing margin are obtained. The maximum envelope radius refers to the radius of the largest circular area occupied by the probe vehicle's body, wheels, sensors, and all other protruding components in a stationary or moving state. This ensures that the probe vehicle will not physically collide with obstacles. The preset safe distance threshold is the minimum theoretical safe distance that the probe vehicle must maintain during operation to compensate for positioning errors, control errors, and terrain modeling errors. The reserved sensing margin is an additional buffer distance set based on experience to detect potential risks in advance; its value is usually proportional to the maximum envelope radius.

[0088] Secondly, based on the safe distance threshold, upper and lower constraint values ​​are determined. The lower constraint value is set to a certain multiple of the safe distance threshold, such as twice, to ensure that the neighborhood space is large enough to contain a sufficient number of effective sampling nodes. This avoids drastic fluctuations in obstacle density calculation results due to a single obstacle caused by an excessively small neighborhood, which could lead to instability in weight adjustment. The upper constraint value is set to another large multiple of the safe distance threshold, such as ten times, to prevent the neighborhood space from becoming too large and including distant, irrelevant terrain in the statistical scope, diluting the obstacle density characteristics of the current local area, while also avoiding excessive computational burden caused by too many node calculations.

[0089] Finally, the target radius is obtained based on the maximum envelope radius, the safe distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0090] In some embodiments, the preliminary calculation formula may be the sum of the maximum envelope radius, the safety distance threshold, and the reserved sensing margin. The preliminary calculated value is compared with pre-determined upper and lower constraint values. If the preliminary calculated value is less than the lower constraint value, the target radius is set as the lower constraint value; if the preliminary calculated value is greater than the upper constraint value, the target radius is set as the upper constraint value; if the preliminary calculated value is between the upper and lower constraint values, the calculated value is directly used as the target radius. Through this process, it is ensured that the target radius meets both the safety and risk avoidance requirements of the detection vehicle and the efficiency and stability requirements of the algorithm calculation, and remains globally fixed throughout the planning process without participating in dynamic adjustments.

[0091] In some embodiments, the essence of R is to define "the extent of the local environment that will affect the safety weight decision of the current node". Its value must simultaneously satisfy two mutually exclusive and complementary core conditions: First, perception coverage: the radius must be sufficient to cover the safety avoidance range of the probe vehicle, ensuring that all obstacles that may threaten driving safety can be perceived in advance, avoiding the failure of weight adjustment due to local missed detection; Second, computational efficiency and feature fidelity: the radius should not be too large to prevent cross-regional sampling from including irrelevant terrain in the statistical range, diluting the dense obstacle features of the current local area, while avoiding redundant calculations that increase the burden on the algorithm.

[0092] The value of R is determined based on the dimensions of the detection vehicle and safety and hazard avoidance requirements, using a unified engineering calculation formula:

[0093]

[0094] in: The maximum envelope radius of the detection vehicle body includes the maximum outer dimensions of all protruding components such as the fuselage, wheels, and sensor extension arms, ensuring coverage of the entire collision risk range of the vehicle body; The algorithm's preset safe distance threshold is the dangerous boundary that the detection vehicle must forcibly avoid, corresponding to the critical distance at which the collision risk increases exponentially. : Reserve a perception margin, typically 0.2 to 0.5 times the fuselage size, to offset terrain modeling errors and control response delays, ensuring that the algorithm can detect dense obstacles in advance and pre-adjust the safety weight to avoid sudden obstacle avoidance at close range.

[0095] To ensure the effectiveness of density calculation and the stability of weight adjustment, this application further sets strict numerical boundaries for R:

[0096] Lower bound constraint: If the radius is too small, the number of effective sampling nodes in the neighborhood will be insufficient, and the density ρ(n) will fluctuate drastically due to the presence of a single obstacle, introducing a large amount of noise. This will lead to frequent jumps in the dynamic weight α(n), causing problems such as path jitter and dense inflection points, which will seriously affect the driving stability.

[0097] Upper limit constraint: If the radius is too large, it will include open areas several times the safety distance in the sampling range, significantly diluting the current local obstacle density features, resulting in insufficient safety weight enhancement in high-risk areas and loss of local adaptive ability; at the same time, the number of neighboring sampling nodes increases with the square of the radius, and an excessively large R will greatly increase the computational load, offsetting the performance advantages of this invention.

[0098] Under the premise of meeting the above constraints, fine-tuning can be carried out in combination with the terrain features of different working environments to achieve optimal adaptation across all scenarios:

[0099] Structured and regular environments (such as open and flat areas): take the smaller value. These environmental obstacles are regularly distributed and have clear boundaries, allowing for precise perception of local risks with a small radius, while minimizing computational overhead.

[0100] Complex and chaotic environments (such as impact crater clusters and areas with dense rock): take a larger value. Expanding the perception range allows for the early prediction of areas with continuous and dense obstacles, enabling a higher safety priority and preventing the path from getting too close to obstacles or making sharp turns.

[0101] It should be noted that this application strictly stipulates that R must be globally fixed and cannot be changed in real time. This is a key design to ensure the stability and engineering practicality of the algorithm: First, the dynamic radius will change the baseline range of the maximum unobstructed distance Dmax within the perception radius, causing ρ(n) calculated by different nodes to be incomparable, completely destroying the adjustment logic of dynamic weights; Second, the fixed radius ensures the uniformity of the global sampling scale, so that the output of the algorithm in different locations and different scenarios has a consistent evaluation standard, which significantly improves the stability and reproducibility of the algorithm and greatly reduces the difficulty of engineering debugging and deployment.

[0102] By comprehensively considering the vehicle's dimensions, safety threshold, and perception margin, the target radius is determined, and an upper and lower limit constraint mechanism is introduced. This achieves a balance between meeting the vehicle's physical avoidance requirements and the algorithm's computational efficiency. The lower limit constraint avoids index fluctuations caused by an excessively small sampling range, while the upper limit constraint prevents the dilution of environmental features caused by an excessively large range. This ensures that the target radius is within a fixed and reasonable range, providing a consistent perception scale for the stable quantification of obstacle density and improving the adaptability and robustness of the path planning algorithm in different terrain scenarios.

[0103] Optionally, in the path planning method for the detection vehicle provided in this application embodiment, the target cost function is obtained based on the distance distribution map and obstacle density, including: calculating a dynamic weight value based on the obstacle density corresponding to the passable grid; calculating a cumulative cost function based on the dynamic weight value and the distance distribution map; calculating a heuristic cost function based on the distance distribution map; and obtaining the target cost function based on the cumulative cost function and the heuristic cost function.

[0104] In some embodiments, the target cost function is obtained based on the distance distribution map and obstacle density, specifically including the following steps:

[0105] If a globally fixed safety weight α is used, this design leads to an irreconcilable contradiction in its adaptability across all scenarios: if α is too large, the algorithm will excessively pursue safe distances, generating a large number of meaningless detours in open and flat areas, significantly increasing the energy consumption and mission time of the probe vehicle, as shown in the part at the end of the path in the upper right corner of the above figure; if α is too small, the algorithm is not sensitive enough to safe distances, and will still generate paths close to obstacles in areas with dense obstacles, failing to effectively avoid collision risks.

[0106] Therefore, dynamic weight values ​​are calculated based on the obstacle density corresponding to the passable grid. During path search, the pre-calculated obstacle density value is read for the currently evaluated passable grid. In some embodiments, a preset nonlinear mapping formula can be used to convert the obstacle density into a dynamic weight value. By introducing an adjustment coefficient, the steepness of the weight increase can be controlled, resulting in a gradual change in weight in sparse obstacle areas and a rapid increase in weight in dense obstacle areas. The resulting dynamic weight value can reflect the safety risk level of the local environment in which the current grid is located in real time.

[0107] Secondly, the cumulative cost function is calculated based on the dynamic weight values ​​and distance distribution map. The cumulative cost function measures the cumulative cost of the path from the starting point to the current grid cell. The cumulative cost function consists of a safety cost and a path length cost. The safety cost is obtained by dividing the dynamic weight values ​​by the Euclidean distance from the current grid cell to the nearest obstacle, while the path length cost is obtained by multiplying the current step size by a fixed length weight coefficient.

[0108] Then, the heuristic cost function is calculated based on the distance distribution map. The heuristic cost function is used to estimate the remaining cost from the current grid cell to the target point, guiding the search direction. For example, the Euclidean distance from the current grid cell to the target point can be used as the value of the heuristic cost function.

[0109] Finally, based on the cumulative cost function and the heuristic cost function, the target cost function is obtained. Adding the previously calculated cumulative cost function to the heuristic cost function yields the target cost function. The target cost function integrates the safety and length information of historical paths with the predicted information of future paths. By continuously iteratively updating the target cost function, the algorithm can gradually converge to the globally optimal path.

[0110] In some embodiments, the safety cost can be calculated using a piecewise function. The safety cost increases slowly when the grid is far from the obstacle; it increases as the grid approaches a safety threshold.

[0111] In some embodiments, the heuristic cost function can be weighted in addition to Euclidean distance, incorporating terrain slope information. For example, in areas with excessively steep slopes, increasing the weight of the heuristic cost guides the algorithm to prioritize paths with gentler slopes, thus achieving both shortest path optimization and smoother driving.

[0112] By introducing dynamic weight values ​​based on obstacle density calculation, the safety penalty term in the cumulative cost function can be adaptively adjusted according to changes in the local environmental risk level. This increases safety priority in obstacle-dense areas while reducing unnecessary safety penalties in open areas, thus effectively avoiding excessive path detours while ensuring a reasonable safe distance between the path and obstacles. Combined with the guiding role of the heuristic cost function on the global path, this achieves synergy between local environmental safety perception and global path efficiency optimization, making the planning results more in line with the dual requirements of safety and traffic efficiency in actual engineering projects.

[0113] Optionally, in the path planning method for the detection vehicle provided in this application embodiment, the dynamic weight value is calculated based on the obstacle density corresponding to the passable grid, including: obtaining the adjustment range and adjustment coefficient of the dynamic weight; and calculating the dynamic weight value based on the adjustment range, adjustment coefficient and obstacle density.

[0114] In some embodiments, a dynamic weight value is calculated based on the obstacle density corresponding to the passable grid, specifically including the following steps:

[0115] First, obtain the adjustment range and adjustment coefficient for the dynamic weights. The adjustment range consists of a preset minimum weight value and a maximum weight value. The minimum weight value is used to assign a lower safety penalty in areas with sparse obstacles, and the maximum weight value is used to assign a higher safety penalty in areas with dense obstacles. The adjustment coefficient is a value greater than 1, used to control the rate at which the weight increases with obstacle density. It can be determined based on experience or simulation tests to ensure that the weight curve is flat in low-density areas and steep in high-density areas.

[0116] Then, dynamic weight values ​​are calculated based on the adjustment interval, adjustment coefficient, and obstacle density. Using a pre-defined nonlinear mapping formula, obstacle density is used as the independent variable, and the adjustment interval and adjustment coefficient are used as parameters to calculate the current dynamic weight value. When the obstacle density is low, the calculated weight value is close to the minimum weight value; when the obstacle density is high, the calculated weight value quickly approaches the maximum weight value. The resulting dynamic weight value smoothly reflects the safety risk level of the local environment, avoiding fluctuations in path planning results caused by abrupt changes in weight.

[0117] In some embodiments, the dynamic weight α nonlinear adjustment formula is as follows: a power function-type continuous nonlinear adaptive formula is adopted to establish a positive mapping relationship between obstacle density and safety weight, achieving the adjustment effect of "gradual increase in low density and steep increase in high density", while ensuring that the weight is continuous without sudden changes, avoiding problems such as path jitter and abrupt inflection points:

[0118]

[0119] The adjustment coefficient γ (γ>1, typical value 2.0~3.0) is used to enhance the differentiated adjustment capability of the weights. The larger γ is, the more gradually α(n) increases in low-density areas, and the more rapidly α(n) rises in high-density areas. This accurately matches the "safety first" engineering requirements, minimizing safety penalties in open areas and drastically increasing safety priority in hazardous areas. The dynamic weights α(n) of all nodes are strictly constrained within a certain range. Within the range, the mechanism avoids the risk of weights exceeding the limit. It can be set based on engineering experience.

[0120] In some embodiments, an obstacle density threshold may be incorporated. (Threshold for open scenes, typical value 0.2~0.3) and (Threshold for dense scenes, typical value 0.7~0.8), the local environment is divided into three typical scenarios, and differentiated weight adjustment strategies are implemented to achieve fine-grained adaptation across all scenarios:

[0121] Open scene ( (Low density): There are very few obstacles in the neighborhood, and the risk of collision is negligible. The algorithm significantly reduces the impact of safety penalties, with the cost function penalizing based on path length. The algorithm prioritizes searching for the shortest path from the starting point to the destination, maximizing traffic efficiency and completely avoiding the excessive detours that traditional fixed high-weight algorithms encounter in open areas.

[0122] Normal scene ( (Medium density): There are a few scattered obstacles in the neighborhood, posing a slight risk of collision. As the density of obstacles increases in an approximately linear and smooth manner, the safety penalty term and the length penalty term are in a balanced state. The algorithm simultaneously takes into account both the safety distance and the path length, ensuring that the path is appropriately far away from obstacles to reserve basic safety redundancy, without sacrificing too much passage efficiency.

[0123] Dense dangerous scenes ( (High density): Obstacles are densely packed and complexly distributed within the neighborhood, resulting in an extremely high risk of collision. Safety penalties Occupying absolute dominance in the cost function, the algorithm will guide the path to actively and significantly move away from all obstacles.

[0124] By obtaining the adjustment range and adjustment coefficient of dynamic weights and calculating them using a preset nonlinear mapping formula based on obstacle density, the dynamic weight values ​​can smoothly transition with changes in the degree of danger in the local environment. In areas with sparse obstacles, the weights are kept low to maintain path efficiency, while in areas with dense obstacles, the weights are rapidly increased to enhance safety redundancy. This continuous and nonlinear adjustment mechanism avoids path jitter caused by sudden changes in weights, thereby improving the stability and rationality of the planning while ensuring path safety.

[0125] Optionally, in the path planning method for the probe provided in this application embodiment, the cumulative cost function is calculated based on the dynamic weight value and the distance distribution map, including: calculating the path length cost based on the grid resolution and length weight; calculating the safety cost based on the distance value in the distance distribution map; and obtaining the cumulative cost function based on the dynamic weight value, the safety cost, and the path length cost.

[0126] In some embodiments, the cumulative cost function is calculated based on dynamic weight values ​​and distance distribution maps, specifically including the following steps:

[0127] First, the path length cost is calculated based on grid resolution and length weight. During path search, each move from the current grid to an adjacent grid incurs a certain distance, determined by the grid map resolution, i.e., the Euclidean or Manhattan distance between the center points of adjacent grids. Simultaneously, a fixed length weight coefficient is preset to balance the proportion of path length in the total cost. Multiplying the distance value of the current step by the length weight coefficient yields the path length cost for the current step. This value reflects the physical distance cost incurred to reach the current grid, aiming to guide the algorithm to prioritize shorter routes.

[0128] Secondly, the safety cost is calculated based on the distance values ​​in the distance distribution map. The Euclidean distance from the current accessible grid cell to the nearest obstacle is read from the distance distribution map. To construct the safety cost, the reciprocal of this distance value can be used; that is, the smaller the distance, the larger the reciprocal, indicating a higher collision risk. Subsequently, the reciprocal of the read distance value is multiplied by the dynamic weight value calculated in the previous step to obtain the safety cost. The dynamic weight value plays a moderating role here: in areas with dense obstacles, the dynamic weight value is larger, amplifying the safety penalty caused by close proximity; in open areas, the dynamic weight value is smaller, reducing the impact of the safety penalty. The resulting safety cost value can quantitatively reflect the current safety risk level of the grid cell.

[0129] Finally, based on the dynamic weight value, the safety cost value, and the path length cost value, the cumulative cost function is obtained. Adding the previously obtained safety cost value to the path length cost value yields the cumulative cost function value for the current grid cell.

[0130] In some embodiments, the cumulative cost function is:

[0131]

[0132] In the formula, Let be the Euclidean distance from node n to the nearest obstacle. This represents the security cost of a node (the closer it is to an obstacle, the higher the security cost). This is a safety weighting coefficient used to adjust the priority of safety costs; These are fixed-length weighting coefficients used to adjust the priority of path length costs; This is the current search step size (i.e., raster resolution).

[0133] By combining the path length cost calculated based on grid resolution and length weight, and the safety cost calculated based on distance distribution map and safety cost mapping, a cumulative cost function containing dynamic weights, safety penalties, and path length information is constructed. This allows the algorithm to simultaneously consider the physical length of the path and the collision risk of the local environment during the search process. In areas with dense obstacles, the algorithm increases the safety priority by increasing the dynamic weight, while maintaining a low safety penalty in open areas to balance traffic efficiency, thereby generating a path that balances safety and efficiency.

[0134] Optionally, in the path planning method for the detection vehicle provided in the embodiments of this application, the method further includes: determining whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; if there is a target passable grid whose distance value is less than the safe distance threshold, then setting the target passable grid as a target state.

[0135] In some embodiments, an absolute safety distance threshold is set when calculating the cost. The safe distance threshold is the minimum safe avoidance distance determined comprehensively based on the kinematic characteristics of the probe vehicle, positioning accuracy error, control execution error, and terrain modeling error. When the actual distance between the probe vehicle and the obstacle is less than this threshold, the collision risk increases significantly, thus requiring special handling of the area. The process involves traversing all passable grids in the distance distribution map and reading the distance value corresponding to each grid. This distance value represents the Euclidean distance from the center point of the grid to the nearest obstacle boundary. The distance value corresponding to each grid is then compared with the safe distance threshold.

[0136] The system determines whether there exists a passable grid whose distance value is less than a safe distance threshold. If the distance value of a passable grid is less than the safe distance threshold, the grid is identified as a high-risk area and marked as a target state. It should be noted that a target state can be either setting the node cost to infinity and removing it from the search space, or prohibiting expansion and removing it from the search space. If the distance value of a passable grid is greater than or equal to the preset safe distance threshold, the corresponding node state is calculated normally, or it is classified as an expandable state.

[0137] In some embodiments, such as Figure 5 The search results for safety-priority paths are shown in the binary obstacle grid. It is evident that the generated paths completely avoid all impassable obstacle areas and actively move away from obstacle boundaries, reserving sufficient safety redundancy.

[0138] In some embodiments, after obtaining the path search results, the path can be optimized and smoothed, for example, as... Figure 6 The processed planned path shown is from Figure 6 As can be seen, the dynamic weighted safety-first path planning method of this application exhibits good overall performance: the path always stays away from the obstacle boundary, the collision risk is low, and the safety redundancy is sufficient; compared with the traditional fixed high-weight algorithm, this algorithm performs better in open areas ( Figure 6 The last part of the path in the upper right corner no longer generates invalid detours, the path is simpler and more efficient, and safety and efficiency are optimized in a coordinated manner.

[0139] In some embodiments, it can be achieved through, as shown in Figure 7 The schematic diagram shown realizes the path planning of the rover. It takes the digital elevation map (DEM) of the lunar surface, the starting point and the ending point of the path as input, and outputs a global safe path that takes into account both safety and traffic efficiency. The whole is composed of four core functional modules working together.

[0140] The first module is the binary obstacle model construction module. This module abandons the traditional continuous cost modeling approach of multi-parameter weighted superposition. First, based on the maximum climbing angle and obstacle crossing height of the exploration vehicle, the terrain is divided into passable areas and impassable obstacle areas. Then, taking into account the positioning error, motion control error and engineering safety margin of the exploration vehicle, the obstacle areas are morphologically expanded to generate a simplified binary obstacle raster map, thereby reducing the complexity of subsequent calculations from the root.

[0141] The second module is the Euclidean distance field (EDT) pre-computation module, which uses the fast sweep method with the best computational efficiency to calculate the Euclidean distance from each passable grid to the nearest obstacle point in batches. After the pre-computation is completed, it can realize the O(1) complexity query of the obstacle distance of any node, providing basic support for the calculation of safety cost.

[0142] The third module is the obstacle density calculation module. Centered on each passable grid, it extracts the local neighborhood space based on a predetermined fixed sensing radius R. After removing invalid obstacle nodes and high-risk nodes, it calculates the average obstacle distance of the effective nodes in the neighborhood. After normalization, it obtains the local obstacle density ρ(n) with a value range of [0,1], thereby quantifying the danger level of the local environment.

[0143] The fourth module is a dynamic weight-driven safety-first global search module, integrating two core capabilities: dynamic adaptive adjustment of safety weights and global path search. This enables real-time linkage between local environment perception and global path optimization. Based on pre-calculated Euclidean distance fields, adjusted according to local obstacle density, and prioritizing safety, this module constructs a path search system that dynamically adapts to all terrain scenarios, ultimately generating a globally optimal path that balances inherent safety and traffic efficiency. The safety-first algorithm follows the traditional A... The total cost function structure decomposes the total cost into cumulative costs. With the cost of inspiration :

[0144]

[0145] Heuristic Cost : Use the Euclidean distance from the node to the destination, satisfying A The algorithm's admissibility condition guarantees that the final path is globally optimal; cumulative cost. The innovation of the safety-first algorithm lies in its integration of safety penalties and path length penalties, which takes the following form: At the same time, an absolute safe distance threshold is set. ,when At this time, the node cost is set to infinity, and it is removed from the search space to form the first layer of security protection.

[0146] The path planning method for the detection vehicle provided in this application relies on the local obstacle density to adjust the dynamic safety weight α(n) in real time. When the distribution of obstacles in the neighborhood is dense and the collision risk increases, α(n) is automatically increased to the upper limit range, so that the safety penalty term in the cost function is reduced. It takes a dominant position, guiding the planned path to proactively avoid various obstacles and increase the distance from obstacle boundaries. Simultaneously, it employs mandatory avoidance rules based on safe distance thresholds to fundamentally prevent vehicles from entering high-risk areas. For example... Figure 8 As shown, both the red path generated by the fixed-weight algorithm and the green path generated by the dynamic-weight algorithm of this application maintain a safe distance from the obstacle, while A in the related technology The yellow path obtained by the algorithm follows the obstacle boundary closely. The planning logic of this application aligns with the safety management principles and manual control approach of lunar exploration missions, effectively mitigating safety hazards caused by rover positioning errors, motion control errors, and terrain modeling errors, and significantly reducing the operational risks of deep space exploration missions from the path planning stage.

[0147] In open areas with flat terrain and sparse obstacles, the dynamic safety weight α(n) automatically decreases to the lower limit, weakening the impact of the safety penalty term, and the algorithm takes the shortest path as the core optimization objective. Simultaneously, this scheme employs binary obstacle modeling, significantly simplifying the terrain cost calculation logic and further reducing overall computation time. Combined with... Figure 8 As can be seen from the comparison, the red path output by the traditional fixed-weight algorithm... Figure 8 In open areas such as the upper right corner, there are still obvious instances of ineffective detours; however, the green path generated by this algorithm is simple and straight, completely avoiding unnecessary detours and achieving a balance between safety and traffic efficiency. This feature can fully support applications such as real-time path planning for lunar rovers on the ground and rapid replanning in case of emergencies, perfectly adapting to the operational characteristics of deep space exploration missions with limited telemetry and control time and high requirements for command response timeliness.

[0148] Using local obstacle density as the core adjustment criterion, coupled with a tiered strategy for three scenarios—open, normal, and densely hazardous—the algorithm achieves smooth weight transitions through a non-linear formula. It can autonomously switch planning logic based on real-time terrain features without manual intervention. A single algorithm can adapt to various heterogeneous and complex terrains, including plains, impact crater areas, and rocky highlands. When facing different obstacle distribution scenarios, core parameters do not require manual readjustment, resulting in a wider range of adaptability. This reduces the workload of algorithm deployment and on-site debugging, improves the versatility and cross-platform portability of the path planning system, and facilitates its application on different models of exploration vehicles operating in different areas.

[0149] The path planning method for a probe vehicle provided in this application traverses a digital elevation map of a target area based on the probe vehicle's accessibility threshold to obtain a target obstacle grid map. The grid map includes passable and impassable grids. The distance between each passable grid and its nearest obstacle is calculated to obtain a distance distribution map. Based on the distance distribution map, the obstacle density corresponding to each passable grid is calculated. A target cost function is obtained based on the distance distribution map and the obstacle density, and a path is solved based on the target cost function to obtain the planned path for the probe vehicle in the target area. This method solves the technical problem of poor probe vehicle path planning in related technologies.

[0150] In this application, the digital elevation map of the target area is traversed based on the mobility threshold of the probe vehicle, simplifying the complex terrain into a target obstacle grid map containing both traversable and impassable grids. This addresses the problem in existing continuous cost models where the impact of terrain parameter differences on mobility safety is over-quantified, leading to planning results that do not conform to engineering intuition. Then, the distance between each traversable grid in the target obstacle grid map and the nearest obstacle point is calculated to obtain a distance distribution map. Based on this distance distribution map, the obstacle density corresponding to each traversable grid is calculated, enabling path planning to directly use the distance to the obstacle as the core safety indicator. This solves the problem in related technologies where the algorithm-generated path closely follows the obstacle boundary, resulting in insufficient safety. Finally, a target cost function is constructed based on the distance distribution map and obstacle density, and path solving is performed based on this target cost function. This effectively solves the technical problem of poor probe vehicle path planning, achieving the effect of improving the safety, adaptability, and rationality of path planning.

[0151] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0152] Example 2

[0153] This application also provides a path planning device for a detection vehicle. It should be noted that the path planning device for the detection vehicle in this application can be used to execute the path planning method for the detection vehicle provided in this application. The path planning device for the detection vehicle provided in this application is described below.

[0154] According to an embodiment of this application, an apparatus for implementing the above-described path planning method for a probe vehicle is also provided, such as... Figure 9 As shown, the device includes: a traversal unit 901, a first calculation unit 902, a second calculation unit 903, and a solution unit 904.

[0155] Traversal unit 901 is used to traverse the digital elevation map of the target area based on the accessibility threshold of the detection vehicle to obtain the target obstacle grid map, wherein the grid in the target obstacle grid map includes passable grids and impassable grids.

[0156] The first calculation unit 902 is used to calculate the distance between each passable grid cell in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map;

[0157] The second calculation unit 903 is used to calculate the obstacle density corresponding to each passable grid cell based on the distance distribution map;

[0158] Solver 904 is used to obtain the target cost function based on the distance distribution map and obstacle density, and to solve the path based on the target cost function to obtain the planned path of the probe in the target area.

[0159] The exploration vehicle path planning device provided in this application embodiment traverses a digital elevation map of the target area based on the exploration vehicle's traversal capability threshold using a traversal unit 901 to obtain a target obstacle grid map. The grid in the target obstacle grid map includes traversable and impassable grids. A first calculation unit 902 calculates the distance between each traversable grid in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map. A second calculation unit 903 calculates the obstacle density corresponding to each traversable grid based on the distance distribution map. A solution unit 904 obtains a target cost function based on the distance distribution map and the obstacle density, and performs path solving based on the target cost function to obtain the planned path of the exploration vehicle in the target area. This solves the technical problem of poor exploration vehicle path planning in related technologies.

[0160] In this application, the digital elevation map of the target area is traversed based on the mobility threshold of the probe vehicle, simplifying the complex terrain into a target obstacle grid map containing both traversable and impassable grids. This addresses the problem in existing continuous cost models where the impact of terrain parameter differences on mobility safety is over-quantified, leading to planning results that do not conform to engineering intuition. Then, the distance between each traversable grid in the target obstacle grid map and the nearest obstacle point is calculated to obtain a distance distribution map. Based on this distance distribution map, the obstacle density corresponding to each traversable grid is calculated, enabling path planning to directly use the distance to the obstacle as the core safety indicator. This solves the problem in related technologies where the algorithm-generated path closely follows the obstacle boundary, resulting in insufficient safety. Finally, a target cost function is constructed based on the distance distribution map and obstacle density, and path solving is performed based on this target cost function. This effectively solves the technical problem of poor probe vehicle path planning, achieving the effect of improving the safety, adaptability, and rationality of path planning.

[0161] Optionally, in the path planning device for the detection vehicle provided in this application embodiment, the traversal unit includes: an identification module, used to identify terrain obstacles on the digital elevation map of the target area and obtain identification results; a division module, used to traverse and divide the digital elevation map according to the traffic capacity threshold and the identification results to obtain an initial obstacle grid map; and an expansion module, used to expand the obstacle grid in the initial obstacle grid map through dilation operations to obtain a target obstacle grid map.

[0162] Optionally, in the path planning device for the detection vehicle provided in this application embodiment, the second calculation unit includes: a first acquisition module, used to acquire the target radius and obtain a neighborhood space based on the target radius with each passable grid as the center; a filtering module, used to filter the grids in the neighborhood space to obtain the filtered target grids; a second acquisition module, used to acquire the target distance value corresponding to the filtered target grid from the distance distribution map; and a first calculation module, used to calculate the target distance value to obtain the obstacle density.

[0163] Optionally, in the path planning device for the detection vehicle provided in this application embodiment, the solution unit includes: a second calculation module, used to calculate based on the obstacle density corresponding to the passable grid to obtain a dynamic weight value; a third calculation module, used to calculate based on the dynamic weight value and the distance distribution map to obtain a cumulative cost function; a fourth calculation module, used to calculate based on the distance distribution map to obtain a heuristic cost function; and a determination module, used to obtain a target cost function based on the cumulative cost function and the heuristic cost function.

[0164] Optionally, in the path planning device for the detection vehicle provided in this application embodiment, the second calculation module includes: a first acquisition submodule, used to acquire the adjustment range and adjustment coefficient of the dynamic weight; and a first calculation submodule, used to calculate the dynamic weight value based on the adjustment range, adjustment coefficient and obstacle density.

[0165] Optionally, in the path planning device for the detection vehicle provided in this application embodiment, the third calculation module includes: a second calculation submodule, used to calculate the path length cost based on grid resolution and length weight; a third calculation submodule, used to calculate the safety cost based on distance values ​​in the distance distribution map; and a first determination submodule, used to obtain the cumulative cost function based on the dynamic weight value, the safety cost, and the path length cost.

[0166] Optionally, in the path planning device for the probe vehicle provided in this application embodiment, the first acquisition module includes: a second acquisition submodule, used to acquire the maximum envelope radius of the probe vehicle, a preset safety distance threshold, and a reserved sensing margin; a second determination submodule, used to determine an upper limit constraint value and a lower limit constraint value based on the safety distance threshold; and a third determination submodule, used to obtain the target radius based on the maximum envelope radius, the safety distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0167] Optionally, in the path planning device for the detection vehicle provided in the embodiments of this application, the device further includes: a judgment unit, used to judge whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; and a setting unit, used to set the target passable grid as a target state if the distance value corresponding to the target passable grid is less than the safe distance threshold.

[0168] It should be noted that the traversal unit 901, the first calculation unit 902, the second calculation unit 903, and the solution unit 904 mentioned above correspond to steps S201 to S204 in Embodiment 1. The four units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0169] Example 3

[0170] Embodiments of this application may provide an electronic device. Figure 10 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 10 As shown, the electronic device may include: one or more ( Figure 10 (Only one is shown) processor 1002, memory 1004, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0171] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0172] The processor can access the information and application programs stored in the memory via the transmission device to execute the following steps: traversing the digital elevation map of the target area based on the vehicle's accessibility threshold to obtain a target obstacle grid map, wherein the grids in the target obstacle grid map include passable grids and impassable grids; calculating the distance between each passable grid in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map; calculating the obstacle density corresponding to each passable grid based on the distance distribution map; obtaining the target cost function based on the distance distribution map and the obstacle density, and solving the path based on the target cost function to obtain the planned path of the vehicle in the target area.

[0173] The processor can access the information and application programs stored in the memory via the transmission device to execute the following steps: traversing the digital elevation map of the target area based on the accessibility threshold of the probe vehicle to obtain a target obstacle raster map, including: identifying terrain obstacles on the digital elevation map of the target area and obtaining identification results; traversing and dividing the digital elevation map according to the accessibility threshold and the identification results to obtain an initial obstacle raster map; and expanding the obstacle raster in the initial obstacle raster map through dilation operations to obtain the target obstacle raster map.

[0174] The processor can access information and applications stored in the memory via a transmission device to perform the following steps: Calculate the obstacle density corresponding to each passable grid cell based on the distance distribution map, including: obtaining the target radius and obtaining a neighborhood space centered on each passable grid cell based on the target radius; filtering the grid cells in the neighborhood space to obtain the filtered target grid cells; obtaining the target distance value corresponding to the filtered target grid cells from the distance distribution map; and calculating the obstacle density based on the target distance value.

[0175] The processor can access the information and application programs stored in the memory via the transmission device to perform the following steps: obtaining the target cost function based on the distance distribution map and obstacle density, including: calculating the dynamic weight value based on the obstacle density corresponding to the passable grid; calculating the cumulative cost function based on the dynamic weight value and the distance distribution map; calculating the heuristic cost function based on the distance distribution map; and obtaining the target cost function based on the cumulative cost function and the heuristic cost function.

[0176] The processor can access the information and application programs stored in the memory via the transmission device to perform the following steps: calculate the dynamic weight value based on the obstacle density corresponding to the passable grid, including: obtaining the adjustment range and adjustment coefficient of the dynamic weight; and calculate the dynamic weight value based on the adjustment range, adjustment coefficient, and obstacle density.

[0177] The processor can access the information and application programs stored in the memory via the transmission device to perform the following steps: calculating the cumulative cost function based on the dynamic weight value and the distance distribution map, including: calculating the path length cost based on the grid resolution and length weight; calculating the security cost based on the distance values ​​in the distance distribution map; and obtaining the cumulative cost function based on the dynamic weight value, the security cost, and the path length cost.

[0178] The processor can call the information and application program stored in the memory through the transmission device to perform the following steps: obtaining the target radius, including: obtaining the maximum envelope radius of the probe vehicle, the preset safe distance threshold, and the reserved sensing margin; determining the upper limit constraint value and the lower limit constraint value based on the safe distance threshold; and obtaining the target radius based on the maximum envelope radius, the safe distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

[0179] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: The method further includes: determining whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; if there is a target passable grid whose distance value is less than the safe distance threshold, then setting the target passable grid to the target state.

[0180] Those skilled in the art will understand that Figure 10 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 10 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 10 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 10 The different configurations shown.

[0181] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0182] Example 4

[0183] Embodiments of this application also provide a computer-readable storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the probe path planning method provided in Embodiment 1.

[0184] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0185] This application also provides a computer program product that, when executed on a data processing device, is suitable for performing the steps of a probe path planning method.

[0186] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0187] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0188] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0189] 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 units can be selected to achieve the purpose of this embodiment according to actual needs.

[0190] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0191] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or 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 application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0192] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A path planning method for a probe vehicle, characterized in that, include: Based on the accessibility threshold of the detection vehicle, the digital elevation map of the target area is traversed to obtain the target obstacle grid map, wherein the grid in the target obstacle grid map includes passable grids and impassable grids. The distance between each passable grid cell and the nearest obstacle point in the target obstacle grid map is calculated to obtain a distance distribution map; Based on the distance distribution map, calculate the obstacle density corresponding to each passable grid cell; Based on the distance distribution map and the obstacle density, a target cost function is obtained, and a path is solved based on the target cost function to obtain the planned path of the probe vehicle in the target area.

2. The method according to claim 1, characterized in that, Based on the traffic capacity threshold of the detection vehicle, the digital elevation map of the target area is traversed to obtain a target obstacle raster map, including: Terrain obstacle identification is performed on the digital elevation map of the target area to obtain the identification results; Based on the accessibility threshold and the identification results, the digital elevation map is traversed and divided to obtain an initial obstacle grid map; The obstacle grid in the initial obstacle grid is expanded by an expansion operation to obtain the target obstacle grid.

3. The method according to claim 1, characterized in that, Based on the distance distribution map, the obstacle density corresponding to each passable grid cell is calculated, including: Obtain the target radius, and use each passable grid cell as the center to obtain the neighborhood space based on the target radius; The grids in the neighborhood space are filtered to obtain the filtered target grids; Obtain the target distance value corresponding to the filtered target grid from the distance distribution map; The obstacle density is obtained by calculating the target distance value.

4. The method according to claim 1, characterized in that, Based on the distance distribution map and the obstacle density, the target cost function is obtained, including: The dynamic weight value is calculated based on the obstacle density corresponding to the passable grid. The cumulative cost function is calculated based on the dynamic weight values ​​and the distance distribution map. The heuristic cost function is calculated based on the distance distribution map. The target cost function is obtained based on the cumulative cost function and the heuristic cost function.

5. The method according to claim 4, characterized in that, Dynamic weight values ​​are calculated based on the obstacle density corresponding to the passable grid, including: Obtain the adjustment range and adjustment coefficient of the dynamic weight; The dynamic weight value is obtained by calculating based on the adjustment range, the adjustment coefficient, and the obstacle density.

6. The method according to claim 4, characterized in that, Based on the dynamic weight values ​​and the distance distribution map, a cumulative cost function is calculated, including: The path length cost is calculated based on the grid resolution and length weight. The safety cost is calculated based on the distance values ​​in the distance distribution map. The cumulative cost function is obtained based on the dynamic weight value, the security cost value, and the path length cost value.

7. The method according to claim 3, characterized in that, To obtain the target radius, including: The maximum envelope radius of the probe vehicle, the preset safe distance threshold, and the reserved sensing margin are obtained. Based on the aforementioned safety distance threshold, the upper limit constraint value and the lower limit constraint value are determined; The target radius is obtained based on the maximum envelope radius, the safe distance threshold, the reserved sensing margin, the upper limit constraint value, and the lower limit constraint value.

8. The method according to claim 4, characterized in that, The method further includes: Determine whether the distance value corresponding to the passable grid in the distance distribution map is less than a preset safe distance threshold; If there is a target passable grid with a distance value less than the safe distance threshold, then the target passable grid is set to target status.

9. A path planning device for a detection vehicle, characterized in that, include: The traversal unit is used to traverse the digital elevation map of the target area based on the accessibility threshold of the detection vehicle to obtain the target obstacle grid map, wherein the grid in the target obstacle grid map includes passable grids and impassable grids. The first calculation unit is used to calculate the distance between each passable grid cell in the target obstacle grid map and the nearest obstacle point to obtain a distance distribution map; The second calculation unit is used to calculate the obstacle density corresponding to each passable grid cell based on the distance distribution map. The solution unit is used to obtain the target cost function based on the distance distribution map and the obstacle density, and to perform path solving based on the target cost function to obtain the planned path of the probe vehicle in the target area.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device containing the computer-readable storage medium to perform the exploration vehicle path planning method according to any one of claims 1 to 8.

11. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program executes the probe path planning method according to any one of claims 1 to 8 when it runs.

Images