An unmanned aerial vehicle path optimization method based on urban wind field environment rasterization

Abstract

This invention relates to the field of UAV path planning technology, specifically a UAV path optimization method based on urban wind field environmental rasterization, comprising the following steps: acquiring building geographic information data and meteorological wind rose diagram data of the target airspace; determining the dominant wind field conditions of the target airspace based on the meteorological wind rose diagram data, and using computational fluid dynamics technology to perform simulation to obtain a three-dimensional wind field distribution dataset of the airspace; and rasterizing the target airspace into a discrete uniform cubic raster model. This invention establishes a high-precision urban wind field model using GIS and CFD technologies, and discretizes it into an environmental raster with wind vector information. It introduces a wind field perception cost function, a dynamic step size mechanism, and straight-line reachability optimization into the path planning, solving the problems of poor adaptability, low search efficiency, and insufficient path smoothness of the traditional A algorithm in complex urban wind fields. This significantly improves flight efficiency while enhancing safety.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) path planning technology, specifically to a UAV path optimization method based on urban wind field environmental gridding. Background Technology

[0002] Unmanned aerial vehicles (UAVs), with their high degree of autonomy and exceptional flexibility, have become a hot research topic in addressing the challenges of transportation and delivery in densely built-up urban low-altitude areas. Wind fields, as a key source of disturbance in the urban low-altitude environment, cause complex airflow phenomena such as wind shear and strong crosswinds, posing a serious threat to the flight stability and controllability of UAVs. Therefore, how to construct safe flight paths adapted to the complex low-altitude urban environment and ensure the efficient and complete execution of transportation missions has become a critical issue that urgently needs to be addressed.

[0003] but:

[0004] Traditional path planning methods are mostly based on static environments or two-dimensional models, which do not fully consider the dynamic impact of wind fields on the ground speed and flight stability of UAVs, resulting in poor path safety, low efficiency and many inflection points.

[0005] Although existing improved algorithms incorporate wind field perception, they still suffer from problems such as fixed search step size, insufficient path smoothness, and high computational complexity.

[0006] Therefore, there is an urgent need for a UAV path planning method that can integrate high-precision wind field modeling, dynamic step size adjustment, and path smoothing optimization. Summary of the Invention

[0007] The purpose of this invention is to provide a method for optimizing UAV paths based on urban wind field environmental gridding, so as to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing UAV paths based on urban wind field environmental gridding, comprising the following steps:

[0009] S1. Environmental Data Acquisition: Obtain building geographic information data and meteorological wind rose diagram data for the target airspace;

[0010] S2. Three-dimensional wind field modeling: Based on the meteorological wind rose diagram data, determine the dominant wind field conditions of the target airspace, construct a three-dimensional digital building model based on the building geographic information data, and combine the determined dominant wind field with computational fluid dynamics technology to perform simulation and obtain a three-dimensional wind field distribution dataset of the airspace.

[0011] S3. Spatial Rasterization: The target spatial domain is rasterized into a discrete uniform cubic raster model, and the wind vector information in the three-dimensional wind field distribution dataset is mapped to each raster.

[0012] S4. Algorithm Model Construction: Establishing an improved A algorithm model with the objective of minimizing total flight time. Path planning model, the improved A The total time cost function of the path planning model integrates wind field perception UAV ground speed calculation, dynamic step size mechanism and straight-line reachability determination.

[0013] S5. Path Planning and Output: Based on the preset starting point and target point, utilize the improved A... The path planning model searches within the grid model and outputs the optimal UAV flight path.

[0014] Preferably, the rasterization process in step S3 specifically includes:

[0015] The grid granularity is Based on this granularity, the spatial domain is divided into X, Y, and Z dimensions respectively. A three-dimensional grid;

[0016] in , This is the floor function;

[0017] The path planning starting point grid is set to The target point grid is set to The path passes through a grid set to ,in ;

[0018] Each 3D grid contains the coordinates of the grid's center point. and the wind speed vector corresponding to that point .

[0019] Preferably, the total time cost function in step S4 Including cumulative costs and the cost of inspiration Total time cost function The expression is:

[0020] ;

[0021] Among them, cumulative cost The total time from the starting point to the current node is expressed as:

[0022] ;

[0023] Heuristic Cost The total time from the current node to the destination is expressed as:

[0024] ;

[0025] in, Indicates drones from the grid Fly to the grid The distance;

[0026] Indicates that the drone is from the grid Fly to the grid The magnitude of the ground speed;

[0027] This represents the expected ground speed from the current grid to the destination. It is the theoretical ground speed amplitude calculated by the UAV in the path segment from the current grid to the destination, based on the wind field conditions of the grid and the constant airspeed constraint of the UAV.

[0028] Preferably, in step S4, the wind field sensing ground speed calculation is based on the vector superposition relationship between airspeed, wind speed, and ground speed, specifically:

[0029] drone ground speed Equal to its airspeed wind speed of the grid The vector sum, i.e. ;

[0030] air speed The amplitude remains constant, and the direction is determined by the ground speed and the wind speed direction. The direction always points to the destination;

[0031] In path segment calculation, based on the constraint that the ground velocity direction points towards the target point, and combined with the wind speed vector, the formula is used. Solve The actual wind field sensing drone ground speed, of which For the first Airspeed within the path segment With ground speed The included angle, For the first Wind speed within the path segment With ground speed The included angle, and Based on the vector composition rule, the method is determined by the constraints of ground speed pointing to the target point, constant air speed amplitude, and measured wind speed vector.

[0032] Preferably, the expression for the dynamic step size mechanism in step S4 is:

[0033] ;

[0034] ;

[0035] in The dynamic step size coefficient represents the distance from the current node to the destination. Distance from start to finish The ratio;

[0036] The initial step size, The distance between the current node and the nearest restricted flight area. This indicates a small offset to prevent a step size of zero.

[0037] Preferably, the straight-line reachability determination in step S4 includes:

[0038] When expanding the current node, backtrack to its grandparent node and check whether the straight line from the grandparent node to the current node passes through any obstacle grid or no-fly zone grid.

[0039] If the straight path is unobstructed, the grandparent node and the current node are directly connected in the path, replacing the original multi-segment broken path, thereby reducing the number of turning points in the path.

[0040] Also includes:

[0041] S6. Online replanning of unsteady wind field: When the UAV detects that the environmental wind field meets any of the following quantitative conditions through sensors or updated data during flight, it is determined that a significant change has occurred and the path replanning process is automatically triggered.

[0042] (1) The measured wind speed at the location of the UAV deviates from the wind speed amplitude of the corresponding grid of the pre-planned path by more than 20% of the preset threshold;

[0043] (2) The wind direction deviation between the measured wind direction at the location of the UAV and the wind direction of the grid corresponding to the pre-planned path exceeds the preset angle threshold of 30°;

[0044] (3) The updated wind field data shows that there are cases in the pre-planned remaining paths that cross the preset flight risk zone;

[0045] Based on the latest wind field grid data, the total time cost function is updated in real time and the optimized path from the current location to the target point is searched again.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention enables UAVs to effectively avoid high-wind-speed risk areas by refining the model of heterogeneous wind fields in cities, thus ensuring flight safety. At the same time, by introducing ground speed perception and dynamic step size mechanisms, the search efficiency is optimized. Experimental results show that this method can reduce the average flight time by about 23.6% and the number of path inflection points by about 58.2%, significantly improving path smoothness and overall flight efficiency, and providing an effective path planning solution for low-altitude traffic management in complex urban environments. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the wind rose in the target area of ​​the UAV path optimization method based on urban wind field environmental gridding according to the present invention.

[0049] Figure 2 This is a schematic diagram of an airspace grid model for a UAV path optimization method based on urban wind field environmental gridding according to the present invention.

[0050] Figure 3 This invention provides an improvement to the UAV path optimization method based on urban wind field environmental gridding. Algorithm flow diagram;

[0051] Figure 4 This is a schematic diagram of a 3D building model of the target area for a UAV path optimization method based on urban wind field environmental rasterization according to the present invention.

[0052] Figure 5 for Figure 4 A schematic diagram of the wind field simulation results in the region;

[0053] Figure 6 This is a schematic diagram of the improved path planning algorithm results of the UAV path optimization method based on urban wind field environmental gridding according to the present invention;

[0054] Figure 7 This is a schematic diagram of the traditional path planning algorithm result of the UAV path optimization method based on urban wind field environmental gridding according to the present invention;

[0055] Figure 8 A schematic diagram of the path planning algorithm results that has wind field awareness but does not consider dynamic step size and straight-line reachability;

[0056] Figure 9 This is a schematic diagram of multi-scenario path results for a UAV path optimization method based on urban wind field environmental gridding according to the present invention;

[0057] Figure 10 This is a schematic diagram of the unsteady wind field path planning result of the UAV path optimization method based on urban wind field environmental gridding according to the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1-10 This invention provides a technical solution: a method for optimizing UAV paths based on urban wind field environmental gridding, comprising the following steps:

[0060] S1. Environmental Data Acquisition: Obtain building geographic information (GIS) data and meteorological wind rose data for the target airspace.

[0061] S2. Three-dimensional wind field modeling: Based on meteorological wind rose diagram data, determine the dominant wind field conditions of the target airspace, construct a three-dimensional digital building model based on building geographic information data, and combine the determined dominant wind field with computational fluid dynamics (CFD) technology to simulate and obtain a three-dimensional wind field distribution dataset of the airspace.

[0062] The construction of the 3D building digital model includes acquiring satellite maps and elevation models with a resolution of no less than 0.5 meters, restoring the polyhedral boundaries and height attributes of the building, and controlling the error of detail restoration within ±1 meter.

[0063] S3. Spatial Rasterization: The target spatial domain is rasterized into a discrete uniform cubic raster model, and the wind vector information in the three-dimensional wind field distribution dataset is mapped to the center point of each raster.

[0064] Rasterization specifically includes:

[0065] The grid granularity is Based on this granularity, the spatial domain is divided into X, Y, and Z dimensions respectively. A three-dimensional grid, grid granularity size Factors such as drone attributes, transportation mission type, and CFD mesh refinement should be considered.

[0066] in , This is the floor function;

[0067] The path planning starting point grid is set to The target point grid is set to The path passes through a grid set to ,in .

[0068] Each 3D grid contains the coordinates of the grid's center point. and the wind speed vector corresponding to that point When a spatial grid contains wind speed information from multiple CFD grids, the average wind speed information from the CFD grids is taken as the wind speed information for the grid. This process significantly impacts the calculation of ground speed and the setting of no-fly zones during subsequent path planning. It effectively transforms raw, high-precision wind field data into a structured grid space, providing quantifiable airspace environmental data support for subsequent path optimization.

[0069] S4. Algorithm Model Construction: Establishing an improved A algorithm model with the objective of minimizing total flight time. Path planning model, improved A The total time cost function of the path planning model integrates wind field perception UAV ground speed calculation, dynamic step size mechanism and straight-line reachability determination. This integrated strategy enables the improved path planning algorithm to reduce flight search time and the number of path inflection points while ensuring UAV flight safety, and optimize the smoothness of the path.

[0070] Total Time Cost Function Including cumulative costs and the cost of inspiration Total time cost function The expression is:

[0071] ;

[0072] Among them, cumulative cost The total time from the starting point to the current node is expressed as:

[0073] ;

[0074] Heuristic Cost The total time from the current node to the destination is expressed as:

[0075] ;

[0076] in, Indicates that the drone is from the grid Fly to the grid The distance;

[0077] Indicates that the drone is from the grid Fly to the grid The magnitude of the ground speed;

[0078] This represents the expected ground speed from the current grid to the destination. It is the theoretical ground speed amplitude calculated by the UAV in the path segment from the current grid to the destination, based on the wind field conditions of the grid and the constant airspeed constraint of the UAV.

[0079] The ground speed calculation of the wind field sensing UAV is based on the vector superposition relationship between airspeed, wind speed, and ground speed, where airspeed... This indicates the speed of the drone relative to the air, wind speed. Ground speed is a measure of the speed of wind relative to the ground. The ground speed calculation for a wind-sensing drone reflects its actual speed relative to the ground as follows:

[0080] Using the parallelogram composition rule, airspeed With wind speed The two adjacent sides that form a parallelogram, and the ground speed This forms a diagonal line, therefore the drone's ground speed Equal to its airspeed wind speed of the grid The vector sum, i.e. ;

[0081] air speed The amplitude remains constant, and the direction is determined by the ground speed and the wind speed direction. The direction always points to the destination, by , can be obtained ;

[0082] In path segment calculation, based on ground speed The constraint of direction pointing towards the target point, combined with wind speed Vector, through formula Solve The actual wind field sensing drone ground speed, of which For the first Airspeed within the path segment With ground speed The included angle, For the first Wind speed within the path segment With ground speed The included angle, and Based on the vector composition rule, the method is determined by the constraints of ground speed pointing to the target point, constant air speed amplitude, and measured wind speed vector.

[0083] Dynamic step size is the engine of node generation, which controls the granularity of exploration. Choosing an appropriate step size according to the situation can speed up the search and improve search efficiency. The distance of the dynamic step size is related to the distance from the current node to the destination and the distance from the current node to the nearest obstacle. The expression of the dynamic step size mechanism is:

[0084] ;

[0085] ;

[0086] in The dynamic step size coefficient represents the distance from the current node to the destination. Distance from start to finish The ratio;

[0087] The initial step size, The distance between the current node and the nearest restricted flight area. This indicates a small offset to prevent a step size of zero.

[0088] Straight-line reachability determination includes:

[0089] When expanding the current node, backtrack to its grandparent node and check whether the straight line from the grandparent node to the current node passes through any obstacle grid or no-fly zone grid.

[0090] If the straight path is unobstructed, the grandparent node and the current node are directly connected in the path, replacing the original multi-segment broken path, thereby reducing the number of turning points in the path.

[0091] S5. Path Planning and Output: Based on the preset starting point and target point, utilize the improved A... The path planning model searches within the grid model and outputs the optimal UAV flight path.

[0092] S6. Online replanning of unsteady wind field: When the UAV detects that the environmental wind field meets any of the following quantitative conditions through sensors or updated data during flight, it is determined that a significant change has occurred and the path replanning process is automatically triggered.

[0093] (1) The measured wind speed at the location of the UAV deviates from the wind speed amplitude of the corresponding grid of the pre-planned path by more than 20% of the preset threshold;

[0094] (2) The wind direction deviation between the measured wind direction at the location of the UAV and the wind direction of the grid corresponding to the pre-planned path exceeds the preset angle threshold of 30°;

[0095] (3) The updated wind field data shows that there are cases in the pre-planned remaining paths that cross the preset flight risk zone;

[0096] Based on the latest wind field grid data, the total time cost function is updated in real time and the optimized path from the current location to the target point is searched again.

[0097] Example 1

[0098] This embodiment uses a typical densely built-up urban area in Nanshan District, Shenzhen as the experimental scenario. The area is approximately 1km × 0.9km, and the airspace height is set at 150m. This embodiment elaborates on the technical solution of the present invention.

[0099] First, based on the clearly defined building layout, geometry, and height information, a 3D building model that closely matches the real-world scene was constructed, such as... Figure 4As shown in the figure, this model defines the physical boundaries of the wind field simulation, ensuring the realism and reliability of the simulation.

[0100] Analysis of the wind rose diagram of Shenzhen generated from meteorological observation data reveals that the prevailing wind in the study area is mainly northerly, with wind speeds primarily at level 3 (4-6 m / s), reaching level 4 (6-8 m / s) in some wind directions. Instantaneous gusts in the study area can reach level 5 (8 m / s) at certain times. Based on the aforementioned 3D building model, CFD wind field simulations were conducted with prevailing wind speeds of 4 m / s, 6 m / s, and 8 m / s, respectively. The simulation results are shown in the figure below. Figure 5 As shown, the three-dimensional wind field distribution characteristics around urban building clusters under three dominant wind speeds are intuitively presented, clearly revealing the coupling effect between wind field and building layout, dominant wind conditions and wind speed gradient change law.

[0101] To verify the algorithm's performance, three sets of comparative experiments were designed. All experiments were conducted in a unified CFD wind field simulation environment with a prevailing wind speed of 8 m / s, ensuring that the experimental variables were limited to the algorithm itself, and that the comparison results were rigorous and valid. The first set compared the improved algorithm with the traditional A... Algorithms were compared, with traditional A... The algorithm uses Euclidean distance as the cost function and ignores wind field influence. The second group compares the improved algorithm with a path planning algorithm that incorporates wind field perception but does not comprehensively consider dynamic step size and straight-line reachability constraints. The third group consists of sub-scene path experiments, verifying the algorithm's performance stability and generalization ability across multiple sub-scenes in the same area, considering different building layouts in the left and right zones. All experiments use matched start and end points and identical UAV parameters in the corresponding scenarios. Evaluation metrics include flight time, flight distance, number of inflection points, and distance traversed through flight risk zones. The results of the improved algorithm are as follows: Figure 6 As shown, the trajectory is smooth overall and crosses no risk zones, fully demonstrating the algorithm's adaptability to wind field environments.

[0102] The results of the first set of comparative experiments are as follows: Figure 7 As shown in Table 1, the path planning results indicators are as follows. Analysis reveals that traditional path planning algorithms, by ignoring wind field influence, enter risk zones, significantly reducing flight safety. Meanwhile, the improved algorithm in this paper, which comprehensively considers wind field perception, dynamic step size, and straight-line reachability, reduces flight time by approximately 42.2% and the number of inflection points by approximately 48.5% while avoiding risk zones.

[0103] Table 1. Path Planning Result Indicators

[0104]

[0105] To systematically verify the comprehensive performance of the improved algorithm proposed in this invention, the second set of experiments compared it with a benchmark algorithm that also possesses wind field perception capabilities for risk avoidance but does not integrate dynamic step size strategy and straight-line reachability detection optimization. The results of the second set of comparative experiments are as follows: Figure 8 As shown in Table 2, the path planning results indicators are as follows. The results indicate that while both algorithms effectively mitigate wind field risks, the improved algorithm reduces the average flight time by approximately 23.6% and the average number of path inflection points by 58.2%, demonstrating a significant advantage in key performance aspects. This not only greatly improves navigation efficiency but also enhances the stability and feasibility of flight control through a significant improvement in path smoothness.

[0106] Table 2 Path Planning Result Indicators

[0107]

[0108] The results of the third group of multi-scenario experiments are as follows: Figure 9 As shown, both the left and right sub-scene paths were planned and generated under a unified wind field environment of 8 m / s, with no risk zone crossings recorded throughout, and both exhibited good trajectory smoothness. Path performance indicators are shown in Table 3. Considering the scene differences of dense buildings and complex street layouts in the right area versus sparse buildings and regular street layouts in the left area, it can be seen that the algorithm maintains stable and excellent performance in sub-scenes with different building layouts, unaffected by subtle differences in local environmental features, fully verifying the algorithm's generalization ability and stability in multiple scenarios.

[0109] Table 3 Path Planning Result Indicators

[0110]

[0111] To address the need for verifying path robustness and replanning mechanisms in response to the characteristics of unsteady wind fields, a path planning experiment was designed that was started only once throughout the entire process. Three sets of gradient steady-state wind fields were used to simulate the continuous dynamic changes in wind speed in a real environment. The section from the starting point to point A used an 8 m / s steady-state wind field; the section from point A to point B switched to a 6 m / s steady-state wind field; and the section from point B to the end point switched to a 4 m / s steady-state wind field. Each wind field was independently simulated using CFD technology, and the wind field boundaries were precisely connected through nodes A and B. The aforementioned gradient wind fields were used as the boundary conditions for the CFD simulation. The path planning algorithm ran continuously throughout the entire process without interruption or restart. When the UAV flew to nodes A and B and sensed a change in the wind field environment, a replanning mechanism was automatically triggered to adjust the path planning strategy in real time to adapt to the new wind field conditions.

[0112] Path planning results under unsteady wind fields are as follows Figure 10As shown, the entire path crosses no risk zones, and the three path segments smoothly connect at nodes A and B, without significant trajectory abrupt changes or redundant inflection points. At node A, the algorithm detects a change in the wind field environment from a simulated environment with a dominant wind speed of 8 m / s to one with a dominant wind speed of 6 m / s, automatically initiating a replanning process. Based on the new wind field grid data, the cost function is recalculated, reducing the number of node expansions to improve flight efficiency. At node B, the grid wind field data is detected to have decreased to a dominant wind speed of 4 m / s, triggering replanning again to optimize path smoothness. This result fully demonstrates that the improved algorithm can accurately adapt to wind field changes through replanning in complex scenarios with continuously and dynamically changing wind speeds. Therefore, the algorithm of this invention can perceive wind field changes in real time and automatically trigger replanning, dynamically adjusting the path strategy to adapt to new wind field conditions. This overcomes the limitation of traditional algorithms that are "only applicable to steady-state wind fields," ensuring a safe, efficient, and smooth flight state even in real dynamic wind field environments.

[0113] In this embodiment, through precise characterization of three-dimensional wind field features and targeted improvements to the path algorithm, multi-objective collaborative optimization of the safety, efficiency, and path quality of UAVs in complex urban environments is achieved.

[0114] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0115] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing a path of an unmanned aerial vehicle based on urban wind field environment rasterization, characterized in that: Includes the following steps: S1. Environmental Data Acquisition: Obtain building geographic information data and meteorological wind rose diagram data for the target airspace; S2. Three-dimensional wind field modeling: Based on the meteorological wind rose diagram data, determine the dominant wind field conditions of the target airspace, construct a three-dimensional digital building model based on the building geographic information data, and combine the determined dominant wind field with computational fluid dynamics technology to perform simulation and obtain a three-dimensional wind field distribution dataset of the airspace. S3. Spatial Rasterization: The target spatial domain is rasterized into a discrete uniform cubic raster model, and the wind vector information in the three-dimensional wind field distribution dataset is mapped to each raster. S4. Algorithm Model Construction: Establishing an improved A algorithm model with the objective of minimizing total flight time. Path planning model, the improved A The total time cost function of the path planning model integrates wind field perception UAV ground speed calculation, dynamic step size mechanism and straight-line reachability determination. S5. Path Planning and Output: Based on the preset starting point and target point, utilize the improved A... The path planning model searches within the grid model and outputs the optimal UAV flight path.

2. The UAV path optimization method based on urban wind field environmental gridding according to claim 1, characterized in that: The rasterization process in step S3 specifically includes: The grid granularity is Based on this granularity, the spatial domain is divided into X, Y, and Z dimensions respectively. A three-dimensional grid; in , This is the floor function; The total number of nodes in the path planning is defined as follows: The starting grid is set to The target point grid is set to The path passes through a grid set to ,in ; Each 3D grid contains the coordinates of the grid's center point. and the wind speed vector corresponding to that point .

3. The UAV path optimization method based on urban wind field environmental gridding according to claim 2, characterized in that: The total time cost function in step S4 Including cumulative costs and the cost of inspiration Total time cost function The expression is: ； Among them, cumulative cost The total time from the starting point to the current node is expressed as: ； Heuristic Cost The total time from the current node to the destination is expressed as: ； in, Indicates that the drone is from the grid Fly to the grid The distance; Indicates that the drone is from the grid Fly to the grid The magnitude of the ground speed; This represents the expected ground speed from the current grid to the destination. It is the theoretical ground speed amplitude calculated by the UAV in the path segment from the current grid to the destination, based on the wind field conditions of the grid and the constant airspeed constraint of the UAV.

4. The UAV path optimization method based on urban wind field environmental gridding according to claim 3, characterized in that: In step S4, the calculation of ground speed based on the wind field sensing is based on the vector superposition relationship between airspeed, wind speed, and ground speed, specifically: drone ground speed equal to its airspeed wind speed of the grid The vector sum, i.e. ; air speed The amplitude remains constant, and the direction is determined by the ground speed and the wind speed direction. The direction always points to the destination; In path segment calculation, based on the constraint that the ground velocity direction points towards the target point, and combined with the wind speed vector, the formula is used. Solve The actual wind field sensing drone ground speed, of which For the first Airspeed within the path segment With ground speed The included angle, For the first Wind speed within the path segment With ground speed The included angle, and Based on the vector composition rule, the method is determined by the constraints of ground speed pointing to the target point, constant air speed amplitude, and measured wind speed vector.

5. The UAV path optimization method based on urban wind field environmental gridding according to claim 4, characterized in that: The expression for the dynamic step size mechanism in step S4 is: ； ； in The dynamic step size coefficient represents the distance from the current node to the destination. Distance from start to finish The ratio; The initial step size, The distance between the current node and the nearest restricted flight area. This indicates a small offset to prevent a step size of zero.

6. The UAV path optimization method based on urban wind field environmental gridding according to claim 5, characterized in that: The straight-line reachability determination in step S4 includes: When expanding the current node, backtrack to its grandparent node and check whether the straight line from the grandparent node to the current node passes through any obstacle grid or no-fly zone grid. If the straight path is unobstructed, the grandparent node and the current node are directly connected in the path, replacing the original multi-segment broken path, thereby reducing the number of turning points in the path.

7. The UAV path optimization method based on urban wind field environmental gridding according to claim 6, characterized in that: Also includes: S6. Online replanning of unsteady wind field: When the UAV detects that the environmental wind field meets any of the following quantitative conditions through sensors or updated data during flight, it is determined that a significant change has occurred and the path replanning process is automatically triggered. (1) The measured wind speed at the location of the UAV deviates from the wind speed amplitude of the corresponding grid of the pre-planned path by more than 20% of the preset threshold; (2) The wind direction deviation between the measured wind direction at the location of the UAV and the wind direction of the grid corresponding to the pre-planned path exceeds the preset angle threshold of 30°; (3) The updated wind field data shows that there are cases in the pre-planned remaining paths that cross the preset flight risk zone; Based on the latest wind field grid data, the total time cost function is updated in real time and the optimized path from the current location to the target point is searched again.

Images