Method and system for intelligent planning and risk assessment of unmanned aerial vehicle route based on large model
By constructing a database of flight path influencing factors and a large visual model, and updating UAV flight paths in real time, the problems of low efficiency and poor reliability in UAV flight path planning have been solved, and efficient and reliable flight path adjustment has been achieved.
Patent Information
- Application Number
- CN202610843582.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-11
- Publication Date
- 2026-08-25
AI Technical Summary
Existing drone flight path planning is inefficient and has poor flight reliability, making it impossible to adjust flight trajectories in real time to cope with disturbances.
The intelligent flight path planning method for UAVs based on large models constructs a feature database of flight path influencing factors, uses a large visual model to update the flight trajectory in real time, acquires and processes real-time images and disturbance factors, and dynamically adjusts the flight path.
It improves the efficiency of flight route planning and flight reliability, and enables the UAV to fly autonomously and stably and respond to obstacles in real time.
Smart Images
Figure CN122631084A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of unmanned aerial vehicle (UAV) route planning technology, and specifically to a method and system for intelligent UAV route planning and risk assessment based on a large model. Background Technology
[0002] Drone flight path planning was developed under the combined forces of explosive market demand, national strategic support, and technological bottlenecks. It is not only related to the flight safety of individual drones, but also a fundamental technology supporting the healthy, orderly, and high-quality development of the trillion-dollar low-altitude economy industry.
[0003] In existing technologies, the planning of drone flight paths generally involves observing the flight environment in advance and constructing corresponding flight paths based on the actual flight environment. However, this method requires a large amount of environmental data collection and processing, resulting in low efficiency and high cost in flight path planning. Furthermore, it cannot adjust the flight trajectory in real time to address disturbances during flight, leading to poor flight performance and reliability.
[0004] In the process of realizing this invention, the inventors of this application discovered that the above-mentioned solutions in the prior art have the defects of low route planning efficiency and poor flight reliability. Summary of the Invention
[0005] The purpose of this invention is to provide a method and system for intelligent planning and risk assessment of UAV routes based on a large model. This method and system for intelligent planning and risk assessment of UAV routes based on a large model has the functions of improving route planning efficiency and ensuring flight reliability.
[0006] To achieve the above objectives, embodiments of the present invention provide a method for intelligent unmanned aerial vehicle (UAV) route planning and risk assessment based on a large model, comprising:
[0007] Acquire data on flight path influencing factors during UAV flight and construct a database of flight path influencing factor characteristics;
[0008] A large visual model is selected, and the visual model is fine-tuned using the aforementioned route influencing factor feature database;
[0009] Obtain the coordinates of the target point during the flight of the UAV;
[0010] Obtain the real-time images and coordinates captured by the drone.
[0011] The current flight trajectory is obtained based on the real-time image, the fine-tuned visual model, the real-time coordinates, and the target point coordinates.
[0012] Obtain the real-time disturbance factors of the currently described UAV;
[0013] A dynamic risk value is obtained based on the real-time disturbance factors of the drone and its flight trajectory.
[0014] The flight trajectory of the UAV is updated based on the dynamic risk value.
[0015] Optionally, the route influencing factor data includes: static obstacle data and dynamic obstacle data.
[0016] Optionally, constructing a database of characteristics of factors influencing flight routes includes:
[0017] The data on factors influencing the flight routes are preprocessed, and feature vectors are extracted.
[0018] Obtain the label corresponding to the feature vector;
[0019] Pair the feature vectors with the corresponding labels and construct an influencing factor feature dataset;
[0020] Construct a feature database of route influencing factors based on the aforementioned feature set of influencing factors.
[0021] Optionally, obtaining the current flight trajectory based on the real-time image, the fine-tuned visual model, the real-time coordinates, and the target point coordinates includes:
[0022] The flight trajectory is constructed based on the real-time coordinates of the UAV and the coordinates of the target point;
[0023] Drive the drone to fly along the flight path;
[0024] The real-time image is input into the fine-tuned visual model to determine whether there are any flight path influencing factors in front of the current drone.
[0025] If a flight path influencing factor is detected ahead of the current drone, the current flight path is updated according to the flight path influencing factor.
[0026] Optionally, updating the current flight path based on the aforementioned route influencing factors includes:
[0027] Determine whether the factors affecting the flight path are static obstacles;
[0028] If the influencing factor on the flight path is determined to be a static obstacle, the edge coordinates of the static obstacle are obtained, wherein the edge coordinates include the top edge coordinates, the left edge coordinates, and the right edge coordinates;
[0029] The relative distance between the UAV and the static obstacle is obtained according to formula (1).
[0030] (1)
[0031] in, The distance between the top edge of the static obstacle and the drone is the relative height distance. The z-axis coordinate of the top edge of the static obstacle is given. The altitude coordinates of the currently described UAV are: The coordinates of the left edge of the static obstacle are the relative horizontal distance between the drone and the current obstacle. The x-coordinate of the left edge of the currently described static obstacle. Let x be the x-axis coordinate of the currently described UAV. The coordinates of the right edge of the obstacle are the relative horizontal distance between the drone and the obstacle. The x-coordinate of the right edge of the currently described static obstacle;
[0032] Select the shortest relative distance between the current UAV and the static obstacle;
[0033] The flight trajectory of the drone is updated based on the shortest relative distance between the drone and the static obstacle.
[0034] Optionally, updating the drone's flight trajectory based on the shortest relative distance between the drone and the static obstacle includes:
[0035] Obtain the preset safety threshold of the static obstacle;
[0036] Determine whether the shortest relative distance between the current drone and the static obstacle is less than or equal to the preset safety threshold;
[0037] If the shortest relative distance between the current drone and the static obstacle is less than or equal to the preset safety threshold, then determine whether the straight-line distance between the current drone and the static obstacle is less than or equal to the preset safety threshold.
[0038] If the current straight-line distance between the drone and the static obstacle is less than or equal to the preset safety threshold, the drone is driven to fly along the direction outside the edge corresponding to the shortest relative distance.
[0039] Determine whether the drone has crossed the static obstacle;
[0040] If it is determined that the UAV has crossed the static obstacle, return to the step of constructing the flight trajectory based on the real-time coordinates of the UAV and the coordinates of the target point;
[0041] If it is determined that the drone has not crossed the static obstacle, the drone is driven to fly in a direction parallel to the static obstacle;
[0042] If it is determined that the straight-line distance between the current drone and the static obstacle is greater than the preset safety threshold, the drone is driven to fly along the current flight trajectory.
[0043] If the shortest relative distance between the drone and the static obstacle is determined to be greater than the preset safety threshold, the drone is driven to fly along the current flight path.
[0044] Optionally, updating the current flight path based on the aforementioned route influencing factors further includes:
[0045] If the factors affecting the flight path are determined not to be static obstacles, then the factors affecting the flight path are determined to be dynamic obstacles.
[0046] Acquire two sets of real-time images of dynamic obstacles taken by the drone at preset time intervals;
[0047] Two sets of dynamic coordinates of the dynamic obstacle are obtained based on two sets of real-time images of the dynamic obstacle.
[0048] The direction of movement of the dynamic obstacle is obtained according to formula (2).
[0049] (2)
[0050] in, Let be the unit vector representing the direction of movement of the dynamic obstacle. Let V be the velocity vector of the dynamic obstacle.
[0051] The drone is driven to avoid the moving obstacle based on its direction of movement.
[0052] Determine whether the drone has crossed the dynamic obstacle;
[0053] If it is determined that the UAV has crossed the dynamic obstacle, return to the step of constructing the flight trajectory based on the real-time coordinates of the UAV and the coordinates of the target point;
[0054] If it is determined that the drone has not crossed the dynamic obstacle, the drone is driven to fly along a direction of movement parallel to the dynamic obstacle.
[0055] Optionally, obtaining a dynamic risk value based on the real-time disturbance factors of the UAV and its flight trajectory includes:
[0056] The meteorological disturbance factor of the UAV is obtained according to formula (3).
[0057] (3)
[0058] in, The weather disturbance factor for the currently described UAV, Given the current wind speed in the flight environment, For safe flight environment wind speed;
[0059] The communication disturbance factor of the UAV is obtained according to formula (4).
[0060] (3)
[0061] in, The communication disturbance factor of the currently described UAV, This refers to the current communication distance between the drone and the controller. This refers to the safe communication distance between the UAV and the controller;
[0062] The dynamic risk value of the current UAV flight is obtained according to formula (5).
[0063] (5)
[0064] in, This represents the dynamic risk value of the current drone flight. , These are meteorological weights and communication weights, respectively.
[0065] Optionally, updating the current flight trajectory of the UAV based on the dynamic risk value includes:
[0066] Determine whether the dynamic risk value of the current drone flight is greater than or equal to the risk threshold;
[0067] If the dynamic risk value of the drone is determined to be greater than or equal to the risk threshold, the drone is driven to hover and an alarm is issued.
[0068] If the dynamic risk value of the current drone flight is determined to be less than the risk threshold, the position coordinates of the drone are checked in real time and the current flight trajectory of the drone is updated.
[0069] On the other hand, the present invention also provides a large-scale model-based intelligent flight path planning and risk assessment system for unmanned aerial vehicles (UAVs), comprising:
[0070] The drone body is equipped with a camera for capturing real-time images of the drone body during flight.
[0071] The controller, connected to the UAV body, is used to execute any of the above-described UAV route intelligent planning and risk assessment methods.
[0072] Through the above technical solution, the UAV route intelligent planning and risk assessment method and system based on a large model provided by this invention collects data on route influencing factors during UAV flight, thereby constructing a corresponding route influencing factor feature database for fine-tuning the visual large model; then, based on the target point coordinates of the UAV flight and the UAV's real-time images and real-time coordinates, it obtains the real-time flight trajectory, acquires the real-time disturbance factors during the UAV flight, and obtains the dynamic risk value based on the real-time disturbance factors. The flight trajectory is then updated based on this dynamic risk value, thus ensuring the autonomous and stable flight of the UAV. By using real-time images and real-time disturbance factors to update the UAV flight trajectory in real time, the dynamic updating of the UAV flight trajectory can be effectively improved, resulting in higher trajectory planning efficiency and higher UAV flight reliability.
[0073] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description
[0074] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:
[0075] Figure 1 This is a flowchart of a method for intelligent planning and risk assessment of unmanned aerial vehicle routes based on a large model, according to one embodiment of the present invention.
[0076] Figure 2 This is a flowchart illustrating the construction of a feature database of flight route influencing factors in a large-model-based intelligent planning and risk assessment method for unmanned aerial vehicle (UAV) routes, according to one embodiment of the present invention.
[0077] Figure 3 This is a flowchart of updating the current flight trajectory in a large-model-based intelligent planning and risk assessment method for unmanned aerial vehicle routes according to an embodiment of the present invention;
[0078] Figure 4 This is a flowchart illustrating how a drone avoids static obstacles in a large-model-based intelligent drone route planning and risk assessment method according to an embodiment of the present invention.
[0079] Figure 5 This is a flowchart illustrating the process of driving UAV flight in a large-model-based intelligent UAV route planning and risk assessment method according to an embodiment of the present invention.
[0080] Figure 6 This is a flowchart illustrating the process of UAV avoiding dynamic obstacles in a UAV route intelligent planning and risk assessment method based on a large model according to an embodiment of the present invention.
[0081] Figure 7 This is a flowchart of obtaining dynamic risk values in a large-model-based intelligent planning and risk assessment method for unmanned aerial vehicle routes according to an embodiment of the present invention;
[0082] Figure 8 This is a flowchart illustrating the determination of the current flight status of a UAV in a large-model-based intelligent flight path planning and risk assessment method according to an embodiment of the present invention. Detailed Implementation
[0083] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.
[0084] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with relevant laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.
[0085] Figure 1 This is a flowchart of a method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) flight paths based on a large model, according to one embodiment of the present invention. Figure 1 The intelligent planning and risk assessment method for drone routes may include:
[0086] In step S1, flight path influencing factor data is acquired during the UAV's flight process, and a flight path influencing factor feature database is constructed. This data can include static obstacle data and dynamic obstacle data. Specifically, static obstacle data can include buildings, power facilities, wind turbines, etc., and the safety threshold range will differ for different static obstacle data. Dynamic obstacle data can include birds, other UAVs, etc. The flight path influencing factor data encountered during historical flights can be summarized to obtain the corresponding flight path influencing factor feature database. Furthermore, the specific forms of the aforementioned flight path influencing factor data include, but are not limited to, images and videos.
[0087] In step S2, a large visual model is selected, and fine-tuning is performed using a route influencing factor feature database. This large visual model may include a Virtual Model (VLM), which can be fine-tuned based on route influencing factor features from the database.
[0088] In step S3, the coordinates of the target point during the UAV's flight are obtained. This can be achieved by constructing a three-dimensional coordinate system to obtain the target point coordinates. Specifically, the target point coordinates can include satellite coordinates, such as those obtained using RTK or radar.
[0089] In step S4, the real-time image and real-time coordinates captured by the UAV are acquired. When the UAV is facing the target point, the real-time image captured by the UAV is the image along its flight path; the real-time coordinates captured by the UAV may include satellite coordinates.
[0090] In step S5, the current flight trajectory is obtained based on the real-time image, the fine-tuned visual model, the real-time coordinates, and the target point coordinates. Specifically, the pixel coordinates of the obstacles in the real-time image can be converted to camera coordinates, then to a 3D coordinate system, and finally to a satellite coordinate system. This allows for the determination of the relative positional relationship between the obstacles and the UAV in the image captured by the UAV, enabling dynamic updates to the UAV's flight trajectory and continuous autonomous optimization of the flight trajectory.
[0091] In step S6, the real-time disturbance factors of the current UAV are obtained. These real-time disturbance factors may include meteorological factors, communication factors, etc.
[0092] In step S7, a dynamic risk value is obtained based on the real-time disturbance factors and flight trajectory of the UAV. Specifically, the dynamic risk value during the flight process can be calculated based on the current meteorological and communication factors of the UAV. This dynamic risk value is also the flight reliability and the degree of influence of disturbance factors.
[0093] In step S8, the current flight trajectory of the UAV is updated based on the dynamic risk value. The greater the disturbance, the higher the dynamic risk value, the more unreliable the flight trajectory or the greater the risk of continued flight, thus necessitating an update to the current UAV's flight trajectory.
[0094] In steps S1 to S8, historical flight path influencing factor data of the UAV during flight is first collected, and features are extracted based on this data to construct a corresponding flight path influencing factor feature library for fine-tuning the large visual model. The target point coordinates and real-time coordinates of the current UAV flight are acquired, and real-time images captured by the UAV are collected. Based on the above, the current flight trajectory of the UAV is obtained and updated in real-time according to the actual flight situation. Real-time disturbance factors during the UAV flight are acquired, and the dynamic risk value of the current UAV flight is determined based on the real-time disturbance factors and the flight trajectory. The UAV's flight trajectory is then updated based on the magnitude of this dynamic risk value.
[0095] In this embodiment of the invention, after obtaining the data on factors influencing flight routes, a feature database of these factors can be constructed. The specific construction steps are as follows: Figure 2 As shown. Specifically, in Figure 2 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0096] In step S10, the data on factors influencing flight routes is preprocessed, and feature vectors are extracted. Preprocessing of the data may include denoising and smoothing; for images and other data, distortion correction may also be performed. Furthermore, deep convolutional neural networks or similar techniques can be used to extract features from images and other data.
[0097] In step S11, the labels corresponding to the feature vectors are obtained. The labels of the feature vectors can include static obstacles, dynamic obstacles, etc., and specifically can include buildings, power facilities, wind turbines, birds, other drones, etc.
[0098] In step S12, the feature vectors are paired with their corresponding labels to construct an influencing factor feature dataset.
[0099] In step S13, a feature database of route influencing factors is constructed based on the feature set of influencing factors.
[0100] In steps S10 to S13, the collected data on flight route influencing factors are preprocessed and features are extracted to obtain corresponding feature vectors. Simultaneously, labels are obtained for the corresponding feature vectors or labels are added to the feature vectors. Based on the feature vectors and labels, an influencing factor feature dataset is constructed, which in turn leads to a flight route influencing factor feature database.
[0101] In this embodiment of the invention, after obtaining the real-time coordinates of the current UAV during flight and the real-time images captured by the UAV, the flight trajectory of the current UAV can be acquired / updated. Specific steps can be as follows: Figure 3 As shown. Specifically, in Figure 3 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0102] In step S50, a flight trajectory is constructed based on the real-time coordinates of the current UAV and the target point coordinates. Specifically, based on the real-time coordinates of the current UAV and the target point coordinates, a straight-line flight trajectory can be constructed according to the principle that the shortest distance between two points is a straight line. If the UAV is in the takeoff phase, it can be driven to the same or similar altitude as the target point, and then a straight-line flight trajectory between the current UAV and the target point can be constructed, which is the initial flight trajectory.
[0103] In step S51, the current drone is driven to fly along the flight path. Specifically, after the flight path of the current drone is established, the drone can be driven to fly in the direction of the flight path.
[0104] In step S52, the real-time image is input into the fine-tuned visual model to determine whether there are any factors affecting the flight path in front of the drone. Specifically, during the drone's flight, real-time images of the area in front of the drone's flight path are collected and input into the fine-tuned visual model to identify and detect obstacles in front of the drone, i.e., to detect factors affecting the flight path.
[0105] In step S53, if a flight path influencing factor is detected ahead of the current drone, the current flight trajectory is updated based on the influencing factor. Specifically, if a flight path influencing factor is identified ahead of the drone's flight trajectory in the real-time image, the current drone's flight trajectory can be updated according to the type and parameters of the influencing factor, so that the drone can avoid obstacles in a timely and effective manner, ensuring the stability and reliability of the drone's flight process.
[0106] In steps S50 to S53, a straight flight trajectory is first constructed based on the real-time coordinates of the current UAV and the target point coordinates, and the UAV is driven to fly along the straight flight trajectory. Simultaneously, the UAV captures real-time images of the path ahead during flight. These real-time images are input into a finely tuned visual model, allowing for the detection and identification of obstacles ahead. This enables diagnostics of factors affecting the UAV's flight path. If obstacles exist ahead, the flight trajectory can be updated based on the obstacle's parameters to avoid them, thus ensuring the stability and reliability of the UAV's flight.
[0107] In this embodiment of the invention, when it is determined that there is an obstacle in front of the drone's flight path, the current flight path can be updated and optimized based on parameters of the obstacle, etc. Specifically, it can be as follows: Figure 4 As shown. Specifically, in Figure 4 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0108] In step S530, it is determined whether the factors affecting the flight path are static obstacles. Generally, factors affecting the flight path can be divided into static obstacles and dynamic obstacles.
[0109] In step S531, if the influencing factor on the flight path is determined to be a static obstacle, the edge coordinates of the static obstacle are obtained. These edge coordinates include the top edge coordinates, left edge coordinates, and right edge coordinates. Specifically, after obtaining the obstacle, the edges of the obstacle can be identified, including the top edge and side edges. The side edges can include the left edge and right edge. Simultaneously, based on the pixel coordinate system to world coordinate system transformation known to those skilled in the art, the three-dimensional coordinates of the top edge, left edge, and right edge of the static obstacle can be obtained. Furthermore, the current UAV's satellite coordinates and the three-dimensional coordinates of the top edge, left edge, and right edge can be transformed into the same coordinate system to obtain the relative positional relationship.
[0110] In step S532, the relative distance between the current UAV and the static obstacle is obtained according to formula (1).
[0111] (1)
[0112] in, The relative height distance between the top edge of the current static obstacle and the drone. This is the z-axis coordinate of the top edge of the current static obstacle. The current altitude coordinates of the drone. The coordinates of the left edge of the current static obstacle are the relative horizontal distance between the drone and the target. This is the x-coordinate of the left edge of the current static obstacle. Let x be the current x-axis coordinate of the drone. The coordinates of the right edge of the current obstacle are the relative horizontal distance between the drone and the obstacle. Let x be the x-coordinate of the right edge of the current static obstacle. Specifically, formula (1) applies to the case where the static obstacle is located on the path of the current flight trajectory, that is, continuing to fly along the current flight trajectory will cause the UAV to collide with the static obstacle.
[0113] In step S533, the shortest relative distance between the current drone and the static obstacle is selected. This shortest relative distance is the minimum of the relative height distance and the relative horizontal distance. Specifically, the shortest edge between the current drone and the static obstacle can include the top edge, left edge, or right edge.
[0114] In step S534, the drone's flight trajectory is updated based on the shortest relative distance between the drone and the static obstacle. Specifically, if the shortest edge between the drone and the static obstacle is determined to be one of the top edge, left edge, or right edge, the drone can be driven to fly around that shortest edge. This method effectively shortens the drone's flight path while avoiding static obstacles, improving the efficiency and reliability of the drone's flight to the target point.
[0115] In steps S530 to S534, the factors influencing the flight path are first assessed to determine whether they are static obstacles. If a static obstacle exists ahead of the UAV's current flight path, the coordinates of its top, left, and right edges can be obtained. Combined with the UAV's real-time coordinates, the closest relative distance between the UAV and the static obstacle can be calculated. Based on this closest relative distance, the shortest path for the UAV to avoid the static obstacle can be planned, and the UAV's flight path can be updated based on this shortest path. This method enables the UAV to achieve the shortest flight path while effectively avoiding static obstacles, thereby improving the UAV's flight efficiency.
[0116] In this embodiment of the invention, after obtaining the shortest relative distance between the current drone and the static obstacle, the flight trajectory of the current drone can be updated. Specific steps can be as follows: Figure 5 As shown. Specifically, in Figure 5 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0117] In step S5340, a preset safety threshold for static obstacles is obtained. Different preset safety thresholds exist for different static obstacles.
[0118] In step S5341, it is determined whether the shortest relative distance between the current drone and the static obstacle is less than or equal to a preset safety threshold. The determination of the shortest relative distance between the current drone and the static obstacle needs to be considered on a case-by-case basis. Specifically, if the static obstacle is on the current drone's flight path, the shortest relative distance between the current drone and the static obstacle should be a negative value. If the static obstacle is not on the current flight path of the drone, then the shortest relative distance between the current drone and the static obstacle should be a positive value.
[0119] In step S5342, if the shortest relative distance between the current drone and the static obstacle is less than or equal to a preset safety threshold, then it is determined whether the straight-line distance between the current drone and the static obstacle is less than or equal to the preset safety threshold. If the shortest relative distance between the current drone and the static obstacle is less than or equal to the preset safety threshold, it indicates that if the drone flies along its flight path, there is a safety risk such as a collision. Furthermore, the straight-line distance between the current drone and the static obstacle can be calculated; specifically, this straight-line distance is the distance between the current drone and the edge corresponding to the shortest relative distance.
[0120] In step S5343, if the straight-line distance between the current drone and the static obstacle is less than or equal to a preset safety threshold, the drone is driven to fly along the outer edge of the shortest relative distance. If the straight-line distance between the drone and the static obstacle is less than or equal to the preset safety threshold, it indicates a potential collision risk if the drone continues to fly. In this case, the drone's flight direction needs to be adjusted to avoid the static obstacle. Specifically, if the static obstacle is on the drone's flight path, the drone can first fly perpendicular to the flight path to the edge of the shortest relative distance, and then fly in an arc along the outer edge of the shortest relative distance to avoid the obstacle. If the static obstacle is not on the drone's flight path, the drone can directly fly in an arc along the outer edge of the shortest relative distance to avoid the obstacle. That is, in this case, the drone needs to be driven to fly along the outer edge of the shortest relative distance.
[0121] In step S5344, it is determined whether the current drone has crossed a static obstacle. Whether the current drone has crossed a static obstacle can be determined based on whether the static obstacle is captured within the field of view of the current drone.
[0122] In step S5345, if it is determined that the current drone has passed a static obstacle, the process returns to the step of constructing a flight trajectory based on the current drone's real-time coordinates and the target point coordinates. If the current drone has not captured any static obstacle, it means that the static obstacle has been passed, and a straight-line flight trajectory can be constructed based on the current drone's real-time coordinates and the target point coordinates as the current flight trajectory.
[0123] In step S5346, if it is determined that the current drone has not crossed the static obstacle, the drone is driven to fly in a direction parallel to the static obstacle. Specifically, if the current drone has captured an image of a static obstacle, it means that it has not crossed the static obstacle. Flying in a direction parallel to the static obstacle maintains a preset safety threshold distance from the static obstacle, ensuring the reliability of the drone's flight.
[0124] In step S5347, if it is determined that the straight-line distance between the current drone and the static obstacle is greater than a preset safety threshold, the drone is driven to fly along the current flight trajectory. Specifically, if the straight-line distance between the current drone and the static obstacle is greater than the preset safety distance, it indicates that the drone is still within a safe flight range, and the drone can simply be driven to fly along the flight trajectory.
[0125] In step S5348, if it is determined that the shortest relative distance between the current drone and the static obstacle is greater than a preset safety threshold, the drone is driven to fly along the current flight path. If the shortest relative distance between the current drone and the static obstacle is greater than the preset safety threshold, it means that there is no risk of collision between the current drone and the static obstacle, and the drone can be driven to fly along the current flight path.
[0126] In this embodiment of the invention, when the drone detects an obstacle ahead, if it is not a static obstacle, it may be a dynamic obstacle. Avoidance of dynamic obstacles can be based on, for example... Figure 6 Follow the steps shown. Specifically, in Figure 6 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0127] In step S5349, if the influencing factor of the flight path is determined to be a static obstacle, then the influencing factor of the flight path is determined to be a dynamic obstacle.
[0128] In step S5350, two sets of real-time images of dynamic obstacles are acquired by the drone at preset time intervals.
[0129] In step S5351, two sets of dynamic coordinates of the dynamic obstacles are obtained based on the two sets of real-time images of the dynamic obstacles. These two sets of coordinates of the dynamic obstacles are also three-dimensional coordinates.
[0130] In step S5352, the movement direction of the dynamic obstacle is obtained according to formula (2).
[0131] (2)
[0132] in, Let be the unit vector representing the direction of movement of the dynamic obstacle. This represents the velocity vector of the dynamic obstacle. Specifically, , , , ;in, This represents the absolute displacement of the dynamic obstacle along the x-axis. This represents the absolute displacement of the dynamic obstacle along the y-axis. This represents the absolute displacement of the dynamic obstacle along the z-axis. , The three-dimensional coordinates of the dynamic obstacle before and after the preset time interval are respectively. For the preset time interval, , , Let x represent the flight velocity components of the UAV on the x-axis, y-axis, and z-axis, respectively.
[0133] In step S5353, the drone is driven to avoid the obstacle based on its movement direction. Specifically, after obtaining the unit vector of the obstacle's movement direction, the drone can be driven to fly in a direction perpendicular to that unit vector to achieve avoidance.
[0134] In step S5354, it is determined whether the current drone has crossed a dynamic obstacle. Whether the current drone has crossed a dynamic obstacle can be determined based on whether the dynamic obstacle is captured within the field of view of the current drone.
[0135] In step S5355, if it is determined that the current drone has passed the dynamic obstacle, the process returns to the step of constructing a flight trajectory based on the current drone's real-time coordinates and the target point coordinates. If the current drone has not captured any dynamic obstacle, it means that the dynamic obstacle has been passed, and a straight-line flight trajectory can be constructed based on the current drone's real-time coordinates and the target point coordinates as the current flight trajectory.
[0136] In step S5356, if it is determined that the current drone has not crossed the dynamic obstacle, the drone is driven to fly along a direction parallel to the movement of the dynamic obstacle. Specifically, if the current drone has captured images of the dynamic obstacle, it means that it has not crossed the obstacle. Flying along a direction parallel to the dynamic obstacle maintains a preset safety threshold distance from the obstacle, ensuring the reliability of the drone's flight.
[0137] In this embodiment of the invention, after obtaining the disturbance factors of the current UAV flight, the dynamic risk value can be calculated based on the disturbance factors and the flight trajectory. The specific calculation steps are as follows: Figure 7 As shown. Specifically, in Figure 7 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0138] In step S70, the meteorological disturbance factor of the current UAV is obtained according to formula (3).
[0139] (3)
[0140] in, For the current weather disturbance factors of drones, Given the current wind speed in the flight environment, Wind speed is considered for safe flight conditions. Specifically, the higher the wind speed, the greater the safety risk to the flight.
[0141] In step S71, the current communication disturbance factor of the UAV is obtained according to formula (4).
[0142] (3)
[0143] in, The current communication disturbance factor for drones, This represents the current communication distance between the drone and its controller. This refers to the safe communication distance between the drone and its controller. Specifically, the greater the communication distance, the greater the safety risk during flight.
[0144] In step S72, the dynamic risk value of the current UAV flight is obtained according to formula (5).
[0145] (5)
[0146] in, This represents the dynamic risk value of the current drone flight. , These are meteorological weights and communication weights. Specifically, the meteorological weights and communication weights can be set according to the actual situation.
[0147] In this embodiment of the invention, after obtaining the dynamic risk value of the current drone flight process, it can be determined whether to update the current drone flight trajectory based on the dynamic risk value. Specific steps can be as follows: Figure 8 As shown. Specifically, in Figure 8 Furthermore, the intelligent route planning and risk assessment method for unmanned aerial vehicles (UAVs) may also include:
[0148] In step S80, it is determined whether the dynamic risk value of the current drone flight is greater than or equal to the risk threshold.
[0149] In step S81, if the dynamic risk value of the current drone is determined to be greater than or equal to the risk threshold, the drone is driven to hover and an alarm is issued. Specifically, if the dynamic risk value is too high, which may be due to high wind speed or excessive communication distance, the drone can be driven to hover and wait, and a warning can be issued.
[0150] In step S82, if the dynamic risk value of the current drone flight is determined to be less than the risk threshold, the drone's position coordinates are checked in real time, and the current drone's flight trajectory is updated. Specifically, if the dynamic risk value is low, the real-time coordinates of the current drone can be verified to determine if it has deviated. If the deviation is large, the drone's flight trajectory can be updated; otherwise, it is not updated.
[0151] On the other hand, the present invention also provides a large-scale model-based intelligent planning and risk assessment system for unmanned aerial vehicle (UAV) routes. Specifically, the system may include the UAV itself and a controller. Specifically, the UAV itself may include a camera.
[0152] The drone is equipped with a camera to capture real-time images during its flight. A controller connected to the drone executes any of the aforementioned intelligent drone flight path planning and risk assessment methods.
[0153] Through the above technical solution, the UAV route intelligent planning and risk assessment method and system based on a large model provided by this invention collects data on route influencing factors during UAV flight, thereby constructing a corresponding route influencing factor feature database for fine-tuning the visual large model; then, based on the target point coordinates of the UAV flight and the UAV's real-time images and real-time coordinates, it obtains the real-time flight trajectory, acquires the real-time disturbance factors during the UAV flight, and obtains the dynamic risk value based on the real-time disturbance factors. The flight trajectory is then updated based on this dynamic risk value, thus ensuring the autonomous and stable flight of the UAV. By using real-time images and real-time disturbance factors to update the UAV flight trajectory in real time, the dynamic updating of the UAV flight trajectory can be effectively improved, resulting in higher trajectory planning efficiency and higher UAV flight reliability.
[0154] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0155] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0156] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0157] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0158] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0159] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0160] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0161] It should also be noted that 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. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0162] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for intelligent flight path planning and risk assessment of unmanned aerial vehicles (UAVs) based on a large model, characterized in that, include: Acquire data on flight path influencing factors during UAV flight and construct a database of flight path influencing factor characteristics; A large visual model is selected, and the visual model is fine-tuned using the aforementioned route influencing factor feature database; Obtain the coordinates of the target point during the flight of the UAV; Obtain the real-time images and coordinates captured by the drone. The current flight trajectory is obtained based on the real-time image, the fine-tuned visual model, the real-time coordinates, and the target point coordinates. Obtain the real-time disturbance factors of the currently described UAV; A dynamic risk value is obtained based on the real-time disturbance factors of the drone and its flight trajectory. The flight trajectory of the UAV is updated based on the dynamic risk value.
2. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 1, characterized in that, The data on factors affecting flight routes includes both static obstacle data and dynamic obstacle data.
3. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 1, characterized in that, The construction of a database of characteristics of factors influencing flight routes includes: The data on factors influencing the flight routes are preprocessed, and feature vectors are extracted. Obtain the label corresponding to the feature vector; Pair the feature vectors with the corresponding labels and construct an influencing factor feature dataset; Construct a feature database of route influencing factors based on the aforementioned feature set of influencing factors.
4. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 1, characterized in that, The current flight trajectory is obtained based on the real-time image, the fine-tuned visual model, the real-time coordinates, and the target point coordinates, including: The flight trajectory is constructed based on the real-time coordinates of the UAV and the coordinates of the target point; Drive the drone to fly along the flight path; The real-time image is input into the fine-tuned visual model to determine whether there are any flight path influencing factors in front of the current drone. If a flight path influencing factor is detected ahead of the current drone, the current flight path is updated according to the flight path influencing factor.
5. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 4, characterized in that, The current flight path is updated based on the aforementioned factors affecting the flight route, including: Determine whether the factors affecting the flight path are static obstacles; If the influencing factor on the flight path is determined to be a static obstacle, the edge coordinates of the static obstacle are obtained, wherein the edge coordinates include the top edge coordinates, the left edge coordinates, and the right edge coordinates; The relative distance between the UAV and the static obstacle is obtained according to formula (1). ，（1） in, The distance between the top edge of the static obstacle and the drone is the relative height distance. The z-axis coordinate of the top edge of the static obstacle is given. The altitude coordinates of the currently described UAV are: The coordinates of the left edge of the static obstacle are the relative horizontal distance between the drone and the current obstacle. The x-coordinate of the left edge of the currently described static obstacle. Let x be the x-axis coordinate of the currently described UAV. The coordinates of the right edge of the obstacle are the relative horizontal distance between the drone and the obstacle. The x-coordinate of the right edge of the currently described static obstacle; Select the shortest relative distance between the current UAV and the static obstacle; The flight trajectory of the drone is updated based on the shortest relative distance between the drone and the static obstacle.
6. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 5, characterized in that, The flight trajectory of the drone is updated based on the shortest relative distance between the drone and the static obstacle, including: Obtain the preset safety threshold of the static obstacle; Determine whether the shortest relative distance between the current drone and the static obstacle is less than or equal to the preset safety threshold; If the shortest relative distance between the current drone and the static obstacle is less than or equal to the preset safety threshold, then determine whether the straight-line distance between the current drone and the static obstacle is less than or equal to the preset safety threshold. If the current straight-line distance between the drone and the static obstacle is less than or equal to the preset safety threshold, the drone is driven to fly along the direction outside the edge corresponding to the shortest relative distance. Determine whether the drone has crossed the static obstacle; If it is determined that the UAV has crossed the static obstacle, return to the step of constructing the flight trajectory based on the real-time coordinates of the UAV and the coordinates of the target point; If it is determined that the drone has not crossed the static obstacle, the drone is driven to fly in a direction parallel to the static obstacle; If it is determined that the straight-line distance between the current drone and the static obstacle is greater than the preset safety threshold, the drone is driven to fly along the current flight trajectory. If the shortest relative distance between the drone and the static obstacle is determined to be greater than the preset safety threshold, the drone is driven to fly along the current flight path.
7. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 6, characterized in that, Updating the current flight path based on the aforementioned route influencing factors also includes: If the factors affecting the flight path are determined not to be static obstacles, then the factors affecting the flight path are determined to be dynamic obstacles. Acquire two sets of real-time images of dynamic obstacles taken by the drone at preset time intervals; Two sets of dynamic coordinates of the dynamic obstacle are obtained based on two sets of real-time images of the dynamic obstacle. The direction of movement of the dynamic obstacle is obtained according to formula (2). ，（2） in, Let be the unit vector representing the direction of movement of the dynamic obstacle. Let V be the velocity vector of the dynamic obstacle. The drone is driven to avoid the moving obstacle based on its direction of movement. Determine whether the drone has crossed the dynamic obstacle; If it is determined that the UAV has crossed the dynamic obstacle, return to the step of constructing the flight trajectory based on the real-time coordinates of the UAV and the coordinates of the target point; If it is determined that the drone has not crossed the dynamic obstacle, the drone is driven to fly along a direction of movement parallel to the dynamic obstacle.
8. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 1, characterized in that, The dynamic risk value is obtained based on the real-time disturbance factors of the drone and its flight trajectory, including: The meteorological disturbance factor of the UAV is obtained according to formula (3). ，（3） in, The weather disturbance factor for the currently described UAV, Given the current wind speed in the flight environment, For safe flight environment wind speed; The communication disturbance factor of the UAV is obtained according to formula (4). ，（4） in, The communication disturbance factor of the currently described UAV, This refers to the current communication distance between the drone and the controller. This refers to the safe communication distance between the UAV and the controller; The dynamic risk value of the current UAV flight is obtained according to formula (5). ，（5） in, This represents the dynamic risk value of the current drone flight. , These are meteorological weights and communication weights, respectively.
9. The method for intelligent planning and risk assessment of unmanned aerial vehicle (UAV) routes according to claim 8, characterized in that, Updating the current flight trajectory of the UAV based on the dynamic risk value includes: Determine whether the dynamic risk value of the current drone flight is greater than or equal to the risk threshold; If the dynamic risk value of the drone is determined to be greater than or equal to the risk threshold, the drone is driven to hover and an alarm is issued. If the dynamic risk value of the current drone flight is determined to be less than the risk threshold, the position coordinates of the drone are checked in real time and the current flight trajectory of the drone is updated.
10. A large-scale model-based intelligent flight path planning and risk assessment system for unmanned aerial vehicles (UAVs), characterized in that, include: The drone body is equipped with a camera for capturing real-time images of the drone body during flight. The controller, connected to the UAV body, is used to execute the UAV route intelligent planning and risk assessment method as described in any one of claims 1-9.