Unmanned aerial vehicle autonomous obstacle avoidance path planning method and device and storage medium

Abstract

The invention discloses an unmanned aerial vehicle autonomous obstacle avoidance path planning method and device and a storage medium, and relates to the technical field of unmanned aerial vehicle autonomous navigation. The method comprises the steps that feature information of a target obstacle is acquired through a YOLOv3 model, and the feature information comprises at least one of the following information: the shape, the volume and the color; acquiring position information of the unmanned aerial vehicle; determining a target obstacle avoidance path based on the feature information and the position information of the unmanned aerial vehicle; and generating a control strategy based on the target obstacle avoidance path, wherein the control strategy is used for controlling the flight path of the unmanned aerial vehicle. The technical problem of low reliability of the obstacle avoidance path provided by the traditional algorithm of the unmanned aerial vehicle in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of autonomous navigation technology for unmanned aerial vehicles, and more specifically, to a method, device and storage medium for autonomous obstacle avoidance path planning for unmanned aerial vehicles. Background Art

[0002] The rapid development of drone technology has brought revolutionary changes to aerial operations and transportation. In these applications, autonomous obstacle avoidance and real-time 3D path planning of drones have become key technologies to ensure flight safety and mission efficiency. With the increasing application of drones in urban environments, agricultural monitoring, disaster relief and other fields, their navigation capabilities in complex and dynamic environments have become particularly important. Traditional path planning methods often rely on simple obstacle avoidance strategies and pre-set flight paths, which are difficult to achieve effective real-time path planning when faced with unknown or dynamically changing environments.

[0003] Existing path planning technologies often face problems such as high computational complexity, slow response speed, and poor adaptability when dealing with complex environments and dynamic obstacles. These problems limit the autonomy and flexibility of drones in complex environments. In addition, since drones need to perform tasks in a changing environment, traditional algorithms often find it difficult to provide reliable obstacle avoidance paths when faced with challenges such as sudden obstacles, complex terrain, and sensor noise.

[0004] The prior art has the following disadvantages:

[0005] (1) Lack of real-time performance: The existing computer vision-based UAV path planning methods have high computational complexity when processing high-resolution images, resulting in insufficient real-time performance and unable to meet the real-time obstacle avoidance requirements in rapidly changing environments.

[0006] (2) High consumption of computing resources: Although deep learning methods perform well in feature extraction and path planning, they usually require a large amount of computing resources, which is a significant disadvantage for resource-constrained platforms such as drones.

[0007] (3) Limited adaptability: Many existing technologies have limited adaptability when faced with unknown or dynamically changing environments, and it is difficult to quickly replan paths to deal with sudden obstacles.

[0008] (4) Strong dependence on GPS: Many drone navigation systems still rely heavily on GPS signals, which can lead to a significant decrease in positioning accuracy in urban canyons or indoor environments.

[0009] (5) Insufficient sensor fusion accuracy: Although sensor fusion technology can improve positioning accuracy, data synchronization and fusion accuracy between different sensors in complex environments remain a challenge.

[0010] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0011] The embodiments of the present application provide a method, device and storage medium for autonomous obstacle avoidance path planning for a drone, so as to at least solve the technical problem of low reliability of obstacle avoidance paths provided by traditional algorithms for drones in the prior art.

[0012] According to one aspect of an embodiment of the present application, a method for autonomous obstacle avoidance path planning for an unmanned aerial vehicle is provided, comprising: using a YOLOv3 model to obtain feature information of a target obstacle, wherein the feature information includes at least one of the following: shape, volume, and color; obtaining the position information of the unmanned aerial vehicle; determining a target obstacle avoidance path based on the feature information and the position information of the unmanned aerial vehicle; and generating a control strategy based on the target obstacle avoidance path, wherein the control strategy is used to control the flight path of the unmanned aerial vehicle.

[0013] Optionally, the YOLOv3 model is used to obtain feature information of the target obstacle, including: obtaining image information of the target environment; performing target detection on the image information of the target environment using the YOLOv3 model to determine target obstacle information, wherein the target obstacle information includes at least one of the following: target obstacle type information and target obstacle position information; based on the image information of the target environment, feature extraction is performed on each target obstacle to determine feature information of each target obstacle.

[0014] Optionally, based on the feature information and the position information of the UAV, a target obstacle avoidance path is determined, including: generating a planned path based on the feature information and the position information of the UAV; determining a minimum cost target path among multiple planned paths based on a target algorithm; performing a simulation test based on the minimum cost target path to obtain a simulation test result, wherein the simulation test result includes at least one of the following: flight speed, flight altitude, battery consumption and flight time of the UAV; based on the simulation test result, optimizing the minimum cost path to obtain the target obstacle avoidance path.

[0015] Optionally, a planned path is generated based on the feature information and the position information of the UAV, including: determining the relative position relationship between the UAV and each obstacle based on the position information of the UAV and the position information of each target obstacle; generating an environmental model based on the feature information, the environmental model being used to simulate the volume and position of each target obstacle; and generating a planned path based on the relative position relationship and the environmental model.

[0016] Optionally, based on the target algorithm, determining the lowest cost target path among multiple planned paths includes: obtaining conditional parameters of each planned path, the conditional parameters including at least one of the following: path length, wind speed, air density and flight altitude; based on the conditional parameters of each planned path, determining the lowest cost initial path among multiple planned paths; optimizing the lowest cost initial path to obtain the lowest cost target path.

[0017] Optionally, based on the simulation test results, the minimum cost path is optimized, including: obtaining the hovering time and number of turns in the simulation test results; obtaining the current power of the drone; and optimizing the minimum cost path based on the hovering time, number of turns and current power of the drone.

[0018] Optionally, based on the simulation test results, the minimum cost initial path is optimized, including: obtaining the flight performance parameters of the UAV, the flight performance parameters including at least one of the following: the maximum speed of the UAV, the turning radius of the UAV and the climb rate of the UAV; based on the flight performance parameters and the simulation test results, the minimum cost initial path is optimized.

[0019] According to another aspect of an embodiment of the present application, a device for autonomous obstacle avoidance path planning for an unmanned aerial vehicle is also provided, including: a first acquisition module, the first acquisition module is used to use a YOLOv3 model to obtain feature information of a target obstacle, wherein the feature information includes at least one of the following: shape, volume, and color; a second acquisition module, the second acquisition module is used to obtain position information of the unmanned aerial vehicle; a determination module, the determination module is used to determine a target obstacle avoidance path based on the feature information and the position information of the unmanned aerial vehicle; an execution module, the execution module is used to generate a control strategy based on the target obstacle avoidance path, and the control strategy is used to control and adjust the flight path of the unmanned aerial vehicle.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present invention.

[0021] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0022] In the embodiments of the present application, based on optimizing the feature extraction efficiency of the YOLOv3 model and the real-time performance of the path planning algorithm, the autonomous navigation and obstacle avoidance capabilities of the UAV in complex environments are ensured, the flight safety and mission success rate of the UAV are improved, and technical support is provided for the application of UAVs in complex environments such as cities, forests, and tunnels, thereby solving the technical problem of low reliability of obstacle avoidance paths provided by traditional UAV algorithms in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0024] Figure 1 is a flow chart of an optional method for autonomous obstacle avoidance path planning for a drone according to an embodiment of the present application;

[0025] Figure 2 is a flow chart of an optional method for autonomous obstacle avoidance path planning for a drone according to an embodiment of the present application;

[0026] Figure 3 It is a schematic diagram of an optional drone autonomous obstacle avoidance path planning device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] According to an embodiment of the present application, an embodiment of a method for autonomous obstacle avoidance path planning for a drone is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] Figure 1 The autonomous obstacle avoidance path planning method for a drone according to an embodiment of the present application is as follows: Figure 1 As shown, the method comprises the following steps:

[0031] Step S100, using the YOLOv3 model to obtain feature information of the target obstacle, wherein the feature information includes at least one of the following: shape, volume, and color.

[0032] In step S100, the drone captures images of the surrounding environment in real time through the camera it carries during flight, and these images are input into the pre-trained YOLOv3 model for processing. The YOLOv3 model can quickly detect and locate multiple targets in the image, and output the bounding box and category information of the target obstacle. Furthermore, the system processes the bounding box information output by YOLOv3 through a deep learning network or other computer vision technologies (such as edge detection, template matching, etc.) to extract the feature information of the obstacle, including but not limited to the shape, volume and color of the obstacle. These feature information will be used for subsequent path planning and obstacle avoidance strategy formulation.

[0033] Step S200, obtaining the location information of the drone.

[0034] In step S200, the drone acquires its own position information in real time during flight through integrated sensors (such as GPS, inertial navigation system INS, visual sensors, etc.). These sensor data are transmitted to the onboard computer and processed by the data fusion algorithm to obtain the precise position of the drone. The position information of the drone is the basis for path planning and obstacle avoidance strategy formulation, ensuring that the drone can correctly adjust the flight path according to its own position and obstacle information.

[0035] Step S300, determining a target obstacle avoidance path based on the feature information and the position information of the UAV.

[0036] In step S300, after obtaining the characteristic information of the target obstacle and the real-time position of the drone, the system enters the path planning stage. The system uses the heuristic A* algorithm or a similar path planning algorithm to calculate a collision-free obstacle avoidance path based on the shape, volume and position information of the obstacle. The path planning algorithm considers the flight performance of the drone and environmental factors, such as flight altitude, wind speed, etc., to ensure that the generated path is both safe and efficient. In addition, the system also updates the path information in real time to respond to dynamic changes in obstacles or newly added obstacle detections, ensuring that the drone can quickly adapt to environmental changes.

[0037] Step S400, generating a control strategy based on the target obstacle avoidance path, the control strategy is used to control the flight path of the UAV.

[0038] In step S400, after the target obstacle avoidance path is determined, a control strategy will be generated based on the results of path planning. The control strategy will guide the UAV on how to adjust the flight direction, speed and altitude to fly along the planned path. These control strategies are converted into specific flight control instructions and sent to the flight control system of the UAV. The flight control system adjusts the flight attitude and speed of the UAV in real time according to the instructions received to ensure that the UAV can avoid obstacles safely and efficiently and fly along the planned path. The system continuously monitors the flight status of the UAV during the flight, and dynamically adjusts the control strategy according to the real-time location information and environmental changes to ensure the flight safety of the UAV and the success of the mission.

[0039] Through the above steps, based on optimizing the feature extraction efficiency of the YOLOv3 model and the real-time performance of the path planning algorithm, the autonomous navigation and obstacle avoidance capabilities of the UAV in complex environments are ensured, the flight safety and mission success rate of the UAV are improved, and technical support is provided for the application of UAVs in complex environments such as cities, forests, and tunnels.

[0040] Optionally, the YOLOv3 model is used to obtain feature information of the target obstacle, including:

[0041] Step S110, acquiring image information of the target environment.

[0042] In step S110, the camera on the drone captures the target environment images on the flight path in real time. These images contain the spatial information around the drone, including possible obstacles. The image data is transmitted to the flight control system or ground control station of the drone through a wireless communication module or a direct connection for processing. The acquired image information should ensure clarity and real-time performance so that the subsequent YOLOv3 model can accurately perform target detection and feature extraction.

[0043] Step S120, using the YOLOv3 model to perform target detection on the image information of the target environment to determine target obstacle information, wherein the target obstacle information includes at least one of the following: target obstacle type information and target obstacle position information.

[0044] In step S120, after receiving the image information of the target environment, the system inputs the image data into the YOLOv3 model for processing. The YOLOv3 model is an advanced target detection algorithm that can detect and locate multiple targets in an image in real time, and output the type information and location information of the target obstacle. The type information helps the drone identify the nature of the obstacle, such as pedestrians, vehicles, trees, etc., while the location information provides the specific coordinates of the obstacle in the image, which can be used to calculate the relative position of the obstacle in the real space. Through the YOLOv3 model, the drone can quickly and accurately identify potential obstacles, providing key information for subsequent obstacle avoidance and path planning.

[0045] Step S130 , based on the image information of the target environment, feature extraction is performed on each target obstacle to determine feature information of each target obstacle.

[0046] In step S130, after determining the type and location information of the target obstacle, feature extraction is further performed on each target obstacle. This process involves analyzing the shape, size, texture, color and other features of the target obstacle in order to more accurately understand and describe the obstacle. Feature extraction can be performed through a deep learning network or other computer vision techniques that can extract more detailed obstacle features from the bounding box output by the YOLOv3 model. Feature information is crucial for the formulation of obstacle avoidance strategies for drones, as it can help drones make more accurate obstacle avoidance decisions during flight and avoid collisions with obstacles.

[0047] Combination Figure 2 As shown in the figure, through the above steps, the YOLOv3 model can be effectively used to obtain the characteristic information of the target obstacle, providing a solid data foundation for the autonomous obstacle avoidance and real-time three-dimensional path planning of the UAV.

[0048] Optionally, determining a target obstacle avoidance path based on the feature information and the position information of the UAV includes:

[0049] Step S310, generating a planned path based on the feature information and the location information of the drone.

[0050] In step S310, the system initializes the path planning algorithm by analyzing the obstacle feature information detected by the YOLOv3 model and the real-time location data of the drone. The algorithm considers the shape, volume, and position of the obstacles in the environmental model, while evaluating the current flight status of the drone (such as speed, altitude, and remaining battery). At this stage, the system generates multiple possible planning paths that bypass obstacles and ensure that the drone does not collide during flight. The path planning algorithm also takes into account the flight performance limitations of the drone, such as maximum speed and turning radius, and the generated path needs to meet these constraints.

[0051] Step S320, based on the target algorithm, determine the lowest cost target path among the multiple planned paths.

[0052] In step S320, after generating multiple planned paths, the system uses a heuristic algorithm (such as the A* algorithm) to perform a cost evaluation on these paths. The cost evaluation may include factors such as path length, flight time, and energy consumption. The system calculates the cost of each path and selects the path with the lowest cost as the target path. This process ensures that the flight path selected by the drone is optimal, which is both safe and can effectively save energy and improve flight efficiency.

[0053] Step S330, performing a simulation test based on the minimum cost target path to obtain a simulation test result, wherein the simulation test result includes at least one of the following: the flight speed, flight altitude, battery consumption and flight time of the drone.

[0054] In step S330, after determining the lowest cost target path, the system will conduct a simulation test to evaluate the feasibility and efficiency of the path. The simulation test is conducted in a virtual environment and can simulate various complex flight conditions, such as wind speed changes, temperature differences, flight altitude restrictions, etc. By simulating the UAV to fly along the target path, the system can collect key data such as the flight speed, flight altitude, battery consumption, and flight time of the UAV, which constitute the simulation test results. The analysis of the simulation test results helps the system evaluate the performance of the path planning algorithm and provide a basis for subsequent path optimization.

[0055] Step S340: Based on the simulation test results, the minimum cost path is optimized to obtain a target obstacle avoidance path.

[0056] In step S340, the determined minimum cost path is optimized based on the simulation test results. The optimization goals may include reducing flight time, reducing energy consumption, improving path smoothness, etc. In this process, the performance of the UAV in the simulation test is analyzed, possible path bottlenecks or flight obstacles are identified, and then the path planning algorithm is adjusted to generate a new optimized path. This process may involve local or global adjustments to the path to obtain the best obstacle avoidance path. The optimized path will be simulated again until it meets the predetermined performance standards to ensure the flight safety and mission efficiency of the UAV in complex environments.

[0057] The above steps form a closed-loop path planning and optimization process. From obtaining obstacle feature information to generating the optimal obstacle avoidance path, each step is based on real-time data and algorithm optimization, ensuring that the UAV can achieve efficient and safe autonomous flight in complex environments. Through these meticulous steps, the present invention aims to provide a comprehensive and reliable real-time three-dimensional path planning method for autonomous obstacle avoidance of UAVs, providing technical support for the wide application of UAVs in agricultural monitoring, urban mapping, logistics distribution, disaster relief and other fields.

[0058] Optionally, based on the feature information and the location information of the UAV, a planned path is generated, including:

[0059] Step S311, based on the position information of the drone and the position information of each target obstacle, determine the relative position relationship between the drone and each obstacle.

[0060] In step S311, the relative position relationship between the drone and each obstacle is calculated based on the real-time position information of the drone and the position information of each target obstacle obtained through the YOLOv3 model. This includes information such as the direction, distance and height difference of the obstacle relative to the drone, providing an accurate positioning basis for subsequent path planning. By accurately calculating the relative position of the drone and the obstacle, the system can better understand the environment in which the drone is located, laying the foundation for generating a safe planned path.

[0061] Step S312: Generate an environment model based on the feature information, where the environment model is used to simulate the volume and position of each target obstacle.

[0062] In step S312, an environment model is generated based on the feature information. Feature information, such as the shape, volume, and color of obstacles, is used to create a three-dimensional environment model that contains detailed information of all detected obstacles. The environment model not only simulates the static position of obstacles, but also takes into account the dynamic changes of obstacles, such as moving pedestrians or vehicles. By building an environment model, the system can intuitively display the obstacles on the UAV's flight path, providing an actionable environment view for the path planning algorithm.

[0063] Step S313, generating a planned path based on the relative position relationship and the environment model.

[0064] In step S313, the system generates a planned path based on the relative position relationship between the drone and the obstacle and the environmental model using a heuristic A* algorithm or other path planning algorithm. The path planning algorithm takes into account the flight performance of the drone, environmental factors, and the spatial distribution of obstacles to ensure that the generated path is optimal, that is, to minimize flight time and energy consumption while avoiding collisions with any obstacles. The algorithm evaluates the costs of different paths and selects the path with the lowest cost as the planned path. This process may involve multiple iterations and optimizations of the path to adapt to dynamic changes in the environment and ensure that the drone can fly safely and efficiently in a complex environment.

[0065] Through the above steps, the present invention describes in detail how to generate a planned path based on feature information and the position information of the UAV in the real-time three-dimensional path planning method for autonomous obstacle avoidance of the UAV, aiming to improve the flight safety and mission execution efficiency of the UAV.

[0066] Optionally, based on a target algorithm, determining a minimum cost target path among a plurality of planned paths includes:

[0067] Step S321, obtaining condition parameters of each planned path, the condition parameters including at least one of the following: path length, wind speed, air density and flight altitude.

[0068] In step S321, in the path planning of the drone, the system will generate multiple possible obstacle avoidance paths based on the obstacle information detected by the YOLOv3 model and the real-time location information of the drone. Next, the system needs to evaluate the cost of these paths to select the most economical and efficient flight path. To this end, the system collects the conditional parameters of each planned path, which may include the total length of the path, the wind speed conditions that may be encountered during the flight, the air density, and the planned flight altitude of the drone. The collection of conditional parameters is completed through the sensors carried by the drone (such as GPS, anemometer, barometer, etc.) and the preset environmental data model, ensuring the real-time and accuracy of the parameters.

[0069] Step S322: determining the lowest cost initial path among the multiple planned paths based on the condition parameters of each planned path.

[0070] In step S322, after collecting the conditional parameters of each planned path, the system uses a heuristic algorithm (such as the A* algorithm) to calculate the total cost of each path. The cost calculation is not only based on the length of the path, but also comprehensively considers the impact of factors such as wind speed, air density, and flight altitude on the energy consumption of the drone. For example, higher wind speeds or different air densities may increase the energy consumption of the drone, and different flight altitudes may also affect the flight efficiency of the drone. The system converts these conditional parameters into cost values ​​through a weighted formula, and then selects the path with the lowest cost value as the initial minimum cost path. This process ensures that the drone can not only avoid obstacles safely, but also complete the flight mission with minimal energy consumption.

[0071] Step S323, optimizing the minimum cost initial path to obtain the minimum cost target path.

[0072] It should be noted that although the minimum cost initial path generated in step S322 has taken into account multiple conditional parameters, in some cases, the initial path may not be optimal. In order to further improve flight efficiency and reduce costs, the minimum cost initial path will be optimized in step S323. The optimization process may include adjusting the flight altitude to take advantage of better aerodynamic conditions, replanning the path to reduce wind resistance or avoid flight areas with higher energy consumption, etc. The system will repeatedly calculate the optimized path cost and compare it with the original minimum cost path until the path with the lowest cost is found as the target path. This optimization process utilizes the real-time data of the drone during flight and continuous monitoring of environmental conditions to ensure that the drone can achieve the lowest cost target path in actual flight, thereby improving flight efficiency and extending battery life.

[0073] The whole process reflects the system's refined management of the UAV flight cost. Through multi-path planning, condition parameter evaluation and path optimization, the system not only ensures the safe flight of the UAV, but also strives to minimize the total cost of the flight mission, reflecting the efficiency and economy of the present invention in the real-time three-dimensional path planning of the UAV autonomous obstacle avoidance.

[0074] Optionally, based on the simulation test results, the minimum cost path is optimized, including:

[0075] Step S341, obtaining the hovering duration and the number of turns in the simulation test results.

[0076] In step S341, the hovering time and the number of turns are extracted from the simulation test results as indicators for evaluating the path quality. The hovering time reflects the time the drone needs to pause to avoid obstacles during flight, while the number of turns reflects the complexity of the path in path planning.

[0077] Step S342, obtaining the current power level of the drone.

[0078] In step S342, the power information will be used to evaluate the energy efficiency of the path planning algorithm and determine whether the drone can complete the entire flight mission without charging in the middle. The system will obtain the drone's power data in real time through the onboard power monitoring module.

[0079] Step S343, optimizing the minimum cost path based on the hovering time, the number of turns and the current battery power of the drone.

[0080] In step S343, the optimization goal is to reduce the hovering time and the number of turns of the drone, while taking into account the power consumption of the drone, to ensure the efficiency of the path planning algorithm and the flight safety of the drone. The system may achieve this goal by adjusting the ratio of straight segments and turning segments of the path, optimizing the obstacle avoidance strategy, or replanning the path.

[0081] The above optimization strategies aim to reduce the energy consumption of drones in complex environments, improve flight efficiency, and ensure flight safety. This step ensures that the drone can maintain sufficient power to deal with emergencies such as emergency landings or returning to base while performing its mission. Through simulation tests and continuous analysis of actual flight data, the system can continuously optimize the path planning algorithm to adapt to the autonomous obstacle avoidance and path planning needs of drones under different missions and environmental conditions.

[0082] Optionally, based on the simulation test results, the minimum cost initial path is optimized, including:

[0083] Step S344, obtaining the flight performance parameters of the UAV, the flight performance parameters including at least one of the following: the maximum speed of the UAV, the turning radius of the UAV and the climb rate of the UAV.

[0084] In step S344, the maximum speed of the drone can help the system plan a path that is consistent with the drone's flight capabilities, avoiding control difficulties caused by setting a flight speed that is too high; the turning radius is used to ensure that the turning action in the path is executable by the drone; and the climb rate takes into account the drone's maneuverability in the vertical direction, ensuring that the drone's vertical flight limitations are taken into account when planning the path.

[0085] Step S345, optimizing the minimum cost initial path based on the flight performance parameters and simulation test results.

[0086] In step S345, based on the simulation test results, the system optimizes the path, with the goal of improving the adaptability and efficiency of the path while ensuring flight safety. The optimization process may include the following strategies:

[0087] Path smoothness optimization: Adjust sharp corners in the path to reduce the maneuvering of the drone and avoid energy loss and increased difficulty in flight control caused by frequent changes in flight direction.

[0088] Flight speed adjustment: Adjust the flight speed based on the maximum speed of the drone and the constraints in the path (such as obstacles, wind speed, etc.) to ensure that the drone operates within a safe speed range while completing the mission as quickly as possible.

[0089] Turn Radius Adjustment: Ensure that the turning maneuvers in the path comply with the physical limitations of the drone and avoid setting a too small turning radius, which may cause the drone to be unable to perform or increase flight risks.

[0090] Climb / descent path optimization: Taking into account the climb and descent rates of the drone, the path in the vertical direction is optimized to ensure that the drone can smoothly perform climb and descent actions and avoid flight obstructions in complex terrain.

[0091] The optimized path through the above steps will be more in line with the actual flight capability of the UAV, can better adapt to complex environments, reduce uncertainty during flight, and improve the success rate and safety of mission execution. This process reflects the advancement and practicality of the present invention in the real-time three-dimensional path planning of autonomous obstacle avoidance of UAVs. Through continuous simulation and optimization, the present invention can provide a more efficient and safer flight path planning solution.

[0092] Example 2

[0093] Combination Figure 3 As shown, an embodiment of the present application also provides an autonomous obstacle avoidance path planning device for a UAV, comprising: a first acquisition module 51, a second acquisition module 52, a determination module 53 and an execution module 54, wherein the first acquisition module 51 is used to obtain feature information of a target obstacle using a YOLOv3 model, wherein the feature information includes at least one of the following: shape, volume and color; the second acquisition module 52 is used to obtain the position information of the UAV; the determination module 53 is used to determine a target obstacle avoidance path based on the feature information and the position information of the UAV; the execution module 54 is used to generate a control strategy based on the target obstacle avoidance path, and the control strategy is used to control and adjust the flight path of the UAV.

[0094] Example 3

[0095] An embodiment of the present application further provides an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of the present invention when running.

[0096] Example 4

[0097] An embodiment of the present application further provides a computer-readable storage medium, which includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present invention.

[0098] Example 5

[0099] An embodiment of the present application further provides a computer program product, including a computer program, which implements the methods in various embodiments of the present invention when executed by a processor.

[0100] Example 6

[0101] An embodiment of the present application further provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium is used to store a computer program, and when the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0102] Example 7

[0103] The embodiments of the present application further provide a computer program, which implements the methods in the above-mentioned embodiments of the present invention when executed by a processor.

[0104] This application achieves at least the following technical effects:

[0105] (1) Improving real-time performance: The present invention aims to improve the real-time obstacle avoidance capability of UAVs in complex environments by integrating the lightweight YOLOv3 target detection algorithm and the heuristic A* path planning algorithm.

[0106] (2) Reduce computing resource consumption: By optimizing the algorithm structure and reducing the demand for computing resources, UAVs can perform complex path planning tasks under resource-constrained conditions.

[0107] (3) Enhanced adaptability: The method of the present invention can quickly respond to environmental changes, adapt to dynamic obstacles, and provide a flexible path planning strategy.

[0108] (4) Reduce dependence on GPS: The present invention uses visual information to perform path planning, reduces dependence on GPS signals, and improves navigation capabilities in environments with limited GPS signals.

[0109] (5) Improving sensor fusion accuracy: The present invention improves the accuracy of sensor data fusion through precise feature extraction and path planning, thereby achieving more accurate positioning and navigation in complex environments.

[0110] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0111] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0112] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0113] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0114] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0115] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0116] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for autonomous obstacle avoidance path planning for a drone, characterized in that: include: Using the YOLOv3 model to obtain feature information of the target obstacle, wherein the feature information includes at least one of the following: shape, volume, and color; Get the location information of the drone; Determine a target obstacle avoidance path based on the feature information and the position information of the drone; A control strategy is generated based on the target obstacle avoidance path, and the control strategy is used to control the flight path of the UAV.

2. The method according to claim 1, characterized in that Acquiring the feature information of the target obstacle using the YOLOv3 model includes: Acquire image information of the target environment; Using the YOLOv3 model to perform target detection on the image information of the target environment to determine target obstacle information, wherein the target obstacle information includes at least one of the following: target obstacle type information and target obstacle position information; Based on the image information of the target environment, feature extraction is performed on each of the target obstacles to determine the feature information of each of the target obstacles.

3. The method according to claim 2, characterized in that Determining the target obstacle avoidance path based on the feature information and the position information of the drone includes: Generate a planned path based on the feature information and the location information of the drone; Based on the target algorithm, determining the lowest cost target path among the plurality of planned paths; Based on the minimum cost target path, a simulation test is performed to obtain a simulation test result, wherein the simulation test result includes at least one of the following: the flight speed, flight altitude, battery consumption and flight time of the UAV; Based on the simulation test results, the minimum cost path is optimized to obtain the target obstacle avoidance path.

4. The method according to claim 3, characterized in that Generating the planned path based on the feature information and the location information of the drone includes: Based on the position information of the drone and the position information of each target obstacle, determining the relative position relationship between the drone and each obstacle; Based on the feature information, an environment model is generated, wherein the environment model is used to simulate the volume and position of each target obstacle; The planned path is generated based on the relative position relationship and the environment model.

5. The method according to claim 3, characterized in that: Determining the lowest cost target path among the plurality of planned paths based on the target algorithm includes: Acquire condition parameters of each of the planned paths, wherein the condition parameters include at least one of the following: path length, wind speed, air density, and flight altitude; Determining a minimum cost initial path among a plurality of the planned paths based on the condition parameters of each of the planned paths; The minimum cost initial path is optimized to obtain the minimum cost target path.

6. The method according to claim 3, characterized in that Based on the simulation test results, the minimum cost initial path is optimized, including: Obtaining the hovering duration and the number of turns in the simulation test results; Get the current power of the drone; The minimum cost path is optimized based on the hovering duration, the number of turns, and the current battery power of the drone.

7. The method according to any one of claims 4 to 6, characterized in that Based on the simulation test results, the minimum cost path is optimized, including: Acquire flight performance parameters of the UAV, wherein the flight performance parameters include at least one of the following: a maximum speed of the UAV, a turning radius of the UAV, and a climbing rate of the UAV; The minimum cost initial path is optimized based on the flight performance parameters and the simulation test results.

8. An autonomous obstacle avoidance path planning device for a drone, characterized in that: include: A first acquisition module, wherein the first acquisition module is used to acquire feature information of the target obstacle using a YOLOv3 model, wherein the feature information includes at least one of the following: shape, volume, and color; A second acquisition module, the second acquisition module is used to obtain the location information of the drone; A determination module, the determination module is used to determine a target obstacle avoidance path based on the feature information and the position information of the drone; An execution module is used to generate a control strategy based on the target obstacle avoidance path, and the control strategy is used to control and adjust the flight path of the drone.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The invention comprises a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images