Path planning method and device for unmanned aerial vehicle and computer program product
The drone path planning is optimized through the ant colony algorithm, which solves the problem of time-consuming and labor-consuming manual operation, and realizes efficient power inspection path planning, which is highly adaptable and meets the needs of power inspection.
Patent Information
- Application Number
- CN202411066452.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-05
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The path method of manually manipulating the drone in the prior art is time-consuming and labor-intensive, and it is difficult to meet the needs of power inspection.
The ant colony algorithm is used to plan the drone path, combine environmental information and drone performance, and through the iterative optimization of multiple sets of training data and ant colony, the optimal inspection path is generated and adjusted in real time to adapt to environmental changes.
It improves the efficiency and path compliance of drone inspections, can find better paths in a short time, adapt to different inspection needs and environmental changes, and reduces flight time and energy consumption.
Smart Images

Figure CN118963385A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of electric power inspection technology, and more specifically, to a path planning method and device for an unmanned aerial vehicle (UAV), a computer-readable storage medium, and a computer program product. Background Art
[0002] In the power system, the operation and maintenance of the transmission system is an important task. At present, power inspection work is often completed by manually controlling drones on site. However, as the demand for inspections continues to increase, the method of manually controlling the path of drones is time-consuming and labor-intensive, and the planned inspection path is difficult to meet the inspection needs. Summary of the invention
[0003] The main purpose of the present application is to provide a path planning method, device, computer-readable storage medium and computer program product for an unmanned aerial vehicle, so as to at least solve the problem that the manual control of the path of the unmanned aerial vehicle in the prior art is time-consuming and labor-intensive, and the planned inspection path is difficult to meet the inspection requirements.
[0004] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a path planning method for an unmanned aerial vehicle is provided, comprising: obtaining relevant information of an unmanned aerial vehicle inspection, wherein the relevant information includes at least one or more of the inspection requirements, the environmental information of the inspection and the performance of the unmanned aerial vehicle, the inspection requirement is the requirement for inspecting a target object, the target object includes one or more of equipment, facilities and areas, the environmental information includes at least one or more of terrain, climate and vegetation, and the performance of the unmanned aerial vehicle includes at least one or more of flight speed, flight altitude, climb rate, descent rate and flight distance; constructing a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical relevant information obtained within a historical time period and a historical optimal path corresponding to the historical relevant information; inputting the relevant information into the path planning model to obtain the optimal path corresponding to the relevant information; controlling the unmanned aerial vehicle to fly according to the optimal path to inspect the target object.
[0005] Optionally, after obtaining relevant information of the drone inspection, the method further includes: based on the inspection requirements, obtaining spatial distribution feature information of the target object to be inspected, wherein the spatial distribution feature information includes at least one or more of position distribution information, morphological feature distribution information and topological relationship information, the position distribution information is the position of the target object, the morphological feature distribution information is the characteristics of the shape of the target object, and the topological relationship information is the distance relationship between multiple target objects; classifying the target objects to be inspected according to the spatial distribution feature information to obtain a classification result.
[0006] Optionally, after obtaining relevant information of the drone inspection, the method also includes: obtaining key track points, wherein the key track points are track points pre-set in the inspection requirements; obtaining obstacle avoidance points corresponding to static obstacles, wherein the obstacle avoidance points are the location points of static obstacles in the target object to be inspected; and constructing a three-dimensional model based on at least the key track points and the obstacle avoidance points.
[0007] Optionally, before constructing the path planning model, the method further includes: obtaining a feature map of the target object to be inspected, wherein the feature map is a histogram based on point cloud features and color gradients; filtering and fusing the feature map according to a target formula to obtain a response map, wherein the response map is used to characterize the response degree of the filter in different feature maps, and the target formula is:
[0008] f′(x)=∑ i=1 α i ×f i (x),
[0009] f′(x) represents the response graph, α i represents the weight of the i-th feature map, f i (x) represents the i-th feature map,
[0010]
[0011] k i represents the peak value of the ith filter response, k i =max(f i (x)), k j represents the peak value of j filter responses, k j =max(f j (x)), N represents the number of feature maps.
[0012] Optionally, before constructing the path planning model, the method also includes: initializing the population, randomly placing M ants, M≥2, wherein the route walked by each ant represents the inspection path of the drone; using the relevant information to determine the next moving position of the ant; when all the ants have reached the inspection end point, obtaining the optimal solution in the iterative process to obtain the optimal path.
[0013] Optionally, after inputting the relevant information into the path planning model and obtaining the optimal path corresponding to the relevant information, the method also includes: controlling the UAV to fly along the optimal path, and acquiring image data of the optimal path during the inspection process in real time; performing image analysis on the image data to determine whether there are dynamic obstacles on the optimal path of the UAV, and obtaining dynamic detection results.
[0014] Optionally, after performing image analysis on the image data to determine whether there is a dynamic obstacle on the optimal path of the UAV and obtaining a dynamic detection result, the method further includes: obtaining obstacle information of the dynamic obstacle, wherein the obstacle information includes at least one or more of the position of the dynamic obstacle, the flight trajectory of the dynamic obstacle, and the overlap information of the dynamic obstacle with the optimal path; generating an obstacle avoidance route according to the obstacle information, and controlling the UAV to fly along the obstacle avoidance route; and when the UAV has passed the dynamic obstacle, controlling the UAV to return to the optimal path to fly.
[0015] According to another aspect of the present application, a path planning device for an unmanned aerial vehicle is provided, comprising: a first acquisition unit, for acquiring relevant information of an unmanned aerial vehicle inspection, wherein the relevant information includes at least one or more of inspection requirements, inspection environment information and performance of the unmanned aerial vehicle, the inspection requirement is the requirement for inspecting a target object, the target object includes one or more of equipment, facilities and areas, the environmental information includes at least one or more of terrain, climate and vegetation, and the performance of the unmanned aerial vehicle includes at least one or more of flight speed, flight altitude, climb rate, descent rate and flight distance; a first construction unit, for constructing a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical relevant information obtained within a historical time period and a historical optimal path corresponding to the historical relevant information; a first processing unit, for inputting the relevant information into the path planning model to obtain an optimal path corresponding to the relevant information; and a first control unit, for controlling the unmanned aerial vehicle to fly along the optimal path to inspect the target object.
[0016] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the path planning methods for the drone.
[0017] According to another aspect of the present application, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any one of the path planning methods for the drone are implemented.
[0018] By applying the technical solution of the present application, an ant colony algorithm is used to plan the inspection path of the drone. The ant colony algorithm is adaptive and can adjust the path planning in real time according to environmental changes to adapt to different situations and requirements. The ant colony algorithm is a heuristic algorithm that can find a better path planning in a shorter time, thereby making the planned inspection path meet the inspection requirements and improving the efficiency of drone inspections. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0020] Figure 1 A hardware structure block diagram of a mobile terminal for executing a path planning method for a drone provided in an embodiment of the present application is shown;
[0021] Figure 2 A schematic diagram of a process flow of a path planning method for a drone provided according to an embodiment of the present application is shown;
[0022] Figure 3 The data classification diagram of the track parameterized modeling of this scheme is shown;
[0023] Figure 4 The following is a schematic diagram showing the logic flow of the present solution;
[0024] Figure 5 The target tracking process flow diagram of this scheme is shown;
[0025] Figure 6 A structural block diagram of a path planning device for a drone provided according to an embodiment of the present application is shown.
[0026] The above drawings include the following reference numerals:
[0027] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION
[0028] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0029] 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.
[0030] 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 sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. 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.
[0031] As introduced in the background technology, the method of manually controlling the path of the drone in the prior art is time-consuming and labor-intensive, and the planned inspection path is difficult to meet the inspection requirements. To solve the above problems, the embodiments of the present application provide a drone path planning method, device, computer-readable storage medium and computer program product.
[0032] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.
[0033] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 1 is a hardware structure block diagram of a mobile terminal of a method for planning a path for a drone according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1Different configurations shown.
[0034] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.
[0035] In this embodiment, a path planning method for a drone running on a mobile terminal, a computer terminal or a similar computing device 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 may be executed in an order different from that shown here.
[0036] Figure 2 FIG. 1 is a flow chart of a method for planning a path for a drone according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:
[0037] Step S201, obtaining relevant information of the drone inspection, wherein the relevant information includes at least one or more of the inspection demand, the inspection environment information and the performance of the drone, the inspection demand is the demand for inspecting the target object, the target object includes one or more of equipment, facilities and areas, the environmental information includes at least one or more of terrain, climate and vegetation, and the performance of the drone includes at least one or more of flight speed, flight altitude, climb rate, descent rate and flight distance;
[0038] Specifically, according to the inspection requirements of power equipment, the geometric features and location information of the equipment in the inspection area can be collected through multiple channels. Specifically, relevant information can be obtained, namely the target object to be inspected, the environment of the target object to be inspected, and the performance parameters of the drone.
[0039] Step S202, constructing a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical related information obtained within a historical time period and a historical optimal path corresponding to the historical related information;
[0040] Specifically, the machine training can be carried out through the ant colony algorithm to obtain a path planning model, and the path planning model can output the optimal path corresponding to the relevant information.
[0041] Step S203, inputting the above relevant information into the above path planning model to obtain the optimal path corresponding to the above relevant information;
[0042] Specifically, relevant information can be input into a trained path planning model, and the path planning model outputs an optimal path.
[0043] Step S204: Control the UAV to fly along the optimal path to inspect the target object.
[0044] Specifically, the drone can be controlled to fly for inspection, and the drone performs inspection along the optimal path.
[0045] Through this embodiment, an ant colony algorithm is used to plan the inspection path of the drone. The ant colony algorithm is adaptive and can adjust the path planning in real time according to environmental changes to adapt to different situations and requirements. The ant colony algorithm is a heuristic algorithm that can find a better path planning in a shorter time, thereby making the planned inspection path meet the inspection requirements and improving the efficiency of drone inspections.
[0046] Specifically, with the continuous development of deep learning technology, the application of machine vision systems in the field of UAV power inspection is becoming more and more widespread. The flexibility and high-altitude bird's-eye view of UAVs make adaptive inspection possible. UAV systems equipped with high-resolution cameras and image processing algorithms can analyze real-time image data and map information to avoid obstacles on the basis of covering the inspection area, minimize flight time and power consumption, determine the optimal safe inspection path, and efficiently complete the inspection task.
[0047] During drone inspections, the inspection targets usually present diverse forms and complex network structures, and the inspection environment and scenes also have the same problems, including obstructions, weather conditions, lighting changes, etc. This not only puts higher requirements on the drone positioning system, but also faces severe challenges in the target recognition and tracking algorithm capabilities in the drone inspection system. In addition, the computing power of the gimbal chip carried by the drone is limited and cannot afford high-complexity mathematical operations. Therefore, in order to improve the autonomy of planning routes and the robustness of target recognition problems in drone inspections, it is urgent to seek parameterized support related to inspection equipment and tracks.
[0048] Therefore, the solution proposed in this application is to improve the autonomy and robustness of UAV inspection route planning and target tracking, and to use a parameterized assisted positioning method based on multi-data fusion to cope with the occlusion challenges in complex inspection environments and improve the autonomy and reliability of UAV path planning.
[0049] The solution of the present application can extract inspection area information and perform parametric modeling; integrate multi-feature target morphology prediction methods to predict the morphology of obstructing objects such as obstacles encountered by the drone, and achieve precise positioning and obstacle avoidance; iteratively optimize the ant colony algorithm based on the improved adaptive information fusion factor to find the optimal flight path in the parametrically modeled inspection area, and maintain and update the modeled model and planned route.
[0050] During the specific implementation process, after obtaining relevant information of the drone inspection, the above method also includes the following steps: based on the above inspection requirements, obtaining the spatial distribution feature information of the above target object to be inspected, wherein the above spatial distribution feature information includes at least one or more of the position distribution information, morphological feature distribution information and topological relationship information, the above position distribution information is the position of the above target object, the above morphological feature distribution information is the characteristics of the shape of the above target object, and the above topological relationship information is the distance relationship between multiple target objects; the above target objects to be inspected are classified according to the above spatial distribution feature information to obtain classification results.
[0051] In this solution, it is necessary to obtain the spatial distribution characteristic information of the target objects to be inspected, and classify the target objects based on this information. By analyzing this information, it can help to classify the target objects, so as to perform the inspection tasks more effectively. By classifying according to the characteristics of the target objects, a more reasonable inspection path and plan can be formulated according to the inspection requirements and spatial distribution characteristic information, thereby improving the efficiency and accuracy of the inspection. Classification can help identify different types of target objects, select appropriate inspection methods and parameters in a targeted manner, and improve the precision and accuracy of the inspection. By classifying the target objects, it can ensure that all types of target objects are fully inspected to avoid omissions or omissions.
[0052] Specifically, the data structure of the inspection equipment and drone tracks is collected and designed, and the location and morphological characteristics of the target are constructed from different angles. The track has a certain dependence on the inspection demand. Generally speaking, there must be track points in the inspection demand area. Therefore, the parametric construction of the track takes the inspection demand into consideration at the same time; the parameter collection mainly comes from the construction drawings, location information, model category and obstacle information of the inspection demand, track points of previous inspections, etc., to jointly build a database for track planning; by analyzing the spatial distribution characteristics of the track points, the collected data is classified according to the location distribution, topological relationship and morphological characteristic distribution.
[0053] Spatial distribution features refer to the information collection of the spatial position of the detection target, its own morphological features, and the topological relationship between it and other detection targets in 3D modeling; Figure 3 As shown in the figure, when modeling a three-dimensional map, it is necessary to perform parameterized extraction of entities. For example, when classifying the location distribution of the detection target, it is marked according to the target number and three-dimensional coordinates; the inspection target's own shape is classified according to the characteristics of the target shape and the mathematical morphological similarity according to three-dimensional physical models such as cylinders, cones, and cuboids; the topological relationship represents the relationship between the detection targets, such as the direct relationship when there are no obstacles, the adjacent relationship when there are obstacles, etc.
[0054] During the specific implementation process, after obtaining relevant information of the drone inspection, the above method also includes the following steps: obtaining key track points, wherein the above key track points are track points pre-set in the above inspection requirements; obtaining obstacle avoidance points corresponding to static obstacles, wherein the above obstacle avoidance points are the location points of static obstacles in the above target object to be inspected; and constructing a three-dimensional model based on at least the above key track points and the above obstacle avoidance points.
[0055] In this solution, before the drone inspection mission is carried out, it is necessary to obtain key track points and obstacle avoidance points of static obstacles, and build a three-dimensional model based on these points. Based on the key track points and obstacle avoidance points, a three-dimensional model can be built, that is, a model that presents the flight path, key positions and obstacle positions of the drone in three-dimensional space. By setting key track points and obstacle avoidance points and building a three-dimensional model, obstacles in flight can be effectively avoided, flight risks can be reduced, and flight safety can be improved. Based on the pre-set track points and obstacle avoidance points, the drone can fly along the planned path, improving flight efficiency and accuracy.
[0056] Key track points are track points set in advance according to inspection requirements, usually the locations of areas or target objects that need to be focused on. By obtaining key track points, it can be ensured that the drone can accurately and efficiently fly to key locations during the inspection process, improving the efficiency and accuracy of the inspection task.
[0057] Static obstacle avoidance points are the location points of static obstacles in the target object to be inspected, which are used to avoid obstacles that the drone may encounter during flight. By obtaining these obstacle avoidance points, it can be ensured that the drone avoids obstacles during flight, ensuring flight safety and successfully completing the inspection mission.
[0058] Specifically, a parameterized model of inspection requirements is constructed by extracting category information, coordinate information, and geometric information of inspection equipment and inspection waypoints.
[0059] Specifically, Figure 4 As shown in the figure, modeling the flight track of a UAV refers to mapping the position information and geometric features of different components in the three-dimensional model into variable parameters by reasonably splitting them to achieve dynamic modeling. The parameter information of the UAV track is extracted as the parameter information of the three-dimensional modeling by extracting the category information and coordinate position information of the track. For example, the key aerial photography track points and extreme obstacle avoidance points are extracted through the inspection requirements, construction drawings, obstacle shapes, etc., to achieve the three-dimensional modeling of the UAV track driven by parameter information. Define the starting position of the UAV as the coordinate origin and establish a three-dimensional coordinate system. Map the position information of each component into a variable parameter through modeling data, and calculate the position change of the component relative to the reference point. At the same time, the route is a continuous entity connected by a series of tracks. Each track point controls the turning angle, direction and position of the route, which is the basic element for constructing a three-dimensional route. By extracting the track, obstacles and inspection requirements, an irregular triangulation algorithm is designed for rapid modeling. According to the real-time component position and obstacle information, the shape and size of the grid are dynamically adjusted to ensure that all areas that need to be inspected are covered.
[0060] Specifically, a three-dimensional Cartesian coordinate system (XYZ) is established with the coordinate origin (0, 0, 0) of the starting position of the drone as the reference point. The X axis represents the horizontal direction, the Y axis represents the vertical direction, and the Z axis represents the height direction. Then, the position change of the actual coordinates of the drone relative to the reference point is expressed as (ΔXref, ΔYref, ΔZref).
[0061] Assume that the coordinates of the component are (Xc, Yc, Zc), calculate the position change relative to the reference point, and obtain the coordinate values in the three-dimensional coordinate system: X'=Xc-Δxref, Y'=Yc-Δyref, Z'=Zc-Δzref.
[0062] Among them, X', Y', Z' represent the transformation coordinates of the component on the X, Y, and Z axes respectively, and (ΔXref, ΔYref, ΔZref) represents the change in the reference transformation origin coordinates.
[0063] According to the geometric shape and size of the component, create the corresponding geometry in the modeling software or import the existing component model, combine, layout and adjust it to build a complete three-dimensional model.
[0064] Specifically, in order to enhance the scalability of the UAV trajectory parameterization model, the occlusion problem in the actual environment and the UAV's perspective is considered, and the target morphology is predicted based on the fusion of multi-feature filters. On this basis, the robustness of the simulated route is verified in combination with the UAV's maneuverability characteristics.
[0065] In the specific implementation process, before building the path planning model, the method further includes the following steps: obtaining a feature map of the target object to be inspected, wherein the feature map is a histogram based on point cloud features and color gradients; filtering and fusing the feature map according to the target formula to obtain a response map, wherein the response map is used to characterize the response degree of the filter in different feature maps, and the target formula is:
[0066] f′(x)=∑ i=1 α i ×f i (x),
[0067] f′(x) represents the above response graph, α i represents the weight of the i-th feature map above, f i (x) represents the i-th feature map mentioned above,
[0068]
[0069] k i represents the peak value of the ith filter response, k i =max(f i (x)), k j represents the peak value of j filter responses, k j =max(f j (x)), N represents the number of the above feature maps.
[0070] In this scheme, by acquiring feature maps and performing filter fusion, the characteristic information of the target object can be described more comprehensively and accurately, which helps to improve the accuracy of UAV recognition of target objects during inspections. Filter fusion can help remove noise and interference in the data, making the data clearer and more reliable, which is conducive to accurate analysis and processing of data. By fusing different types of feature information, the description of the target object from multiple aspects can be comprehensively considered, improving the comprehensiveness and accuracy of data analysis.
[0071] The feature map is a histogram based on point cloud features and color gradients. Point cloud features are the three-dimensional point cloud data of the target object obtained by laser scanning or photography, and color gradients can be used to describe the color distribution of the target object. Combining these features into a histogram form can more intuitively present the feature information of the target object.
[0072] Filter fusion is a data processing technology that can remove noise, smooth data, enhance features, etc. by filtering feature maps, and ultimately obtain clearer and more accurate feature information. The fusion operation can integrate feature information from different sources, comprehensively consider the description of the target object by multiple features, and improve the reliability and accuracy of the data.
[0073] Specifically, according to the inspection needs and inspection targets, parameter modeling is performed for the rapid inspection area, and the effectiveness of the simulated route of the drone is verified by combining factors such as the maneuverability and visual occlusion of the drone. Due to the difference in perspective caused by the changes in drone inspections, the inspection targets or obstacle information often have occlusion problems. In order to meet the safety and robustness requirements of path planning, the target tracking technology combined with multi-feature fusion is used for morphological prediction, and multiple feature information is integrated to reduce the limitations and misjudgment risks of a single feature, improve the system's perception of environmental changes, noise interference and abnormal conditions, and enhance the system's reliability in predicting occluded objects; according to different feature types, the corresponding kernel filter is used for feature training. The features include the RGB visualization image features of the inspection photos, the point cloud features of the parameterized modeling, and the histogram features of the color gradient. The filters are calculated with the corresponding feature images respectively to obtain the kernel correlation filter response maps (i.e., feature maps) of different features, and the weights are dynamically allocated according to the importance and reliability of the features, and the final response map is generated by fusion.
[0074] In addition, a threshold δ can be set. When the similarity between the response prediction map and the actual target is high enough, that is, greater than the threshold δ, the filter parameters related to the feature are updated:
[0075]
[0076] Among them, α′ i is the ith updated filter, μ is the learning rate, which is a constant. is the filter associated with the i-th feature of the previous frame captured.
[0077] In addition, the starting position and maneuverability of the UAV can be defined, the occlusion problem in the target tracking process can be comprehensively considered, and the simulated inspection flight path of the UAV on different parameterized models can be generated by parametric modeling of the inspection area to verify the accuracy and reliability of the track parameterized model.
[0078] Specifically, the starting position of the UAV can be determined, as well as the maneuvering characteristics of the UAV, such as flight speed, flight altitude, maximum climb / descent rate, etc.
[0079] Obtain the location and shape information of obstacles, perform parametric modeling on the inspection area, divide the area into discrete grids, and define corresponding parameters for each grid, including route length, heading angle, height limit, etc. Among them, parametric modeling can flexibly adjust the three-dimensional model according to actual inspection needs and obstacle addition and deletion information.
[0080] Parametric modeling, taking into account the location and size of obstacles during modeling, planning flight paths based on ant colony algorithms, generating flight paths for drone inspections on different parametric models to avoid collisions between drones and obstacles. The flight paths under different inspection environments are saved in the path library, which is convenient for repeated calls and regional range path updates.
[0081] Use the modeling software Revit tool for 3D modeling, simulate the flight of the drone, observe the flight trajectory, check whether it meets the expected inspection requirements and obstacle avoidance strategies, and evaluate the accuracy and reliability of the path.
[0082] In some embodiments, before constructing the path planning model, the method further includes the following steps: initializing the population, randomly placing M ants, M≥2, wherein the route walked by each of the ants represents the inspection path of the drone; using the relevant information to determine the next moving position of the ants; when all of the ants have reached the inspection end point, obtaining the optimal solution in the iterative process to obtain the optimal path.
[0083] In this scheme, an ant colony needs to be initialized and M ants are randomly placed at the starting position. Each ant represents the inspection path of the drone, that is, the ant simulates the flight trajectory of the drone by moving. By using relevant information, such as the distribution of target objects, environmental information, and drone performance, the next moving position of each ant can be determined. The ants choose the moving direction based on the pheromone concentration and heuristic information to find a better inspection path. After all ants have reached the inspection end point, the optimal solution in the iteration process can be obtained based on the movement path and effect of the ants in the iteration process, that is, the optimal path. This optimal path is obtained through continuous search and optimization by the ant colony algorithm, and is expected to provide the optimal flight path planning for drone inspection tasks. The ant colony algorithm has global search capabilities and can continuously search for better solutions in the search space, which helps to find the best drone inspection path. Because the ant colony algorithm has certain randomness and adaptability, it can effectively overcome the local optimal solution trap and improve the robustness of the algorithm. The ant colony algorithm can continuously adjust the path selection strategy according to the pheromone concentration and the movement effect of the ants found during the search process to achieve adaptive optimization of path planning.
[0084] Specifically, the ant colony algorithm is used to solve the path planning based on parameterized model simulation. The inspection cost of the drone is set as the optimization function. The maneuverability characteristics of the drone and the obstacle position constraints are used as constraints. The solution is based on the ant colony algorithm. The specific implementation steps are as follows:
[0085] 1. Initialize the population, pheromone factor, heuristic function factor, pheromone volatility factor, pheromone constant, maximum number of iterations, place M ants, each ant represents a path;
[0086] 2. Probabilistically select the next moving position for each ant and determine whether the path is completed. If not, continue to move. If yes, select the best path among M ants.
[0087] 3. Calculate the path length L of each ant n , record the optimal solution in the current algorithm iteration process, that is, the shortest path;
[0088] 4. Determine whether the algorithm meets the maximum number of iterations. If yes, output the best path. If no, update the pheromone and repeat steps 1-3.
[0089] Among them, the pheromone volatilization factor is updated during the algorithm iteration process, and the pheromone factor connecting each route point is updated using an improved adaptive adjustment method:
[0090]
[0091]
[0092] Among them, τ i,j (t+1) is the pheromone value on the path from i to j after t+1 cycles, τ i,j (t)×(1-ρ) is the pheromone value of the tth cycle, ρ is the pheromone volatility factor, 0≤ρ≤1, n is the nth ant, is the pheromone value of the nth ant on the path from i to j; Q is the initialized pheromone constant, L n is the total length of the path of the nth ant in this cycle, The pheromone attenuation factor decreases as the number of algorithm cycles increases, enhancing the convergence ability of the algorithm.
[0093] Specifically, modeling software is used to simulate various inspection environments, taking into account the maneuverability characteristics of various UAVs and the actual inspection environment; secondly, under different conditions and environments, the bionic evolutionary algorithm and ant colony algorithm are continuously optimized, and the minimization of the UAV inspection cost function is set to iterate and update the path planning algorithm; finally, when UAVs equipped with camera equipment are tracking dynamic obstacles, the problem of target occlusion is considered, and the feature filter is fused to predict the target shape and position to enhance the obstacle avoidance capability.
[0094] In addition, the actual planned path of the UAV and the parameter changes in the inspection area can be periodically updated to the parameterized track model. Through maintenance and training, a track planning library can be established to meet the needs of personnel to judge, manage and strengthen the planned routes.
[0095] During the actual inspection process, there may be changes in obstacles, as well as dynamic obstacles such as kites and birds. In some embodiments, after the above-mentioned relevant information is input into the above-mentioned path planning model to obtain the optimal path corresponding to the above-mentioned relevant information, the above-mentioned method also includes the following steps: controlling the above-mentioned UAV to fly according to the above-mentioned optimal path, and obtaining the image data of the above-mentioned optimal path during the inspection process in real time; performing image analysis on the above-mentioned image data to determine whether there are dynamic obstacles on the above-mentioned optimal path of the above-mentioned UAV, and obtaining dynamic detection results.
[0096] In this solution, according to the optimal path planning, the drone is controlled to fly along the optimal path. The optimal path is optimized. Flying along the optimal path can effectively reduce flight time and energy consumption and improve the efficiency of the inspection task. During the flight, the image data on the flight path of the drone is obtained in real time. These image data can be obtained through the camera or sensor of the drone, and contain real-time information of the inspection environment. Image analysis is performed on the acquired image data, and computer vision technology is used to identify and detect whether there are dynamic obstacles on the optimal path. Dynamic obstacles may be moving vehicles, personnel or other obstacles, which pose a threat to the flight safety of the drone. Through image analysis, dynamic detection results can be obtained, that is, whether there are dynamic obstacles on the optimal path. According to the detection results, the flight path of the drone can be adjusted in time or other measures can be taken to ensure flight safety and complete the inspection task. Dynamic obstacles on the optimal path can timely detect possible flight obstacles, avoid collisions and accidents, and improve flight safety. Through image analysis technology, dynamic obstacles can be automatically identified, manual intervention can be reduced, and inspection efficiency and timeliness can be improved. Combined with image analysis technology, automatic monitoring and processing of the drone inspection process can be achieved, reducing the manual burden and improving the quality of inspection task completion.
[0097] Specifically, the drone needs to be equipped with corresponding cameras or sensors to collect image data in real time during the flight. These data can be in the form of video streams, image sequences, etc. The collected image data is preprocessed, including denoising, image enhancement, image registration, etc., to improve the subsequent image analysis effect. Computer vision technology, such as target detection algorithms (such as YOLO, SSD, Faster R-CNN, etc.), is used to analyze the image data and identify the targets in the image, including static and dynamic targets. For the identified targets, target tracking algorithms (such as Kalman filtering, multi-target tracking, etc.) are used to track the position and motion trajectory of the target in continuous frames to achieve continuous tracking of the target. By analyzing the motion characteristics of the target, it is determined whether the target is a dynamic obstacle. Common methods include speed estimation, trajectory prediction, etc. to identify the motion state of the target. According to the recognition results of dynamic obstacles, when a dynamic obstacle is detected on the drone path, an alarm can be issued to avoid collision or dangerous situations.
[0098] In some embodiments, after performing image analysis on the image data to determine whether there are dynamic obstacles on the optimal path of the UAV and obtaining dynamic detection results, the method further includes the following steps: obtaining obstacle information of the dynamic obstacle, wherein the obstacle information includes at least one or more of the position of the dynamic obstacle, the flight trajectory of the dynamic obstacle, and the overlap information of the dynamic obstacle with the optimal path; generating an obstacle avoidance route according to the obstacle information, and controlling the UAV to fly along the obstacle avoidance route; and controlling the UAV to return to the optimal path when the UAV has passed the dynamic obstacle.
[0099] In this scheme, obstacle information of dynamic obstacles is obtained, including the location of the obstacle, flight trajectory, and information on coincidence with the optimal path. This information helps to accurately understand the location and movement trajectory of the obstacle and provide a basis for obstacle avoidance. According to the obstacle information obtained, an obstacle avoidance route is generated using an obstacle avoidance algorithm. The obstacle avoidance route is a flight path that bypasses dynamic obstacles to ensure that the drone passes through the obstacle area safely. The drone is guided to fly along the generated obstacle avoidance route. The drone should adjust its heading and flight altitude according to the calculated obstacle avoidance route to avoid dynamic obstacles and ensure flight safety. After the drone successfully avoids the dynamic obstacle, the drone is controlled to return to the original optimal path for flight. This helps to maintain the efficiency and accuracy of the flight mission and avoid excessive detours. By generating an obstacle avoidance route and controlling the drone to avoid dynamic obstacles, the risk of collision can be effectively reduced and flight safety can be improved. By avoiding dynamic obstacles in a timely manner, the drone can continue to fly along the optimal path to ensure the successful completion of the mission. With the help of the algorithm, the obstacle avoidance route is automatically generated to achieve automatic obstacle avoidance of the drone, reduce manual intervention, and improve flight efficiency.
[0100] Specifically, the drone flies according to the planned route. When the imaging device or sensor device carried by the drone detects an obstacle on the path to point B at point A, it obtains the location information and shape characteristics of the obstacle, mathematically models the obstacle, and marks the obstacle-related information in the path grid of the three-dimensional model to perform short-term planning of the path algorithm. After passing the obstacle, the drone's track returns to the pre-planned path.
[0101] Specifically, Figure 5 As shown, tasks can be generated, parametric modeling can be performed, trajectory planning schemes can be set, multi-feature filter prediction targets can be combined with target features in three-dimensional modeling, inspection targets can be identified and inspection images can be taken. If there is occlusion, morphological prediction can be performed to determine the shape and flight trajectory of dynamic obstacles. If there is no occlusion, the inspection can be ended when the inspection is completed.
[0102] Specifically, the solution of the present application forms a parametric simulation technology of the inspection environment and UAV flight routes through the study of the data structure and parametric modeling method of UAV inspection routes and based on the parametric modeling of the inspection equipment.
[0103] Based on construction drawings, radar sensors and other equipment data, the location and geometric information of tracks, inspection equipment, obstacles, etc. are extracted and mapped into variable parameters to achieve high-precision, parametric modeling of the inspection area, providing data support for UAV inspection simulation and target identification.
[0104] In order to solve the problem of target occlusion during UAV inspection, multi-feature filters are fused to predict target morphology, which enhances the interpretability and robustness of the track parameterization model.
[0105] By designing three-dimensional parametric modeling of the trajectory, combining the three-dimensional model with the properties of inspection needs, drones, obstacles and other objects, combining the target morphology prediction method to simulate the problem of drones encountering obstacles and occlusion, and solving the parametrically modeled drone path planning model based on the improved adaptive pheromone adjustment ant colony algorithm, the life cycle of drone trajectory planning management is improved, the calculation amount of similar route planning is reduced, and it is possible to quickly provide drones with safe flight routes.
[0106] The embodiment of the present application also provides a path planning device for a drone. It should be noted that the path planning device for the drone of the embodiment of the present application can be used to execute the path planning method for a drone provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and those that have been described will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.
[0107] The following is an introduction to the path planning device for a drone provided in an embodiment of the present application.
[0108] Figure 6 is a structural block diagram of a path planning device for a drone according to an embodiment of the present application. Figure 6 As shown, the device comprises:
[0109] A first acquisition unit 10 is used to acquire relevant information of the UAV inspection, wherein the relevant information includes at least one or more of the inspection demand, the inspection environment information and the performance of the UAV, the inspection demand is the demand for inspecting the target object, the target object includes one or more of equipment, facilities and areas, the environmental information includes at least one or more of terrain, climate and vegetation, and the performance of the UAV includes at least one or more of flight speed, flight altitude, climb rate, descent rate and flight distance;
[0110] A first construction unit 20 is used to construct a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical related information obtained within a historical time period and a historical optimal path corresponding to the historical related information;
[0111] The first processing unit 30 is used to input the above-mentioned relevant information into the above-mentioned path planning model to obtain the optimal path corresponding to the above-mentioned relevant information;
[0112] The first control unit 40 is used to control the UAV to fly along the optimal path to inspect the target object.
[0113] Through this embodiment, an ant colony algorithm is used to plan the inspection path of the drone. The ant colony algorithm is adaptive and can adjust the path planning in real time according to environmental changes to adapt to different situations and requirements. The ant colony algorithm is a heuristic algorithm that can find a better path planning in a shorter time, thereby making the planned inspection path meet the inspection requirements and improving the efficiency of drone inspections.
[0114] During the specific implementation process, the above-mentioned device also includes a second acquisition unit and a classification unit. The second acquisition unit is used to obtain the spatial distribution feature information of the above-mentioned target object to be inspected based on the above-mentioned inspection requirements after obtaining the relevant information of the drone inspection, wherein the above-mentioned spatial distribution feature information at least includes one or more of position distribution information, morphological feature distribution information and topological relationship information, the above-mentioned position distribution information is the position of the above-mentioned target object, the above-mentioned morphological feature distribution information is the characteristics of the shape of the above-mentioned target object, and the above-mentioned topological relationship information is the distance relationship between multiple above-mentioned target objects; the classification unit is used to classify the above-mentioned target objects to be inspected according to the above-mentioned spatial distribution feature information to obtain a classification result.
[0115] In this solution, it is necessary to obtain the spatial distribution characteristic information of the target objects to be inspected, and classify the target objects based on this information. By analyzing this information, it can help to classify the target objects, so as to perform the inspection tasks more effectively. By classifying according to the characteristics of the target objects, a more reasonable inspection path and plan can be formulated according to the inspection requirements and spatial distribution characteristic information, thereby improving the efficiency and accuracy of the inspection. Classification can help identify different types of target objects, select appropriate inspection methods and parameters in a targeted manner, and improve the precision and accuracy of the inspection. By classifying the target objects, it can ensure that all types of target objects are fully inspected to avoid omissions or omissions.
[0116] During the specific implementation process, the above-mentioned device also includes a third acquisition unit, a fourth acquisition unit and a second construction unit. The third acquisition unit is used to obtain key track points after obtaining relevant information of the drone inspection, wherein the above-mentioned key track points are track points pre-set in the above-mentioned inspection requirements; the fourth acquisition unit is used to obtain obstacle avoidance points corresponding to static obstacles, wherein the above-mentioned obstacle avoidance points are the location points of static obstacles in the above-mentioned target object to be inspected; the second construction unit is used to construct a three-dimensional model based on at least the above-mentioned key track points and the above-mentioned obstacle avoidance points.
[0117] In this solution, before the drone inspection mission is carried out, it is necessary to obtain key track points and obstacle avoidance points of static obstacles, and build a three-dimensional model based on these points. Based on the key track points and obstacle avoidance points, a three-dimensional model can be built, that is, a model that presents the flight path, key positions and obstacle positions of the drone in three-dimensional space. By setting key track points and obstacle avoidance points and building a three-dimensional model, obstacles in flight can be effectively avoided, flight risks can be reduced, and flight safety can be improved. Based on the pre-set track points and obstacle avoidance points, the drone can fly along the planned path, improving flight efficiency and accuracy.
[0118] In the specific implementation process, the device further includes a fifth acquisition unit and a fusion unit. The fifth acquisition unit is used to obtain a feature map of the target object to be inspected before building a path planning model, wherein the feature map is a histogram based on point cloud features and color gradients; the fusion unit is used to filter and fuse the feature map according to a target formula to obtain a response map, wherein the response map is used to characterize the response degree of the filter in different feature maps, and the target formula is:
[0119] f′(x)=∑ i=1 α i ×f i (x),
[0120] f′(x) represents the above response graph, α i represents the weight of the i-th feature map above, f i(x) represents the i-th feature map mentioned above,
[0121]
[0122] k i represents the peak value of the ith filter response, k i =max(f i (x)), k j represents the peak value of j filter responses, k j =max(f j (x)), N represents the number of the above feature maps.
[0123] In this scheme, by acquiring feature maps and performing filter fusion, the characteristic information of the target object can be described more comprehensively and accurately, which helps to improve the accuracy of UAV recognition of target objects during inspections. Filter fusion can help remove noise and interference in the data, making the data clearer and more reliable, which is conducive to accurate analysis and processing of data. By fusing different types of feature information, the description of the target object from multiple aspects can be comprehensively considered, improving the comprehensiveness and accuracy of data analysis.
[0124] In some embodiments, the above-mentioned device also includes an initial change unit, a first determination unit and a sixth acquisition unit. The initial change unit is used to initialize the population and randomly place M ants before constructing the path planning model, where M≥2, and the route walked by each of the above-mentioned ants represents the inspection path of the above-mentioned drone; the first determination unit is used to use the above-mentioned relevant information to determine the next moving position of the above-mentioned ants; the sixth acquisition unit is used to obtain the optimal solution in the iterative process when all of the above-mentioned ants have reached the inspection end point, so as to obtain the above-mentioned optimal path.
[0125] In this scheme, an ant colony needs to be initialized and M ants are randomly placed at the starting position. Each ant represents the inspection path of the drone, that is, the ant simulates the flight trajectory of the drone by moving. By using relevant information, such as the distribution of target objects, environmental information, and drone performance, the next moving position of each ant can be determined. The ants choose the moving direction based on the pheromone concentration and heuristic information to find a better inspection path. After all ants have reached the inspection end point, the optimal solution in the iteration process can be obtained based on the movement path and effect of the ants in the iteration process, that is, the optimal path. This optimal path is obtained through continuous search and optimization by the ant colony algorithm, and is expected to provide the optimal flight path planning for drone inspection tasks. The ant colony algorithm has global search capabilities and can continuously search for better solutions in the search space, which helps to find the best drone inspection path. Because the ant colony algorithm has certain randomness and adaptability, it can effectively overcome the local optimal solution trap and improve the robustness of the algorithm. The ant colony algorithm can continuously adjust the path selection strategy according to the pheromone concentration and the movement effect of the ants found during the search process to achieve adaptive optimization of path planning.
[0126] During the actual inspection process, there may be changes in obstacles, as well as dynamic obstacles such as kites and birds. In some embodiments, the above-mentioned device also includes a second control unit and a second determination unit. The second control unit is used to input the above-mentioned relevant information into the above-mentioned path planning model to obtain the optimal path corresponding to the above-mentioned relevant information, and then control the above-mentioned UAV to fly according to the above-mentioned optimal path, and obtain the image data of the above-mentioned optimal path during the inspection process in real time; the second determination unit is used to perform image analysis on the above-mentioned image data, determine whether there are dynamic obstacles on the above-mentioned optimal path of the above-mentioned UAV, and obtain dynamic detection results.
[0127] In this solution, according to the optimal path planning, the drone is controlled to fly along the optimal path. The optimal path is optimized. Flying along the optimal path can effectively reduce flight time and energy consumption and improve the efficiency of the inspection task. During the flight, the image data on the flight path of the drone is obtained in real time. These image data can be obtained through the camera or sensor of the drone, and contain real-time information of the inspection environment. Image analysis is performed on the acquired image data, and computer vision technology is used to identify and detect whether there are dynamic obstacles on the optimal path. Dynamic obstacles may be moving vehicles, personnel or other obstacles, which pose a threat to the flight safety of the drone. Through image analysis, dynamic detection results can be obtained, that is, whether there are dynamic obstacles on the optimal path. According to the detection results, the flight path of the drone can be adjusted in time or other measures can be taken to ensure flight safety and complete the inspection task. Dynamic obstacles on the optimal path can timely detect possible flight obstacles, avoid collisions and accidents, and improve flight safety. Through image analysis technology, dynamic obstacles can be automatically identified, manual intervention can be reduced, and inspection efficiency and timeliness can be improved. Combined with image analysis technology, automatic monitoring and processing of the drone inspection process can be achieved, reducing the manual burden and improving the quality of inspection task completion.
[0128] In some embodiments, the device further includes a seventh acquisition unit, a third control unit and a fourth control unit, wherein the seventh acquisition unit is used to perform image analysis on the image data to determine whether there is a dynamic obstacle on the optimal path of the UAV, and obtain obstacle information of the dynamic obstacle after obtaining the dynamic detection result, wherein the obstacle information includes at least one or more of the position of the dynamic obstacle, the flight trajectory of the dynamic obstacle, and the overlap information of the dynamic obstacle with the optimal path; the third control unit is used to generate an obstacle avoidance route according to the obstacle information, and control the UAV to fly along the obstacle avoidance route; the fourth control unit is used to control the UAV to return to the optimal path to fly when the UAV has passed the dynamic obstacle.
[0129] In this scheme, obstacle information of dynamic obstacles is obtained, including the location of the obstacle, flight trajectory, and information on coincidence with the optimal path. This information helps to accurately understand the location and movement trajectory of the obstacle and provide a basis for obstacle avoidance. According to the obstacle information obtained, an obstacle avoidance route is generated using an obstacle avoidance algorithm. The obstacle avoidance route is a flight path that bypasses dynamic obstacles to ensure that the drone passes through the obstacle area safely. The drone is guided to fly along the generated obstacle avoidance route. The drone should adjust its heading and flight altitude according to the calculated obstacle avoidance route to avoid dynamic obstacles and ensure flight safety. After the drone successfully avoids the dynamic obstacle, the drone is controlled to return to the original optimal path for flight. This helps to maintain the efficiency and accuracy of the flight mission and avoid excessive detours. By generating an obstacle avoidance route and controlling the drone to avoid dynamic obstacles, the risk of collision can be effectively reduced and flight safety can be improved. By avoiding dynamic obstacles in a timely manner, the drone can continue to fly along the optimal path to ensure the successful completion of the mission. With the help of the algorithm, the obstacle avoidance route is automatically generated to achieve automatic obstacle avoidance of the drone, reduce manual intervention, and improve flight efficiency.
[0130] The above-mentioned path planning device for the UAV includes a processor and a memory, and the above-mentioned first acquisition unit, first construction unit, first processing unit and first control unit are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.
[0131] The processor includes a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters can be adjusted to solve the problem that the manual control of the path of the drone in the existing technology is time-consuming and labor-intensive, and the planned inspection path is difficult to meet the inspection requirements.
[0132] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0133] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the path planning method of the drone.
[0134] An embodiment of the present invention provides a processor, which is used to run a program, wherein the path planning method of the drone is executed when the program is run.
[0135] An embodiment of the present invention provides a device, the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the path planning method steps of the drone are implemented. The device in this article can be a server, a PC, a PAD, a mobile phone, etc.
[0136] A computer program product includes a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the path planning method for the drone described above in each embodiment of the present application are implemented.
[0137] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
[0138] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0139] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0140] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0142] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0143] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0144] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules 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 technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0145] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0146] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:
[0147] 1) The path planning method of the drone of the present application adopts an ant colony algorithm to plan the inspection path of the drone. The ant colony algorithm is adaptive and can adjust the path planning in real time according to environmental changes to adapt to different situations and requirements. The ant colony algorithm is a heuristic algorithm that can find a better path planning in a shorter time, thereby making the planned inspection path meet the inspection requirements and improving the efficiency of drone inspections.
[0148] 2) The path planning device of the drone of the present application adopts an ant colony algorithm to plan the inspection path of the drone. The ant colony algorithm is adaptive and can adjust the path planning in real time according to environmental changes to adapt to different situations and requirements. The ant colony algorithm is a heuristic algorithm that can find a better path planning in a shorter time, thereby making the planned inspection path meet the inspection requirements and improving the efficiency of drone inspections.
[0149] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A path planning method for an unmanned aerial vehicle, characterized in that: include: Obtaining relevant information about the inspection of the drone, wherein the relevant information includes at least one or more of inspection requirements, inspection environment information, and performance of the drone, wherein the inspection requirement is a requirement for inspecting a target object, wherein the target object includes one or more of equipment, facilities, and areas, wherein the environmental information includes at least one or more of terrain, climate, and vegetation, and wherein the performance of the drone includes at least one or more of flight speed, flight altitude, climb rate, descent rate, and flight distance; Constructing a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical related information obtained within a historical time period and a historical optimal path corresponding to the historical related information; Inputting the relevant information into the path planning model to obtain an optimal path corresponding to the relevant information; The drone is controlled to fly along the optimal path to inspect the target object.
2. The method according to claim 1, characterized in that After obtaining relevant information about the drone inspection, the method further includes: Based on the inspection requirements, spatial distribution feature information of the target object to be inspected is obtained, wherein the spatial distribution feature information includes at least one or more of position distribution information, morphological feature distribution information and topological relationship information, the position distribution information is the position of the target object, the morphological feature distribution information is the characteristics of the shape of the target object, and the topological relationship information is the distance relationship between multiple target objects; The target objects to be inspected are classified according to the spatial distribution feature information to obtain a classification result.
3. The method according to claim 1, characterized in that After obtaining relevant information about the drone inspection, the method further includes: Acquire key track points, wherein the key track points are track points pre-set in the inspection requirements; Obtaining an obstacle avoidance point corresponding to a static obstacle, wherein the obstacle avoidance point is a location point of a static obstacle in the target object to be inspected; A three-dimensional model is constructed based on at least the key track points and the obstacle avoidance points.
4. The method according to claim 1, characterized in that Before building the path planning model, the method further includes: Acquire a feature map of the target object to be inspected, wherein the feature map is a histogram based on point cloud features and color gradients; The feature graphs are filtered and fused according to the target formula to obtain a response graph, wherein the response graph is used to characterize the response degree of the filter in different feature graphs. The target formula is: f′(x)=∑ i=1 α i ×f i (x), f′(x) represents the response graph, α i represents the weight of the i-th feature map, f i (x) represents the i-th feature map, k i represents the peak value of the ith filter response, k i =max(f i (x)), k j represents the peak value of j filter responses, k j =max(f j (x)), N represents the number of feature maps.
5. The method according to claim 1, characterized in that Before building the path planning model, the method further includes: Initialize the population and randomly place M ants, M ≥ 2, where the route walked by each ant represents the inspection path of the drone; Using the relevant information, determining the next moving position of the ant; When all the ants have reached the inspection endpoint, the optimal solution in the iterative process is obtained to obtain the optimal path.
6. The method according to claim 1, characterized in that After inputting the relevant information into the path planning model to obtain the optimal path corresponding to the relevant information, the method further includes: Controlling the UAV to fly along the optimal path, and acquiring image data of the optimal path during the inspection process in real time; Perform image analysis on the image data to determine whether there are dynamic obstacles on the optimal path of the drone, and obtain a dynamic detection result.
7. The method according to claim 6, characterized in that After performing image analysis on the image data to determine whether there are dynamic obstacles on the optimal path of the drone and obtaining a dynamic detection result, the method further includes: Obtaining obstacle information of the dynamic obstacle, wherein the obstacle information includes at least one or more of the position of the dynamic obstacle, the flight trajectory of the dynamic obstacle, and the coincidence information of the dynamic obstacle with the optimal path; Generate an obstacle avoidance route according to the obstacle information, and control the UAV to fly along the obstacle avoidance route; When the UAV has passed the dynamic obstacle, the UAV is controlled to return to the optimal path for flight.
8. A path planning device for an unmanned aerial vehicle, characterized in that: include: A first acquisition unit is used to acquire relevant information of the UAV inspection, wherein the relevant information includes at least one or more of the inspection demand, the inspection environment information and the performance of the UAV, the inspection demand is the demand for inspecting the target object, the target object includes one or more of equipment, facilities and areas, the environmental information includes at least one or more of terrain, climate and vegetation, and the performance of the UAV includes at least one or more of flight speed, flight altitude, climb rate, descent rate and flight distance; A first construction unit is used to construct a path planning model, wherein the path planning model is obtained by training with an ant colony algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical related information obtained within a historical time period and a historical optimal path corresponding to the historical related information; A first processing unit, configured to input the relevant information into the path planning model to obtain an optimal path corresponding to the relevant information; The first control unit is used to control the UAV to fly along the optimal path to inspect the target object.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the path planning method for the drone described in any one of claims 1 to 7.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the path planning method for a drone described in any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Multi-UAV route planning and dynamic obstacle avoiding method based on combination of Voronoi ant colony algorithm with artificial potential field method
CN109521794A
Multi-rotor unmanned aerial vehicle inspection path planning method
CN111256703A
Power transmission line unmanned aerial vehicle inspection path planning and defect analysis cloud platform
CN114326825A
Multi-feature fusion target tracking and positioning method based on response peak value
CN114972437A
Unmanned aerial vehicle obstacle avoidance path planning method and device, computer equipment and storage medium
CN115686052A
Cited By
Multi-AGV path planning method and system, storage medium and product
CN119272963A
A multi-AGV path planning method, system, storage medium and product
CN119272963B
Unmanned aerial vehicle autonomous obstacle avoidance path planning method and device and storage medium
CN119596937A
Unmanned aerial vehicle air base station intelligent scheduling method based on low earth orbit satellite internet and related device
CN119647916A
Unmanned aerial vehicle path planning method and device, electronic equipment and storage medium
CN120333467A
