An unmanned aerial vehicle inspection planning method and system based on power failure features

Abstract

The application relates to a kind of unmanned aerial vehicle inspection planning methods and systems of power failure feature perception, it is related to route planning field, by obtaining the data when target power tower fails, dynamically adjusting preset safe flight fence and marking key points, constructing initial inspection route after random planning of inspection planning constraint, then based on key points and stopping point, the fitness of inspection is calculated, and then the optimal inspection route is obtained to control unmanned aerial vehicle to execute inspection, solve the technical problems that unmanned aerial vehicle cannot effectively perceive failure and dynamically adjust inspection route and safe flight fence when power inspection, leading to the technical problems of insufficient safety and effectiveness of inspection, realize the dynamic adaptation of unmanned aerial vehicle inspection route to failure, improve the safety and effectiveness of inspection.

Description

Technical Field

[0001] This invention relates to the field of route planning, and more particularly to a method and system for planning unmanned aerial vehicle (UAV) inspections based on power fault characteristics. Background Technology

[0002] With the rapid development of power systems and the continuous expansion of power grids, power poles, as important supporting structures for transmission lines, are crucial for ensuring the reliability of the entire power system through safe and stable operation. However, due to various factors such as the natural environment, equipment aging, and human error, power poles inevitably experience various faults during operation, such as pole deformation and component damage. If these faults are not detected and addressed in a timely manner, they may lead to more serious power accidents, or even large-scale power outages, causing significant impacts on the social economy and people's lives.

[0003] Traditional methods of inspecting power poles mainly rely on regular manual inspections. This method is not only labor-intensive and inefficient, but also greatly limited by natural conditions such as weather and terrain, making it difficult to achieve comprehensive and timely inspections. With the rapid development of drone technology, drone inspection, due to its advantages of high efficiency, flexibility, and low cost, is gradually being applied to the inspection of power poles. Drone inspection can quickly cover large areas, performing high-altitude, multi-angle photography and inspection of poles, effectively improving inspection efficiency and accuracy.

[0004] However, most existing drone inspection technologies remain at the stage of preset routes or manual control, lacking the ability to detect and dynamically adjust to power pole faults. When a power pole experiences a fault, such as deformation, the safe distance changes. Existing preset route inspection methods cannot adapt to this change, resulting in insufficient safe distance between the drone and the faulty pole during inspection, potentially leading to safety accidents. Furthermore, due to the lack of effective fault detection, drones cannot accurately identify faulty components and locations, leading to decreased inspection quality and an inability to promptly detect and address potential safety hazards. Summary of the Invention

[0005] This invention addresses the technical problem in existing technologies where unmanned aerial vehicles (UAVs) cannot effectively detect faults and dynamically adjust inspection routes and safe flight fences during power line inspections, resulting in insufficient inspection safety and effectiveness. The invention provides a UAV inspection planning method and system for power fault feature perception to solve this problem.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides a method for planning UAV inspections based on power fault feature perception. The method includes: acquiring fault data when a power fault occurs on a target power pole; integrating and adjusting a preset safe flight fence to obtain an adjusted safe flight fence sequence and marking multiple key points; constructing inspection planning constraints based on the adjusted safe flight fence sequence; randomly planning a first inspection route for inspecting the target power pole, wherein the first inspection route includes multiple first stopping points; calculating and analyzing the first inspection fitness of the first inspection route based on the multiple key points and the multiple first stopping points; optimizing the inspection route to obtain an optimal inspection route; and controlling the UAV to inspect the target power pole.

[0007] Secondly, the present invention provides a UAV inspection planning system for power fault feature perception. The system includes: a data acquisition module, used to acquire fault data when a power fault occurs on a target power pole, integrate and adjust a preset safe flight fence to obtain an adjusted safe flight fence sequence, and mark multiple key points; a route planning module, used to construct inspection planning constraints based on the adjusted safe flight fence sequence, and randomly plan a first inspection route for inspecting the target power pole, wherein the first inspection route includes multiple first stop points; a fitness calculation module, used to calculate and analyze the first inspection fitness of the first inspection route based on the multiple key points and the multiple first stop points; and an inspection control module, used to optimize the inspection route, obtain the optimal inspection route, and control the UAV to inspect the target power pole.

[0008] The beneficial effects of this invention are as follows: by acquiring data when a target power pole fails, dynamically adjusting the preset safe flight fence and marking key points, constructing inspection planning constraints, randomly planning the initial inspection route, and then calculating the inspection fitness based on key points and stopping points, thereby optimizing the optimal inspection route to control the UAV to perform the inspection, the UAV inspection route is dynamically adapted to the fault characteristics of the power pole, thus improving the safety and effectiveness of the inspection. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating a method for planning unmanned aerial vehicle (UAV) inspections based on power fault feature perception, provided by the present invention.

[0010] Figure 2 This is a schematic diagram of the structure of a UAV inspection and planning system for power fault feature perception provided by the present invention.

[0011] Figure labeling: Data acquisition module 11, route planning module 12, fitness calculation module 13, inspection control module 14. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0014] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0015] Example 1: like Figure 1 As shown, this embodiment of the invention provides a UAV inspection planning method based on power fault feature perception, applied to a UAV inspection planning system based on power fault feature perception. The method includes: S10: Obtain fault data when a power failure occurs on the target power pole, integrate and adjust the preset safe flight fence, obtain the adjusted safe flight fence sequence, and mark multiple key points.

[0016] For example, power poles are an important component of power transmission lines, primarily used to support and secure transmission conductors, ensuring the safe and stable transmission of electricity. Power poles are typically made of materials such as steel or concrete, possessing sufficient strength and stability to withstand the forces of conductors, wind, ice, and snow. Furthermore, through their rational layout and height settings, power poles ensure sufficient safe distances between conductors and between conductors and the ground, preventing electrical accidents.

[0017] In the context of drone power line inspection, a safe flight fence is a pre-defined virtual or physical flight boundary established to ensure that the drone does not collide with power poles and their associated facilities, such as conductors and insulators, during the inspection process. This boundary is typically set based on factors such as the actual location, height, shape of the power pole, and the surrounding environment. The main function of a safe flight fence is to ensure the flight safety of the drone during the inspection process, preventing it from colliding with power facilities due to flying into dangerous areas, resulting in equipment damage or drone crashes. Furthermore, by setting up a safe flight fence, the drone can be guided to conduct inspections along a predetermined route, avoiding unnecessary flights and hovering, thereby improving inspection efficiency.

[0018] Specifically, in this solution, during the inspection of power poles, when a power failure occurs on the target power pole, relevant fault data is first acquired. This data includes key characteristic information such as fault type, fault location, and fault severity. Then, based on this fault data, a pre-set safety flight fence around the power pole is integrated and adjusted. The pre-set safety flight fence is essentially a safe flight range constructed around the target pole, which can be divided by establishing a coordinate system around the target pole. The UAV flying outside this pre-set safety flight fence ensures that it maintains a safe distance from the power pole and its ancillary facilities, thereby guaranteeing flight safety. The integrated adjustment involves dynamically changing the boundaries, shape, and other parameters of the pre-set safety flight fence based on the fault data to adapt to the inspection needs under fault conditions, ultimately obtaining an adjusted safety flight fence sequence.

[0019] The adjusted safe flight fence sequence refers to a series of flight fence-related elements with a specific order and rules, modified and optimized to ensure flight safety. The sequence indicates that the settings and parameters of these flight fences are arranged or combined in a certain order; the adjustment signifies that the original safe flight fence sequence has been modified based on new requirements, environmental changes, and other factors.

[0020] Meanwhile, the adjusted safety flight fence is further analyzed, and several key points are marked. These key points are usually the coordinates of the largest deformation after a tower failure, such as the coordinates of the apex where a tower is severely tilted or bent due to a fault. These key points often correspond to the more important locations of the tower failure, providing important basis for subsequent inspection route planning and fault analysis. For example, if a power tower tilts due to strong winds, after the system obtains fault data such as the tilt angle and direction, it will adjust the safety flight fence accordingly to avoid the tilted area and mark the coordinates of the apex with the most severe tilt as a key point for focused attention during drone inspections.

[0021] S20: Based on the adjusted safe flight fence sequence, construct inspection planning constraints and randomly plan a first inspection route for inspecting the target power pole, wherein the first inspection route includes multiple first stop points.

[0022] Furthermore, after obtaining the adjusted safe flight fence sequence for the target power pole, these fence sequences will be combined into a single overall safe flight fence. Then, based on this overall safe flight fence, inspection planning constraints will be constructed, stipulating that the drone must fly outside the overall safe flight fence, thereby ensuring the drone's flight safety throughout the entire inspection process.

[0023] Subsequently, the system randomly plans the first inspection route for the target power pole. This route is not fixed but generated based on a random algorithm to explore different inspection path possibilities. During the planning process, the structural characteristics and fault features of the target power pole are fully considered, and multiple first stopping points are rationally set. These first stopping points are key nodes on the inspection route. When the UAV flies to these stopping points, it will hover to collect detailed images of the power pole using its onboard image acquisition equipment. For example, if a component on the target power pole is deformed due to a fault, the system will specifically set a stopping point near that area when randomly planning the inspection route, allowing the UAV to hover and collect images of that area, providing crucial information for subsequent fault analysis and processing.

[0024] S30: Calculate and analyze the first inspection adaptability of the first inspection route based on the multiple key points and multiple first stop points.

[0025] In detail, after completing the initial inspection route planning for the target power poles and determining multiple initial stop points, the initial inspection route is calculated and analyzed based on the marked key points to evaluate its suitability for the initial inspection. Suitability here is a comprehensive indicator that considers multiple aspects, among which the distance between key points and initial stop points is one of the important evaluation factors.

[0026] Specifically, the distance between each first stopping point and its corresponding key point is calculated and compared with a preset standard detection distance. Ideally, the closer the distance between the stopping point and the key point is to the standard detection distance, the more accurately the drone can approach the fault or important detection location during the inspection process, thus acquiring clearer and more effective image data, and resulting in higher detection quality. For example, if the standard detection distance is set to 5 meters, when the distance between a certain first stopping point and its corresponding key point is exactly 5 meters, it means that the drone can acquire images of the key point's area at the optimal distance when hovering at that stopping point, and the detection quality is relatively ideal.

[0027] In addition to distance, other factors that may affect inspection efficiency and quality are also considered, such as the total length of the inspection route and flight time. After assessing the suitability of the first inspection route, if the inspection time is found to be too long or there is other room for optimization, the system will make optimization adjustments. By replanning the route and adjusting the location of the stopping points, the inspection time will be reduced, the overall inspection efficiency will be improved, and more efficient and accurate power pole inspections will be achieved while ensuring the quality of the inspection.

[0028] S40: Optimize the inspection route, obtain the optimal inspection route, and control the drone to inspect the target power pole.

[0029] Specifically, after completing the fitness calculation and analysis of the first inspection route, the inspection route will be further optimized based on the fitness evaluation results. The optimization process involves comprehensively adjusting the position, order, and inspection path of multiple first stop points in the first inspection route. Using intelligent algorithms, such as genetic algorithms and particle swarm optimization, the route schemes that can further improve inspection efficiency and quality are explored while satisfying inspection planning constraints (such as safe flight fence restrictions).

[0030] Specific optimization goals include shortening the total inspection time, reducing the drone's flight distance, and ensuring the rational distribution of stop points near key points. For example, if an inspection path has redundancy or detours, the system will trim or replan it so that the drone can reach the next stop point more directly; if there are multiple stop points near a key point but their distance distribution is uneven, the positions of the stop points will be adjusted to more evenly cover the area where the key point is located, thereby improving the inspection quality.

[0031] After multiple rounds of iterative optimization, the optimal inspection route was finally obtained. This route, while meeting safe flight constraints, can complete the inspection of the target power poles in the shortest time and with the highest efficiency. Subsequently, the system will send control commands to the drone to guide it to perform inspection operations according to the optimal inspection route, ensuring that the drone can accurately reach each stopping point during flight and complete a comprehensive and efficient inspection of the target power poles.

[0032] In a preferred embodiment, fault data when a power failure occurs on the target power pole is acquired, and a preset safe flight fence is integrated and adjusted to obtain an adjusted safe flight fence sequence. Several key points are then marked, including: Acquire fault data when a power failure occurs on the target power pole, including power failure data and environmental data.

[0033] Obtain the preset safe flight fence of the target power pole, wherein the preset safe flight fence includes safe flight boundary coordinates.

[0034] The fault data and the preset safe flight fence are combined and input into multiple safe flight fence generators to generate multiple adjustable safe flight fences, which serve as an adjustable safe flight fence sequence.

[0035] Multiple key points are marked on multiple adjusted safe flight fences within the adjusted safe flight fence sequence to obtain multiple key points.

[0036] Optionally, when a power failure occurs on the target power pole, a data acquisition process is initiated to collect fault-related data, which includes two main categories: power failure data and environmental data. Power failure data includes characteristic information such as fault type (e.g., short circuit, open circuit), time of occurrence, and abnormal changes in fault current or voltage. Environmental data involves weather conditions at the time of the fault (e.g., wind speed, wind direction, rainfall), temperature, humidity, and other factors that may affect the drone's flight and the power pole's condition.

[0037] At the same time, the system acquires information about the pre-set safe flight fence of the target power pole. This pre-set safe flight fence is defined by safe flight boundary coordinates, which clearly delineate the safe flight range that the drone should maintain under normal circumstances, so as to ensure that the drone will not collide with the power pole and its ancillary facilities during the inspection.

[0038] The system combines collected fault data with preset safe flight fence information and inputs them into multiple safe flight fence generators. These generators, built based on different algorithms or models, process and analyze the input data to generate multiple adjusted safe flight fences, forming an adjusted safe flight fence sequence. Each adjusted safe flight fence is specifically tailored to the fault and environmental data to meet inspection requirements under fault conditions.

[0039] Finally, key points are marked for multiple safety flight fence adjustments within the safety flight fence sequence. These key points are typically locations where the tower has undergone significant deformation after a fault, exhibiting obvious fault characteristics, or areas requiring focused attention and data collection during inspections. For example, if a fault causes a section of the tower to tilt severely, the system will mark the apex or key turning point of that tilted section as a key point, enabling the UAV to accurately locate and collect relevant images or data during inspections, providing strong support for subsequent fault analysis and processing.

[0040] In a preferred embodiment, the construction steps of the plurality of safe flight fence generators include: Within the inspection data of power poles experiencing power failures, a set of preset safe flight fences and a set of sample failure data are collected. Then, a set of sample adjusted safe flight fences is constructed after different sample failure data, resulting in a set of sample adjusted safe flight fences.

[0041] Within the sample preset safe flight fence set, sample fault data set, and sample adjusted safe flight fence set, data is divided into a preset proportion to obtain the first safe flight fence generation training data. This process is repeated to obtain the Kth safe flight fence generation training data, where K is a positive integer.

[0042] We employ deep learning to construct a network structure of K safe flight fence generators.

[0043] Training data is generated sequentially from the first safe flight fence to the Kth safe flight fence. The network parameters of the K safe flight fence generators are trained and optimized respectively, and the system is completed after the test converges.

[0044] Specifically, in constructing multiple safe flight fence generators, it is necessary to collect a key sample set from inspection data of power poles experiencing power failures. A pre-set safe flight fence set is collected, containing the coordinate information of the pre-set safe flight boundaries for power poles under different scenarios. Simultaneously, a sample fault data set is collected, covering various types of power failures and environmental parameters at the time of the failure.

[0045] Furthermore, based on the sample preset safe flight fence and sample fault data, sample adjustment safe flight fences corresponding to different sample fault data are further constructed. For example, when the fault type is short circuit and accompanied by strong wind environment, the preset safe flight fence is adjusted according to the fault characteristics and environmental factors to obtain the corresponding sample adjustment safe flight fence, and finally form a set of sample adjustment safe flight fences.

[0046] Subsequently, the data is divided into three sets: a pre-set safe flight fence set, a sample fault data set, and a sample adjusted safe flight fence set. According to a preset ratio, for example, 70% of the data is used as training data, corresponding portions are extracted from each of the three sets to construct the first safe flight fence training data. Continuing to divide the data according to different ratios yields the second safe flight fence training data, and so on, up to the Kth safe flight fence training data (K being a positive integer). This batch-based data division method helps to gradually optimize the generator's performance.

[0047] After the data is prepared, a network structure of K safety flight fence generators is constructed using deep learning technology. Deep learning models can handle complex nonlinear relationships and are suitable for generating adjusted safety flight fences based on fault data and preset safety flight fences. The network parameters of each of the K safety flight fence generators are trained and optimized sequentially using training data from the first to the Kth safety flight fence generation. During training, the network parameters are continuously adjusted so that the generators can accurately generate adjusted safety flight fences that conform to the actual situation based on the input preset safety flight fences and fault data. When the generators reach convergence on the test data, i.e., the output results are stable and meet expectations, it indicates that the K safety flight fence generators have been successfully constructed and are ready for practical application. For example, when facing new power pole faults, these generators can quickly generate accurate adjusted safety flight fences based on real-time acquired fault data and preset safety flight fences, providing safety assurance for drone inspections.

[0048] In a preferred embodiment, key points are marked on multiple adjusted safe flight fences within the adjusted safe flight fence sequence to obtain multiple key points, including: Within each adjusted safe flight fence, the distance between each adjusted safe flight boundary coordinate and the preset safe flight fence is calculated, and the adjusted safe flight boundary coordinate with the largest distance is selected as the key point.

[0049] The calculation process yields multiple key points within the adjusted safe flight fence sequence.

[0050] In detail, during the process of marking key points on multiple adjusted safety flight fences within a sequence to obtain multiple key points, calculations are required for each adjusted safety flight fence in the sequence. Specifically, the calculations involve determining the distance between each adjusted safety flight boundary coordinate within that fence and the preset safety flight fence. For example, assuming an adjusted safety flight fence has three adjusted safety flight boundary coordinates A, B, and C, the distances between A and the preset safety flight fence, B and C, and so on, need to be calculated separately. After completing the distance calculations, the adjusted safety flight boundary coordinate with the largest distance is selected from the calculated distances and identified as the key point of the adjusted safety flight fence. The key point with the largest distance represents the location within the adjusted safety flight fence where the deformation is greatest after a tower failure. This facilitates the inspection and observation of the location with the greatest deformation of the power tower, improving the effectiveness of inspection and observation. Following the same method, the above operations are performed on each adjusted safety flight fence in the sequence, ultimately calculating and processing to obtain multiple key points corresponding to multiple adjusted safety flight fences within the sequence.

[0051] In a preferred embodiment, the inspection planning constraints are constructed based on the adjusted safe flight fence sequence, and a first inspection route for inspecting the target power pole is randomly planned, including: Obtain the union of the adjusted safe flight fence sequences to obtain the fused safe flight fence, and construct inspection planning constraints, wherein the inspection planning constraints include that the inspection route does not enter the fused safe flight fence.

[0052] A first inspection route that satisfies the inspection planning constraints is generated by random planning, wherein the first inspection route is used to inspect the target power pole and includes multiple first stop points.

[0053] Furthermore, a union operation is performed on the obtained adjusted safe flight fence sequence. For example, assuming there are three fences in the adjusted safe flight fence sequence, namely fence 1, fence 2, and fence 3, the boundary ranges of these three fences are integrated, and the overlapping parts are removed to obtain a unified fused safe flight fence that covers all fence boundaries.

[0054] Furthermore, inspection planning constraints are constructed based on the integrated safe flight fence. These constraints clearly stipulate that inspection routes must not enter the area defined by the integrated safe flight fence. This is an important prerequisite for ensuring the safety of inspection flights and avoiding entry into dangerous or restricted areas.

[0055] Subsequently, a stochastic programming algorithm is used to generate a first inspection route for inspecting the target power poles, while satisfying the inspection planning constraints. For example, if the target power poles are distributed in a specific area, the stochastic programming algorithm will comprehensively consider factors such as pole location and incorporate safe flight fence boundaries to plan a route that meets the constraints and effectively covers the target power poles. The first inspection route includes multiple first stop points, which can be set according to the actual distribution of power poles and inspection requirements to ensure that a comprehensive inspection of each target power pole can be carried out.

[0056] The significance of stochastic route planning lies in its ability to efficiently explore feasible solution space and quickly generate the first inspection route that meets actual needs when facing uncertainties such as the complex distribution of target power poles and the irregular boundaries of the integrated safety flight fence in the inspection planning. This is achieved through random sampling and algorithm iteration, while satisfying the inspection planning constraints (i.e., the inspection route does not enter the integrated safety flight fence). This avoids getting trapped in local optima, effectively balances computational efficiency and solution quality, and provides a reliable foundation for the efficient and safe execution of subsequent inspection tasks.

[0057] In a preferred embodiment, the first inspection adaptability of the first inspection route is calculated and analyzed based on the plurality of key points and the plurality of first stop points, including: The first inspection time is obtained when the drone performs flight inspections according to the first inspection route, wherein the first inspection time includes the flight time and the dwell time at the plurality of first stop points.

[0058] Calculate the ratio of the preset inspection time to the first inspection time to obtain the first-time adaptability.

[0059] Based on the multiple key points and multiple first stop points, calculate the first detection adaptability of the first inspection route.

[0060] The first inspection fitness is calculated based on the first time fitness and the first detection fitness.

[0061] Preferably, to calculate and analyze the first inspection fitness of the first inspection route, it is necessary to obtain the first inspection time of the UAV flying along the first inspection route. This time includes the flight time and the dwell time at multiple first stop points. For example, if the UAV takes 2 hours to fly and the total dwell time at each first stop point is 1 hour, then the first inspection time is 3 hours.

[0062] Next, the ratio of the preset inspection time to the first inspection time is calculated to obtain the first-time adaptability. Assuming the preset inspection time is 4 hours and the first inspection time is 3 hours, the ratio is 3 / 4. This value reflects the inspection time efficiency; the larger the ratio, the more efficient the inspection is in the time dimension.

[0063] Simultaneously, based on multiple key points and multiple first stop points, the first detection fitness of the first inspection route is calculated. Key points represent important boundary positions for safe flight, while first stop points are associated with the detection of target power poles. By analyzing the spatial relationship distance factors between key points and stop points, the rationality of the inspection route in terms of the detection target can be quantitatively evaluated, thereby obtaining the first detection fitness.

[0064] The first inspection fitness is calculated by combining the first time fitness and the first detection fitness. The first time fitness focuses on inspection time efficiency, reflecting the completion of the inspection task in the time dimension; the first detection fitness focuses on the rationality of the inspection route for target detection, reflecting the performance of the inspection task in terms of detection effectiveness. By assigning different weights to both and combining them to calculate the first inspection fitness, the overall performance of the inspection route can be comprehensively measured. The specific weight configuration method can be set according to the actual situation and is not specifically limited here.

[0065] The first inspection adaptability serves as a comprehensive indicator for evaluating the quality of the first inspection route. It provides a clear basis for subsequent route optimization, helps improve inspection efficiency, ensures flight safety, and enhances inspection quality, ensuring that inspection tasks can be completed efficiently and accurately.

[0066] In a preferred embodiment, the first detection fitness of the first inspection route is calculated based on the plurality of key points and the plurality of first stop points, including: Obtain the standard detection distance.

[0067] Calculate the distances between the multiple key points and the multiple first stopping points to obtain multiple first actual detection distances.

[0068] Calculate the absolute deviation coefficients between the plurality of first actual detection distances and the standard detection distances, and calculate the plurality of detection standard approximation coefficients.

[0069] The mean of the approximation coefficients of the multiple detection standards is calculated to obtain the first detection fitness.

[0070] Specifically, when calculating the first detection fitness, a standard detection distance is first obtained. This distance is a pre-set benchmark distance used to measure the reasonableness of the detection effect, for example, set to 50 meters. Next, the distances between multiple keypoints and multiple first stopping points are calculated to obtain multiple first actual detection distances. Assuming there is one keypoint and one first stopping point, the actual distance between them, calculated using spatial coordinates, is 60 meters; this is one of the first actual detection distances.

[0071] Then, the absolute deviation coefficients between the first actual detection distance and the standard detection distance are calculated separately. The calculation method is to divide the absolute difference between the first actual detection distance and the standard detection distance by the standard detection distance. Taking the above example, the absolute difference is 60-50=10 meters, and the absolute deviation coefficient is 10÷50=0.2.

[0072] Based on this, the detection standard approximation coefficient is calculated, which is 1 minus the absolute deviation coefficient, i.e., 1 - 0.2 = 0.8. The larger this value is, the closer the first actual detection distance is to the standard detection distance, and the better the detection effect. The average value of multiple detection standard approximation coefficients calculated for all key points and the first stopping point is the first detection fitness.

[0073] For example, if there are 5 key points combined with the first stop point, the approximation coefficients of the 5 detection standards are calculated to be 0.8, 0.85, 0.75, 0.9, and 0.82, respectively. The average value is (0.8 + 0.85 + 0.75 + 0.9 + 0.82) ÷ 5 = 0.824, which is the first detection fitness.

[0074] The first detection fitness, as a quantitative indicator, can intuitively reflect the rationality of the inspection route in terms of the inspection target, provide a clear basis for the subsequent optimization and adjustment of the inspection route, help improve the inspection effect, and ensure that the inspection of the target power poles is more accurate and efficient.

[0075] The UAV inspection planning method for power fault feature perception provided in this invention has at least the following technical effects: 1. By acquiring fault data when a power failure occurs on a target power pole, and combining it with a preset safe flight fence, the system dynamically integrates and adjusts to generate an adjusted safe flight fence sequence, and further marks key points. This allows for flexible adjustment of the drone's safe flight range based on real-time fault data and environmental changes, ensuring that drone inspections are both safe and efficient in complex and ever-changing power failure scenarios, effectively improving the safety and adaptability of inspections.

[0076] 2. Multiple safe flight fence generators are employed, based on deep learning technology and trained and optimized using a large amount of sample data. This constructs a network model that can intelligently generate and adjust safe flight fences according to different fault data, realizing intelligent planning of inspection routes. It can quickly generate the optimal inspection route that meets safe flight constraints based on fault characteristics and environmental conditions, greatly improving the planning efficiency and accuracy of inspection routes.

[0077] 3. By comprehensively considering the time adaptability and detection adaptability of the inspection, the generated inspection route is comprehensively evaluated, and the inspection route is optimized accordingly to obtain the optimal inspection route. This not only focuses on the time efficiency of the inspection, but also emphasizes the detection quality of the inspection, ensuring that the UAV can comprehensively and accurately perceive the characteristics of power faults during the inspection process. This provides strong support for subsequent fault diagnosis and repair, and significantly improves the overall efficiency of UAV inspection for power fault characteristic perception.

[0078] Example 2: like Figure 2 As shown, based on the same inventive concept as the UAV inspection planning method for power fault feature perception provided in Embodiment 1, this embodiment of the invention also provides a UAV inspection planning system for power fault feature perception, the system comprising: The data acquisition module 11 is used to acquire fault data when the target power pole experiences a power failure, integrate and adjust the preset safe flight fence, obtain the adjusted safe flight fence sequence, and mark multiple key points.

[0079] The route planning module 12 is used to construct inspection planning constraints based on the adjusted safe flight fence sequence, and randomly plan a first inspection route for inspecting the target power pole, wherein the first inspection route includes multiple first stop points.

[0080] The fitness calculation module 13 is used to calculate and analyze the first inspection fitness of the first inspection route based on the multiple key points and multiple first stop points.

[0081] The inspection control module 14 is used to optimize the inspection route, obtain the optimal inspection route, and control the drone to inspect the target power pole.

[0082] Furthermore, the data acquisition module 11 is also used to perform the following steps: The process involves: acquiring fault data when a power failure occurs on the target power pole, including power failure data and environmental data; acquiring a preset safe flight fence for the target power pole, including safe flight boundary coordinates; combining the fault data and the preset safe flight fence, inputting them into multiple safe flight fence generators to generate multiple adjustable safe flight fences, forming an adjustable safe flight fence sequence; and marking key points on the multiple adjustable safe flight fences within the adjustable safe flight fence sequence to obtain multiple key points.

[0083] Furthermore, the data acquisition module 11 is also used to perform the following steps: Within the inspection data of power poles experiencing power faults, a set of preset safe flight fences and a set of sample fault data are collected. Adjusted safe flight fences are then constructed based on different fault data to obtain a set of adjusted safe flight fences. Data is divided into preset proportions within these sets to obtain the first set of safe flight fence generation training data. This division continues to obtain the Kth set of safe flight fence generation training data, where K is a positive integer. Deep learning is used to construct a network structure for K safe flight fence generators. The network parameters of each of the K generators are trained and optimized sequentially using the first to the Kth set of safe flight fence generation training data. The network is completed after testing and convergence.

[0084] Furthermore, the data acquisition module 11 is also used to perform the following steps: Within each adjusted safe flight fence, the distance between each adjusted safe flight boundary coordinate and the preset safe flight fence is calculated, and the adjusted safe flight boundary coordinate with the largest distance is selected as the key point; the calculation process is used to obtain multiple key points of multiple adjusted safe flight fences within the adjusted safe flight fence sequence.

[0085] Furthermore, the route planning module 12 is also used to perform the following steps: Obtain the union of the adjusted safe flight fence sequence to obtain the fused safe flight fence, and construct inspection planning constraints, wherein the inspection planning constraints include that the inspection route does not enter the fused safe flight fence; randomly plan and generate a first inspection route that satisfies the inspection planning constraints, wherein the first inspection route is used to inspect the target power pole and includes multiple first stop points.

[0086] Furthermore, the fitness calculation module 13 is also used to perform the following steps: The first inspection time of the UAV during flight inspection along the first inspection route is obtained, wherein the first inspection time includes flight time and dwell time at the plurality of first dwell points; the ratio of the preset inspection time to the first inspection time is calculated to obtain the first time fitness; the first detection fitness of the first inspection route is calculated based on the plurality of key points and the plurality of first dwell points; the first inspection fitness is calculated based on the first time fitness and the first detection fitness.

[0087] Furthermore, the fitness calculation module 13 is also used to perform the following steps: Obtain the standard detection distance; calculate the distance between the multiple key points and the multiple first stopping points to obtain multiple first actual detection distances; calculate the absolute deviation coefficient between the multiple first actual detection distances and the standard detection distance, and calculate multiple detection standard approximation coefficients; calculate the mean of the multiple detection standard approximation coefficients to obtain the first detection fitness.

[0088] Through the foregoing detailed description of a UAV inspection planning method for power fault feature perception, those skilled in the art can clearly understand that this embodiment is a UAV inspection planning system for power fault feature perception. As the system disclosed in the embodiment corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to in the method section.

[0089] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for planning UAV inspections based on power fault characteristic perception, characterized in that, A drone inspection planning system for power fault characteristic perception, the method comprising: The system acquires fault data when a power outage occurs on the target power pole, integrates and adjusts the preset safe flight fence, obtains the adjusted safe flight fence sequence, and marks several key points, including: Acquire fault data when a power failure occurs on the target power pole, including power failure data and environmental data; Obtain the preset safe flight fence of the target power pole, wherein the preset safe flight fence includes safe flight boundary coordinates; The fault data and the preset safe flight fence are combined and input into multiple safe flight fence generators to generate multiple adjustable safe flight fences, which are used as an adjustable safe flight fence sequence. Key points were marked on multiple adjusted safe flight fences within the adjusted safe flight fence sequence to obtain multiple key points, including: Within each adjusted safe flight fence, the distance between each adjusted safe flight boundary coordinate and the preset safe flight fence is calculated, and the adjusted safe flight boundary coordinate with the largest distance is selected as the key point; The calculation process yields multiple key points within the adjusted safe flight fence sequence. Based on the adjusted safe flight fence sequence, an inspection planning constraint is constructed, and a first inspection route for inspecting the target power pole is randomly planned, wherein the first inspection route includes multiple first stop points; Based on the multiple key points and multiple first stop points, the first inspection adaptability of the first inspection route is calculated and analyzed. Optimize the inspection route to obtain the optimal inspection route, and control the drone to inspect the target power pole.

2. The UAV inspection planning method for power fault feature perception according to claim 1, characterized in that, The construction steps for the multiple safe flight fence generators include: Within the inspection data of power poles experiencing power failures, a set of preset safe flight fences and a set of sample failure data are collected. Then, a set of sample adjusted safe flight fences is constructed after different sample failure data to obtain a set of sample adjusted safe flight fences. Within the sample preset safe flight fence set, the sample fault data set, and the sample adjusted safe flight fence set, data is divided into a preset proportion to obtain the first safe flight fence generation training data. The data is then further divided to obtain the Kth safe flight fence generation training data, where K is a positive integer. We employ deep learning to construct a network structure of K safe flight fence generators; Training data is generated sequentially from the first safe flight fence to the Kth safe flight fence. The network parameters of the K safe flight fence generators are trained and optimized respectively, and the system is completed after the test converges.

3. The UAV inspection planning method for power fault feature perception according to claim 1, characterized in that, Based on the adjusted safe flight fence sequence, inspection planning constraints are constructed, and a first inspection route for inspecting the target power pole is randomly planned, including: Obtain the union of the adjusted safe flight fence sequences to obtain the fused safe flight fence, and construct inspection planning constraints, wherein the inspection planning constraints include that the inspection route does not enter the fused safe flight fence; A first inspection route that satisfies the inspection planning constraints is generated by random planning, wherein the first inspection route is used to inspect the target power pole and includes multiple first stop points.

4. The UAV inspection planning method for power fault feature perception according to claim 1, characterized in that, Based on the aforementioned multiple key points and multiple first stop points, the first inspection adaptability of the first inspection route is calculated and analyzed, including: The first inspection time is obtained when the drone performs flight inspection according to the first inspection route, wherein the first inspection time includes the flight time and the dwell time at the plurality of first stop points; Calculate the ratio of the preset inspection time to the first inspection time to obtain the first-time adaptability; Based on the multiple key points and multiple first stop points, calculate the first detection adaptability of the first inspection route; The first inspection fitness is calculated based on the first time fitness and the first detection fitness.

5. The UAV inspection planning method for power fault characteristic perception according to claim 4, characterized in that, Based on the multiple key points and multiple first stop points, the first detection fitness of the first inspection route is calculated, including: Obtain the standard detection distance; Calculate the distances between the multiple key points and the multiple first stopping points to obtain multiple first actual detection distances; Calculate the absolute deviation coefficients between the plurality of first actual detection distances and the standard detection distances respectively, and calculate the plurality of detection standard approximation coefficients; The mean of the approximation coefficients of the multiple detection standards is calculated to obtain the first detection fitness.

6. A UAV inspection planning system for power fault characteristic perception, characterized in that, The system is used to implement the UAV inspection planning method for power fault feature perception according to any one of claims 1-5, the system comprising: The data acquisition module is used to acquire fault data when a power failure occurs on the target power pole, integrate and adjust the preset safe flight fence, obtain the adjusted safe flight fence sequence, and mark multiple key points, including: Acquire fault data when a power failure occurs on the target power pole, including power failure data and environmental data; Obtain the preset safe flight fence of the target power pole, wherein the preset safe flight fence includes safe flight boundary coordinates; The fault data and the preset safe flight fence are combined and input into multiple safe flight fence generators to generate multiple adjustable safe flight fences, which are used as an adjustable safe flight fence sequence. Key points were marked on multiple adjusted safe flight fences within the adjusted safe flight fence sequence to obtain multiple key points, including: Within each adjusted safe flight fence, the distance between each adjusted safe flight boundary coordinate and the preset safe flight fence is calculated, and the adjusted safe flight boundary coordinate with the largest distance is selected as the key point; The calculation process yields multiple key points within the adjusted safe flight fence sequence. The route planning module is used to construct inspection planning constraints based on the adjusted safe flight fence sequence, and randomly plan a first inspection route for inspecting the target power pole, wherein the first inspection route includes multiple first stop points; The fitness calculation module is used to calculate and analyze the first inspection fitness of the first inspection route based on the multiple key points and multiple first stop points. The inspection control module is used to optimize the inspection route, obtain the optimal inspection route, and control the drone to inspect the target power pole.

7. The UAV inspection planning system for power fault characteristic perception according to claim 6, characterized in that, The construction steps for the multiple safe flight fence generators include: Within the inspection data of power poles experiencing power failures, a set of preset safe flight fences and a set of sample failure data are collected. Then, a set of sample adjusted safe flight fences is constructed after different sample failure data to obtain a set of sample adjusted safe flight fences. Within the sample preset safe flight fence set, the sample fault data set, and the sample adjusted safe flight fence set, data is divided into a preset proportion to obtain the first safe flight fence generation training data. The data is then further divided to obtain the Kth safe flight fence generation training data, where K is a positive integer. We employ deep learning to construct a network structure of K safe flight fence generators; Training data is generated sequentially from the first safe flight fence to the Kth safe flight fence. The network parameters of the K safe flight fence generators are trained and optimized respectively, and the system is completed after the test converges.

Images