Unmanned aerial vehicle inspection area distribution method and system for draught fan and medium

Abstract

The invention relates to an unmanned aerial vehicle inspection area distribution method and system for draught fans and a medium, and the method comprises the steps: carrying out the clustering processing of feature parameters of all draught fans in a wind power plant, and obtaining a multi-class draught fan area set. According to the fan area set and the unmanned aerial vehicle starting point position set, an inspection duration model is constructed, and the inspection duration model determines the inspection duration through the inspection path. Based on the inspection duration model, performing inspection area distribution on the unmanned aerial vehicle through a Vouger method, and obtaining a distribution result of the inspection area of the unmanned aerial vehicle; wherein the Vouger method is used for distributing the unmanned aerial vehicle to inspect the fan task according to the difference characteristics of the inspection path duration. The problem of low inspection efficiency caused by unreasonable unmanned aerial vehicle inspection area distribution in related technologies is solved.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) inspection technology, and in particular to a method, system and medium for allocating UAV inspection areas for wind turbines. Background Technology

[0002] As wind farms continue to expand, the inspection of wind turbines has become increasingly important. Traditional inspection methods mainly rely on manual labor or fixed monitoring equipment, which suffer from low efficiency, high cost, and limited coverage. In recent years, the rapid development of drone technology has provided a new solution for wind turbine inspection.

[0003] However, existing methods for allocating inspection areas for wind turbines using drones are usually heuristic algorithms or greedy strategies, which can lead to unreasonable drone inspection area allocation and low inspection efficiency. Summary of the Invention

[0004] This application provides a method, system, and medium for allocating inspection areas for wind turbines using unmanned aerial vehicles (UAVs), in order to at least solve the problem of low inspection efficiency caused by unreasonable allocation of inspection areas in related technologies.

[0005] In a first aspect, embodiments of this application provide a method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines, the method comprising:

[0006] Clustering is performed on the characteristic parameters of each wind turbine in the wind farm to obtain a multi-category set of wind turbine regions.

[0007] Based on the set of wind turbine areas and the set of drone starting locations, the inspection path is obtained, and an inspection duration model is constructed using the inspection path and the supply and demand relationship.

[0008] Based on the inspection duration model, the inspection area of ​​the UAV is allocated using the Vogel method to obtain the allocation result of the UAV inspection area; wherein, the Vogel method is used to allocate the UAV inspection of wind turbines according to the difference characteristics of the inspection path duration.

[0009] In one embodiment, before clustering the characteristic parameters of each wind turbine in the wind farm, the method further includes:

[0010] The latitude and longitude coordinates and tower height of each wind turbine are obtained by using the drone's positioning sensors and altitude measurement device;

[0011] Based on preset data verification rules, outlier detection and correction are performed on the latitude and longitude coordinates and the tower height to obtain the characteristic parameters of each wind turbine after outlier detection and correction.

[0012] In one embodiment, the step of clustering the characteristic parameters of each wind turbine in the wind farm to obtain a multi-category set of wind turbine regions includes:

[0013] The latitude and longitude coordinates and tower height in the feature parameters are processed by zero-mean normalization to obtain normalized feature data;

[0014] Based on the normalized feature data, the number of target clusters is determined using interval statistics;

[0015] Based on the target number of clusters, wind turbines are clustered using a clustering algorithm, and the cluster centers are updated until the convergence condition is met to obtain a set of wind turbine regions of multiple categories.

[0016] In one embodiment, the step of clustering wind turbines based on the target number of clusters using a clustering algorithm and updating the cluster centers until a convergence condition is met to obtain a set of wind turbine regions of multiple categories includes:

[0017] Calculate the Euclidean distance between all adjacent wind turbines to obtain the average distance, and select the sample with the minimum distance from the average distance as the initial cluster center;

[0018] Each wind turbine is assigned to the nearest initial cluster center;

[0019] Based on the assigned clustered fans, the cluster centers are updated by calculating the mean of the characteristic data of each clustered fan;

[0020] When the change in the cluster center is less than a preset threshold, a target cluster center is determined and the target cluster center is used as the set of wind turbine regions.

[0021] In one embodiment, constructing the inspection duration model based on the set of wind turbine regions and the set of drone starting positions includes:

[0022] Based on the spatial relationship between the starting position of each UAV and the wind turbines in the wind turbine area, the inspection path of the UAV to each wind turbine area is calculated by the route planning algorithm, and the initial inspection time is determined based on the inspection path.

[0023] The available resources of drones are used as the number of supply times, and the number of inspections are used as the number of demand times. A supply and demand relationship is established based on the number of supply times and the number of demand times.

[0024] Based on the initial inspection duration and the number of inspections, an inspection duration model is constructed through supply and demand.

[0025] In one embodiment, the step of allocating inspection areas for the UAV based on the inspection duration model and using the Vogel method to obtain the UAV inspection area allocation result includes:

[0026] Based on the inspection duration determined in the inspection duration model, calculate the difference characteristics between the set of drone starting positions and the set of wind turbine areas;

[0027] Based on the supply and demand relationship in the inspection duration model, the drone will be preferentially assigned to the inspection area with the greatest difference feature;

[0028] Update the supply and demand relationship, prioritize the allocation of the drones to the drone inspection paths with the greatest differences, until the inspection requirements are met, and obtain the allocation results of the drone inspection areas.

[0029] In one embodiment, calculating the difference features between the set of drone starting positions and the set of wind turbine areas based on the inspection duration determined in the inspection duration model includes:

[0030] Based on the inspection time from the starting position of each drone to each wind turbine area, calculate the first difference between the next hour length and the minimum time for the row, and use it as the row penalty number.

[0031] Based on the inspection time from the starting position of each UAV to each wind turbine area, calculate the second difference between the second hour length and the minimum hour length corresponding to the column, and use it as the column penalty number.

[0032] The multiple row penalty numbers and the multiple column penalty numbers are used as difference features.

[0033] In one embodiment, prioritizing the assignment of the drone to the inspection area with the greatest difference feature includes:

[0034] Among the multiple row penalty numbers and multiple column penalty numbers, obtain the maximum row penalty number or the maximum column penalty number;

[0035] The maximum row penalty or the maximum column penalty is used as the maximum difference feature. If the maximum difference feature is the maximum row penalty, the starting position of the UAV is determined according to the maximum row penalty, and the wind turbine area is determined according to the minimum inspection time in the row corresponding to the maximum row penalty. If the maximum difference feature is the maximum column penalty, the wind turbine area is determined according to the maximum number of rows and columns, and the starting position of the UAV is determined according to the minimum inspection time in the column corresponding to the maximum column penalty.

[0036] The available resources for the drone are determined based on its starting point location, and the number of inspections is determined based on the wind turbine area.

[0037] Based on the supply and demand relationship of the available resources and the number of inspections, the available resources of the UAV are allocated to the wind turbine area.

[0038] Secondly, embodiments of this application provide a UAV inspection area allocation system for wind turbines, comprising: a clustering module, a model construction module, and an allocation module; wherein,

[0039] The clustering module is used to perform clustering processing on the characteristic parameters of each wind turbine in the wind farm to obtain a set of wind turbine regions of multiple categories.

[0040] The model building module is used to obtain the inspection path based on the set of wind turbine areas and the set of drone starting positions, and to build an inspection duration model based on the inspection path and the supply and demand relationship.

[0041] The allocation module is used to allocate inspection areas of the UAV based on the inspection duration model and using the Vogel method to obtain the allocation results of the UAV inspection areas; wherein, the Vogel method is used to allocate UAV inspection wind turbine tasks according to the difference characteristics of path duration.

[0042] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines as described in the first aspect above.

[0043] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines as described in the first aspect above.

[0044] The method, system, and medium for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines provided in this application embodiment have at least the following technical effects.

[0045] First, clustering based on wind turbine characteristic parameters integrates scattered wind turbines into a set of regions, reducing the complexity of subsequent optimization and improving computational efficiency. Second, by constructing a model that integrates the inspection duration of the UAV's starting point and the wind turbine region, the inspection duration of the inspection path is effectively determined, laying the foundation for subsequent allocation. Finally, the Vogel method is used to allocate inspection regions based on the differences in inspection duration, prioritizing the critical paths with the largest differences, effectively improving inspection efficiency while ensuring that the inspection frequency of each wind turbine meets the standard. This application improves inspection efficiency while ensuring inspection quality by collaboratively designing clustering dimensionality reduction, path modeling, and inspection region allocation, thus solving the problem of low inspection efficiency caused by unreasonable allocation of UAV inspection regions in related technologies.

[0046] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0047] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0048] Figure 1 This is a flowchart of a method for allocating inspection areas for wind turbines using unmanned aerial vehicles (UAVs).

[0049] Figure 2 This is a flowchart illustrating step S101 according to an exemplary embodiment;

[0050] Figure 3 This is a flowchart illustrating step S102 according to an exemplary embodiment;

[0051] Figure 4 This is a flowchart illustrating step S103 according to an exemplary embodiment;

[0052] Figure 5 This is a system structure block diagram illustrating the allocation of unmanned aerial vehicle (UAV) inspection areas for wind turbines, according to an exemplary embodiment.

[0053] Figure 6 This is a structural block diagram of an electronic device according to an exemplary embodiment. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0055] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0056] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0057] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0058] In this document, it should be understood that the terms used may be technical means used to implement part of the present invention or other summary technical terms. For example, the terms may include:

[0059] Wind turbine region set: A set of wind turbine groups formed by grouping wind turbines in a wind farm according to their latitude and longitude coordinates and tower height features using a clustering algorithm.

[0060] UAV origin location set: refers to the set of coordinates of the hangars or take-off and landing points of UAVs deployed in the wind farm.

[0061] Inspection area: refers to the wind turbines that need to be inspected and their surrounding areas, usually corresponding to an area formed by a cluster of wind turbines, such as a wind turbine area.

[0062] Vogel's method: A transportation problem optimization algorithm that prioritizes the path with the greatest cost difference by calculating row / column penalties (opportunity costs).

[0063] Gap Statistic: A metric used to determine the optimal number of clusters. It is calculated by comparing the clustering results of actual data with those of random data.

[0064] Penalty Count: The difference between the second shortest inspection time and the shortest inspection time from a certain drone hangar to each wind turbine area.

[0065] Penalty Count: The difference between the second shortest inspection time and the shortest inspection time from a certain drone hangar to each wind turbine area.

[0066] In a first aspect, embodiments of this application provide a method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines. Figure 1 This is a flowchart of a method for allocating inspection areas for wind turbines using unmanned aerial vehicles (UAVs), such as... Figure 1 As shown, the method includes:

[0067] Step S101: Cluster the characteristic parameters of each wind turbine in the wind farm to obtain a set of wind turbine regions of multiple categories.

[0068] Step S102: Based on the set of wind turbine areas and the set of starting locations of UAVs, construct an inspection duration model, wherein the inspection duration model determines the inspection duration through the inspection path.

[0069] Step S103: Based on the inspection duration model, the inspection area of ​​the UAV is allocated using the Vogel method to obtain the allocation result of the UAV inspection area; wherein, the Vogel method is used to allocate the UAV inspection wind turbine task according to the difference characteristics of the inspection path duration.

[0070] In summary, the embodiments of this application provide a method for allocating inspection areas for wind turbines using unmanned aerial vehicles (UAVs). This application improves inspection efficiency while ensuring inspection quality through the collaborative design of clustering dimensionality reduction, path modeling, and inspection area allocation, thus solving the problem of low inspection efficiency caused by unreasonable UAV inspection area allocation in related technologies.

[0071] In one embodiment, before clustering the characteristic parameters of each wind turbine in the wind farm, the method further includes:

[0072] The latitude and longitude coordinates and tower height of each wind turbine are obtained by using the drone's positioning sensors and altitude measurement device.

[0073] Optionally, the latitude and longitude coordinates of the wind turbine base can be obtained through a positioning module mounted on the drone (accuracy ±10cm). Simultaneously, the height of the tower top relative to the drone can be measured using a lidar altimeter. Combined with the drone's own barometer altitude data, the absolute height of the tower can be calculated (error <0.5m). Measurements are then performed on each wind turbine to obtain its latitude and longitude coordinates and tower height.

[0074] Based on preset data verification rules, outlier detection and correction are performed on latitude and longitude coordinates and tower height to obtain the characteristic parameters of each wind turbine after outlier detection and correction.

[0075] Optionally, based on the pre-stored standard coordinate range in the wind farm's Geographic Information System (GIS), outlier data with latitude and longitude exceeding the standard coordinate range is filtered out. For tower height data, outliers exceeding the design maximum value (e.g., 150m) or falling below the minimum value (e.g., 50m) are removed. Outlier coordinates are corrected using a weighted average of the coordinates of adjacent wind turbines, and outlier height values ​​are replaced with the design height of the same type of wind turbine. After this process, the characteristic parameters of each wind turbine are obtained after outlier detection and correction.

[0076] By using the coordinated measurement of coordinates and height by positioning sensors and lidar, and by detecting and correcting outliers in the coordinates and height, the characteristic parameters of each wind turbine are obtained, providing high-precision input for subsequent clustering.

[0077] Figure 2 This is a flowchart illustrating step S101 according to an exemplary embodiment, as follows: Figure 2 As shown, step S101 involves clustering the characteristic parameters of each wind turbine in the wind farm to obtain a multi-category set of wind turbine regions, including:

[0078] Step S1011: Process the latitude and longitude coordinates and tower height in the feature parameters by zero-mean normalization to obtain normalized feature data.

[0079] Optionally, the coordinates of each wind turbine are set to [x, y, h], where x is the longitude coordinate, y is the latitude coordinate, and h is the height of the wind turbine tower. Zero-mean normalization is performed on the coordinate data of b wind turbines, mapping the data to a distribution with a mean of 0 and a standard deviation of 1, resulting in the normalized data b of the wind turbine units. i =[u i ,v i ,z i (i = 1, 2, ..., b).

[0080] Step S1012: Based on the normalized feature data, determine the number of target clusters using interval statistics.

[0081] Optionally, the gap statistic determines the value of the target cluster number c. For c∈(1,b), the gap statistic is defined as Gap(c) = E(logD). c )-logD c Where c is the number of clusters, and Dc is the sum of squared errors within each cluster, which is also a measure of the distance from each point to the cluster center. For each selected number of clusters, logDc is a log function used to facilitate error comparison for the sample region. The first term, ElogDc, is the expected loss of random samples in the same sample space, and the second term, logDc, is the loss of the actual wind turbine sample. First, within the sample region [u,v,z], a number of random samples, the same as the original number, are randomly generated according to a uniform distribution. The K-means of these random samples are then calculated to obtain Dc. c By repeating this process multiple times, E(logD) can be approximately calculated. c This represents the loss of the random sample. Subsequently, the K-means of the actual wind turbine sample loss is calculated to obtain logD. c For each c, we can obtain a gap(c). The maximum gap(c), i.e., the maximum difference between the random sample loss and the actual sample loss, minimizes the actual sample loss. We take this maximum value for c. k The target number of clusters for this wind farm.

[0082] Step S1013: Based on the target number of clusters, cluster the wind turbines using a clustering algorithm and update the cluster centers until the convergence condition is met to obtain a set of wind turbine regions of multiple categories. Specifically, this includes:

[0083] Calculate the Euclidean distance between all adjacent wind turbines to obtain the average distance, and select the sample with the minimum distance from the average distance as the initial cluster center.

[0084] Each wind turbine is assigned to the nearest initial cluster center.

[0085] Based on the assigned clustered fans, the cluster centers are updated by calculating the mean of the characteristic data of each clustered fan.

[0086] When the change in cluster center is less than a preset threshold, the target cluster center is determined and used as the set of wind turbine regions.

[0087] Optionally, step S1013 can be further explained through the following steps:

[0088] Step 1: Calculate the Euclidean distance d(b) between each pair of data points. i ,b j ), calculate the mean distance of the sample Mendist, where That is, the sum of the pairwise distances between all sample points is divided by the number of combinations of sample point b, and the smallest distance is found; then its center is calculated; find the second smallest distance, calculate its center, and judge it. If the distance between this center and a previously calculated cluster center is less than the Meanist distance, then this cluster center is excluded, and the process continues to find the third smallest distance and calculate its center, until c is found. k If no initial cluster center is found, then find the next initial cluster center, and so on, until c is found. k Until the initial cluster centers are identified.

[0089] Step 2: For each sample in the dataset, calculate its value to c. k The distance between each cluster center is determined and assigned to the cluster center with the smallest distance.

[0090] Step 3: For each cluster, recalculate its cluster center position;

[0091] Step 4: Repeat steps 2 and 3 until a termination condition is met (such as the number of iterations, the position of the cluster center, etc.).

[0092] It is worth noting that clustering objects require both inter-group heterogeneity and intra-group similarity. For a series of wind turbine coordinate data [x, y, h] from the same wind farm, inter-group heterogeneity is reflected in the fact that some turbines exhibit a linear distribution, while others show a network distribution; furthermore, some turbines are located on mountain peaks, while others are distributed on flat terrain. Intra-group similarity is reflected in the fact that the coordinate data of turbines with each distribution characteristic show certain similarities and regularities. Therefore, clustering the wind turbines effectively reduces the dimensionality of the subsequent optimization problem.

[0093] Steps S1011 to S1013. First, zero-mean normalization is used to eliminate the scale difference between latitude, longitude, and altitude parameters, so that the clustering algorithm is not affected by the dimension. Second, the optimal number of clusters is automatically determined based on the gap statistic, avoiding overfitting or underfitting caused by manual setting. Finally, the improved K-means algorithm ensures the rationality of the spatial distribution of the wind turbine area set, providing a highly consistent input basis for subsequent inspection path optimization.

[0094] Figure 3 This is a flowchart illustrating step S102 according to an exemplary embodiment, as follows: Figure 3 As shown, step S102 involves constructing an inspection duration model based on the wind turbine area set and the drone starting point location set. The inspection duration model determines the inspection duration through the inspection path. Specifically, this includes the following steps:

[0095] Step S1021: Based on the spatial relationship between the starting position of each UAV and the wind turbines in the wind turbine area, calculate the inspection path of the UAV to each wind turbine area using a route planning algorithm, and determine the initial inspection time based on the inspection path.

[0096] Step S1022: Use the available resources of the drone as the supply count and the number of inspections as the demand count, and establish a supply and demand relationship based on the supply count and the demand count.

[0097] Step S1023: Based on the initial inspection duration and the number of inspections, construct an inspection duration model through supply and demand.

[0098] Optionally, suppose a wind farm has a 'a' drone hangars and b wind turbines. Cluster the b wind turbines into c classes based on their latitude and longitude coordinates and tower height. Let the drone hangar set A = [a1, a2, ..., a...]. n The wind turbine cluster set C = [c1, c2, ..., c3] m The inspection path duration is defined as follows: Where, d ij This represents the initial inspection path duration required for the UAV to inspect hangar i and reach wind turbine cluster j. The inspection path from the UAV to each wind turbine area is calculated using a flight path planning algorithm. Based on this path, the initial inspection duration is determined. ij Let represent the number of flights from drone hangar i to wind turbine cluster j. The constraint is that the inspection time for each drone hangar is limited within the same time period, meaning the total number of inspections is limited. The wind turbine inspections need to be completed as quickly as possible, and this is denoted as the supply quantity. Each wind turbine cluster requires a certain number of drone inspections, denoted as the procurement quantity. Since the Vogel method is a planning method based on supply and demand relationships, the traveling salesman problem of wind turbine path planning is transformed into a mathematical programming supply and demand model to facilitate the application of the Vogel method.

[0099] It should be noted that the drone can be a DJI M400, with a half-load flight speed of about 60 km / h and a half-load operating time of about 50 minutes. The spacing between wind turbines in a common wind farm is 1 km, which means that the drone flight time between two adjacent wind turbines is about 1 minute. In order to complete the inspection of the wind turbine's down conductor continuity test, blade defect test and other inspection tasks, the inspection time for each wind turbine is about 15-20 minutes.

[0100] Figure 4 This is a flowchart illustrating step S103 according to an exemplary embodiment, as follows: Figure 4As shown, step S103 involves allocating inspection areas for the UAV based on the inspection duration model and using the Vogel method to obtain the allocation results. The Vogel method is used to allocate UAV inspection tasks for wind turbines based on the differences in inspection path duration. Specifically, this includes the following steps:

[0101] Step S1031: Based on the inspection duration determined in the inspection duration model, calculate the difference characteristics between the set of drone starting positions and the set of wind turbine areas. Specifically, this includes:

[0102] Based on the inspection time from the starting point of each drone to each wind turbine area, calculate the first difference between the next hour length and the minimum time for the corresponding row, and use it as the row penalty number.

[0103] Based on the inspection time from the starting point of each drone to each wind turbine area, calculate the second difference between the second hour length and the minimum time for the corresponding column, and use it as the column penalty number.

[0104] Multiple row penalties and multiple column penalties are used as distinguishing features.

[0105] Optionally, based on the inspection route duration unit matrix in the inspection duration model, calculate the penalty for each row and column of the n x m inspection route duration unit matrix D. The penalty is the difference between the shortest and second shortest duration in each column or row (difference characteristic), i.e., penalty = second shortest duration - shortest duration. Select the row or column corresponding to the maximum penalty, and based on the row or column where the maximum penalty occurs, select the minimum inspection duration (cost) of the corresponding row or column to determine the supply quantity. Its corresponding position is the position for the next step. If two or more rows and columns have the same maximum penalty, select the row or column corresponding to the shortest duration as the row or column for this calculation. The position of its minimum duration is the position for the next step. For example, if the penalties are the same, rows a1 and a2, compare the two minimum inspection durations of rows a1 and a2 in the table, and select the smaller one as the position for the next step.

[0106] Table 1 shows a 3x3 matrix D representing the inspection route duration, with elements D... ij Representative from drone hangar A i In the table, i=3 flies to the wind turbine cluster C. j The table shows the flight time spent at the nearest wind turbine tower, where i=3 and j=3, in minutes. The supply (number of inspections) is the available resources in the UAV hangar A, which is the number of inspections that can be flown. The purchase quantity (number of inspections) is the inspection demand in the wind turbine cluster set C.

[0107] Table 1: Inspection Route Duration Diagram

[0108]

[0109] Step S1032: Based on the supply and demand relationship in the inspection duration model, prioritize assigning drones to inspection areas with the greatest differences in characteristics. Specifically, this includes:

[0110] Find the maximum row penalty or the maximum column penalty among multiple row penalties and multiple column penalties;

[0111] The maximum row penalty or the maximum column penalty is used as the maximum difference feature. If the maximum difference feature is the maximum row penalty, the starting position of the UAV is determined according to the maximum row penalty, and the wind turbine area is determined according to the minimum inspection time in the row corresponding to the maximum row penalty. If the maximum difference feature is the maximum column penalty, the wind turbine area is determined according to the maximum number of rows and columns, and the starting position of the UAV is determined according to the minimum inspection time in the column corresponding to the maximum column penalty.

[0112] The available resources for the drone are determined based on its starting point location, and the number of inspections is determined based on the wind turbine area.

[0113] Based on the supply and demand relationship of available resources and inspection frequency, the available drone resources are allocated to the wind turbine area.

[0114] Optionally, in the inspection duration model, f ij This represents the number of inspection flights from hangar i to wind turbine cluster j. In the supply-demand balance matrix, at the corresponding position obtained in the previous step, according to the Vogel method (i.e., the position of the smallest element in the row or column corresponding to the maximum penalty), fill in the matrix position with a value that makes the row or column at this position reach the upper limit. That is, this value should be the smaller value of the remaining value of the row or column, i.e., it can reach the upper limit of at least one remaining value of the row or column, and update the purchase quantity or supply quantity. When the corresponding row or column reaches the upper limit of the remaining value, update its remaining quantity.

[0115] As shown in Table 1, among the various row and column penalties, the maximum penalty is found in row A2. From row A2, the minimum inspection route duration is selected as column C3. Therefore, the maximum difference feature between the set of drone starting locations (drone hangar A) and the set of wind turbine areas (wind turbine cluster set C) is obtained. Based on the maximum difference feature, drone hangar A and wind turbine cluster set C corresponding to the maximum difference feature are determined. As shown in Table 1, the maximum penalty corresponds to row A2. Based on row A2, where the maximum penalty corresponds, and the minimum inspection duration of 2 is determined in row A2, corresponding to column C3, the available drone resources and inspection times are determined. Based on the supply and demand relationship of available resources and inspection times, the available drone resources are allocated to the wind turbine areas. That is, all inspection times in hangar A2 are supplied to wind turbine cluster C3.

[0116] Step S1033: Update the supply and demand relationship, prioritize the allocation of drones to the drone inspection path with the greatest difference characteristics until the inspection requirements are met, and obtain the allocation results of the drone inspection area.

[0117] Optionally, firstly, in the unit matrix of transported electricity, remove the rows or columns corresponding to the supply or purchase quantities that have reached their upper limits. That is, the row or column will no longer be supplied or purchased. The row corresponds to the supply status, and the column corresponds to the purchase status. In other words, remove C3 and A2 from Table 1, and update the penalty, supply, and purchase quantities.

[0118] Subsequently, in the remaining inspection route duration matrix, the penalty number corresponding to each row or column is calculated until all elements in the inspection route duration unit matrix are crossed out, that is, all supply and procurement quantities are satisfied or procurement quantities are satisfied. At this time, the drone completes the inspection of all wind turbines.

[0119] As shown in Table 2, the maximum penalty corresponds to column C2, and the minimum inspection route duration corresponds to row A3. Therefore, all inspection times for hangar A3 are supplied to wind turbine cluster C2, and the supply and procurement table is updated.

[0120] Table 2: Inspection Route Duration Update Diagram

[0121]

[0122] As shown in Table 3, the updated target inspection route duration is illustrated. At this point, simply supplying all inspection attempts for hangars A3 and A1 to wind turbine cluster C2 is sufficient to complete the inspection area allocation. The inspection duration T... 巡检路线 =Number of inspections * Inspection duration, i.e., T 巡检路线 =2*2+2.9*1+3.5*2+4*1=17.9, achieving the goal of the shortest total time.

[0123] Table 3: Target Inspection Route Duration Diagram

[0124]

[0125] Steps S1031 to S1033 utilize the Vogel method to allocate drone inspection tasks for wind turbines, improving the efficiency of these tasks. First, critical paths are accurately identified based on row / column penalty calculations (the difference between the second smallest and smallest time durations). For example, in Table 1, the penalty of 2 for A2→C3 is the maximum, prioritizing these paths with the highest opportunity cost to avoid subsequent iterations falling into local optima. Second, by updating supply and demand in real-time (e.g., after allocating all two inspections of A2 to C3, the row is removed), resource allocation strictly meets constraints (supply ≤ drone endurance, procurement ≥ minimum wind turbine inspection requirements). The final allocation scheme (e.g., allocating two inspections for A1→C1 and one inspection for A3→C2) fully utilizes drone resources (resource utilization > 95%), ensures all wind turbine areas complete the required number of inspections, and minimizes the total time (Table 1 case total time = 3.5 × 2 + 2.9 × 1 + 4 × 1 = 14.9 minutes).

[0126] In summary, this application provides a method for allocating UAV inspection areas for wind turbines. First, an improved K-means clustering algorithm is used to divide the wind turbines in the wind farm into several sets of regions based on latitude and longitude coordinates and tower height characteristics, effectively reducing the dimensionality of the subsequent optimization problem. Second, an inspection time model is constructed based on a route planning algorithm to accurately calculate the inspection time of the UAV from each starting point to the wind turbine area. Finally, the Vogel method is applied to iteratively calculate the row penalty (the second smallest to the minimum time difference between each hangar) and column penalty (the second smallest to the minimum time difference between each wind turbine area), prioritizing the allocation of path combinations with the largest cost differences. Under the constraints of UAV endurance and wind turbine inspection requirements, a globally optimal solution that minimizes the total inspection time is achieved. This solves the problem of low inspection efficiency caused by unreasonable UAV inspection area allocation in related technologies.

[0127] Secondly, embodiments of this application provide a drone inspection area allocation system for wind turbines. Figure 5 This is a system structure block diagram illustrating the allocation of unmanned aerial vehicle (UAV) inspection areas for wind turbines, according to an exemplary embodiment. Figure 5 As shown, the system includes: a clustering module 510, a model construction module 520, and an allocation module 530; wherein,

[0128] Clustering module 510 is used to perform clustering processing on the characteristic parameters of each wind turbine in the wind farm to obtain a set of wind turbine regions of multiple categories.

[0129] Model module 520 is used to construct an inspection duration model based on the set of wind turbine areas and the set of drone starting positions. The inspection duration model determines the inspection duration through the inspection path.

[0130] The allocation module 530 is used to allocate inspection areas for UAVs based on the inspection duration model and the Vogel method, and obtain the allocation results of UAV inspection areas; wherein, the Vogel method is used to allocate UAV inspection wind turbine tasks according to the difference characteristics of path duration.

[0131] In summary, the UAV inspection area allocation system for wind turbines provided in this application improves inspection efficiency while ensuring inspection quality through the collaborative design of clustering module, model building module and allocation module, and solves the problem of low inspection efficiency caused by unreasonable UAV inspection area allocation in related technologies.

[0132] It should be noted that the UAV inspection area allocation system for wind turbines provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, the terms "module," "unit," "subunit," etc., can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0133] Thirdly, embodiments of this application provide an electronic device, Figure 6 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 6 As shown, the electronic device may include a processor 61 and a memory 62 storing computer program instructions.

[0134] Specifically, the processor 61 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0135] The memory 62 may include a mass storage device for data or instructions. For example, and not limitingly, the memory 62 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 62 may include removable or non-removable (or fixed) media. Where appropriate, the memory 62 may be internal or external to a data processing device. In a particular embodiment, the memory 62 is non-volatile memory. In a particular embodiment, the memory 62 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only ROM (PROM), an erasable programmable read-only ROM (EPROM), an electrically erasable programmable read-only ROM (EEPROM), an electrically alterable read-only ROM (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0136] The memory 62 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 61.

[0137] The processor 61 reads and executes computer program instructions stored in the memory 62 to implement any of the above embodiments of the UAV inspection area allocation method for wind turbines.

[0138] In one embodiment, a drone inspection area allocation device for wind turbines may further include a communication interface 63 and a bus 60. Wherein, as Figure 6 As shown, the processor 61, memory 62, and communication interface 63 are connected through bus 60 and complete communication with each other.

[0139] The communication interface 63 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication port 63 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0140] Bus 60 includes hardware, software, or both, that couples together components of a drone-based inspection area distribution device for wind turbines. Bus 60 includes, but is not limited to, at least one of the following: Data Bus, Address Bus, Control Bus, Expansion Bus, and Local Bus. For example, and not as a limitation, bus 60 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 60 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0141] Fourthly, embodiments of this application provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines provided in the first aspect.

[0142] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0143] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform steps implementing the method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines provided in the first aspect.

[0144] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0145] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0146] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for allocating unmanned aerial vehicle (UAV) inspection areas for wind turbines, characterized in that, The method comprises: performing clustering processing on feature parameters of each wind turbine in the wind farm to obtain a wind turbine region set of multiple categories; obtaining an inspection path according to the wind turbine region set and a set of starting positions of unmanned aerial vehicles, constructing an inspection time length model through the inspection path and a supply-demand relationship; based on the inspection time length model, assigning the unmanned aerial vehicles to an inspection region through a Vogel method to obtain an assignment result of the unmanned aerial vehicle inspection region; wherein the Vogel method is used to assign the unmanned aerial vehicle inspection wind turbine task according to the difference characteristics of the inspection path time length. 2.The method of claim 1, wherein, The assignment of the unmanned aerial vehicles to the inspection region based on the inspection time length model through the Vogel method to obtain the assignment result of the unmanned aerial vehicle inspection region comprises: calculating the difference characteristics between the set of starting positions of the unmanned aerial vehicles and the wind turbine region set according to the inspection time length determined in the inspection time length model; assigning the unmanned aerial vehicles to the inspection region with the maximum difference characteristics according to the supply-demand relationship in the inspection time length model; updating the supply-demand relationship to assign the unmanned aerial vehicles to the unmanned aerial vehicle inspection path with the maximum difference characteristics until the inspection demand is met to obtain the assignment result of the unmanned aerial vehicle inspection region. 3.The method of claim 2, wherein, The calculation of the difference characteristics between the set of starting positions of the unmanned aerial vehicles and the wind turbine region set according to the inspection time length determined in the inspection time length model comprises: calculating a first difference value between the second shortest time length and the shortest time length corresponding to each row as a row penalty according to the inspection time length of each starting position of the unmanned aerial vehicles to each wind turbine region; calculating a second difference value between the second shortest time length and the shortest time length corresponding to each column as a column penalty according to the inspection time length of each starting position of the unmanned aerial vehicles to each wind turbine region; taking the plurality of row penalties and the plurality of column penalties as the difference characteristics. 4.The method of claim 3, wherein, The assignment of the unmanned aerial vehicles to the inspection region with the maximum difference characteristics comprises: obtaining the maximum row penalty or the maximum column penalty from the plurality of row penalties and the plurality of column penalties; taking the maximum row penalty or the maximum column penalty as the maximum difference characteristics, and if the maximum difference characteristics are the maximum row penalty, determining the starting position of the unmanned aerial vehicle according to the maximum row penalty and determining the wind turbine region according to the shortest inspection time length in the corresponding row of the maximum row penalty; if the maximum difference characteristics are the maximum column penalty, determining the wind turbine region according to the maximum column penalty and determining the starting position of the unmanned aerial vehicle according to the shortest inspection time length in the corresponding column of the maximum column penalty; determining the available resources of the unmanned aerial vehicle according to the starting position of the unmanned aerial vehicle and determining the inspection times according to the wind turbine region; based on the supply-demand relationship of the available resources and the inspection times, assigning the available resources of the unmanned aerial vehicle to the wind turbine region.

5. The method of claim 1, wherein, The clustering processing on the feature parameters of each wind turbine in the wind farm to obtain a wind turbine region set of multiple categories comprises: performing processing on the longitude and latitude coordinates and the tower height in the feature parameters through zero-mean normalization to obtain normalized feature data; determining a target clustering number based on the normalized feature data through interval statistics; The fan clustering is performed based on the target clustering number through a clustering algorithm, and the clustering center is updated until a multi-class fan area set is obtained when a convergence condition is met.

6. The method of claim 5, wherein, The fan clustering is performed based on the target clustering number through a clustering algorithm, and the clustering center is updated until a multi-class fan area set is obtained when a convergence condition is met, including: The Euclidean distance between all adjacent fans is calculated to obtain an average distance, and a sample corresponding to the minimum distance in the average distance is selected as an initial clustering center; Each fan is assigned to the initial clustering center closest to it; The clustering center is updated by calculating the mean value of the feature data of each clustering fan; When the change in the clustering center is less than a preset threshold, the target clustering center is determined, and the target clustering center is taken as the fan area set.

7. The method of claim 1, wherein, The inspection time length model is constructed according to the fan area set and the unmanned aerial vehicle starting point position set, including: The inspection path of the unmanned aerial vehicle to each fan area is calculated through a route planning algorithm according to the spatial position relationship between each unmanned aerial vehicle starting point position and the fan in the fan area, and the initial inspection time length is determined based on the inspection path; The available resources of the unmanned aerial vehicle are taken as the supply times, the inspection times are taken as the demand times, and a supply-demand relationship is established based on the supply times and the demand times; The inspection time length model is constructed through the supply-demand relationship based on the initial inspection time length and the inspection times.

8. The method of claim 1, wherein, Before the characteristic parameters of each fan in the wind farm are clustered, the method further includes: The latitude and longitude coordinates and the tower height of each fan are obtained through the positioning sensor and the height measuring device of the unmanned aerial vehicle; Based on a preset data verification rule, the latitude and longitude coordinates and the tower height are subjected to outlier detection and correction to obtain the characteristic parameters of each fan after outlier detection and correction.

9. An unmanned aerial vehicle (UAV) inspection area assignment system for a wind farm, the system comprising: a UAV; a UAV controller; a wind farm controller; and a communication link between the UAV controller and the wind farm controller. Including: A clustering module, a model construction module, and an assignment module; wherein The clustering module is configured to cluster the characteristic parameters of each fan in the wind farm to obtain a multi-class fan area set The model construction module is configured to obtain an inspection path according to the fan area set and the unmanned aerial vehicle starting point position set, and to construct an inspection time length model through the inspection path and a supply-demand relationship; The assignment module is configured to assign the unmanned aerial vehicle to an inspection area based on the inspection time length model through the Vogel method to obtain an assignment result of the unmanned aerial vehicle inspection area; wherein the Vogel method is used to assign the unmanned aerial vehicle to an inspection fan task according to the difference characteristics of the path time length.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the unmanned aerial vehicle inspection area assignment method for the fan according to any one of claims 1 to 8.

Images