Wind farm operation and maintenance scheduling method and system and computer device

Abstract

This invention provides a wind farm operation and maintenance scheduling method, system, and computer equipment. Based on the environmental information of the wind farm, a detection mode is determined. When the detection mode is determined to be an early warning detection mode, images of each wind turbine are acquired to obtain early warning images. The coating early warning area of ​​the wind turbine blades is determined based on the attached water film morphology in the early warning images, and an early warning operation and maintenance queue is generated based on the area of ​​the coating early warning area. When the detection mode is determined to be a tracking detection mode, a tracking image group at multiple time points is acquired through a cyclic acquisition strategy. Newly added icing areas at each time point are extracted from the tracking image group as coating failure areas. The coating failure areas are sorted by time point to obtain a failure sequence, and the failure level is determined based on the failure sequence. A failure operation and maintenance queue is generated based on the coating failure area and the failure level. This invention realizes the location of local failure areas of the anti-icing coating and the determination of the failure level.

Description

Technical Field

[0001] This invention relates to data processing technology, and more particularly to a wind farm operation and maintenance scheduling method, system, and computer equipment. Background Technology

[0002] Wind turbine blades are exposed to complex natural environments for extended periods and are highly susceptible to icing. Therefore, anti-icing coatings are typically applied to the blade surfaces. The integrity of these anti-icing coatings directly affects the power generation efficiency and operational safety of wind turbines.

[0003] Currently, the inspection of anti-icing coatings on wind turbine blades mainly relies on conventional methods, such as manually inspecting the coating quality by carrying instruments close to the blade. This method is inefficient and highly dependent on manual labor. Existing inspection solutions often only determine which blades are iced after large-scale icing has occurred. However, once a localized area of ​​the coating on a wind turbine blade is damaged and icing occurs first, the icing can spread and even cover the entire blade. This phenomenon means that current technology can only detect the entire blade being iced afterward, without being able to determine which localized area of ​​the coating has failed. Because the specific location and severity of the failure cannot be determined, current maintenance scheduling lacks targeted data support, often leading to blind replacement of the entire coating during repairs, significantly increasing maintenance costs.

[0004] Therefore, accurately locating the partially failed areas of the anti-icing coating and determining its severity level, thereby improving maintenance efficiency and reducing maintenance costs, has become a critical issue that urgently needs to be addressed. Summary of the Invention

[0005] This invention provides a wind farm operation and maintenance scheduling method, system, and computer equipment, which can accurately locate some of the failure areas of the anti-icing coating and determine its failure severity level, thereby improving operation and maintenance efficiency and reducing operation and maintenance costs.

[0006] A first aspect of the present invention provides a wind farm operation and maintenance scheduling method, comprising: The detection mode is determined based on the environmental information of the wind farm, and the detection mode includes an early warning detection mode and a tracking detection mode. When the detection mode is determined to be the early warning detection mode, images are acquired from each wind turbine to obtain early warning images. The early warning area of ​​the wind turbine blade coating is determined based on the early warning images, and an early warning maintenance queue is generated based on the coating early warning area. When the detection mode is determined to be the tracking detection mode, multiple tracking image groups are obtained by acquiring images of the wind turbine according to the cyclic acquisition strategy. Based on the tracking image groups, the coating failure areas of the wind turbine blades and the failure levels of each coating failure area are determined, and a failure maintenance queue is generated according to the coating failure areas and failure levels.

[0007] Optionally, in one possible implementation of the first aspect, the determination of the detection mode based on the environmental information of the wind farm includes an early warning detection mode and a tracking detection mode, comprising: Obtain preset detection conditions, including water attachment detection conditions and freezing detection conditions; The environmental information of the wind farm is retrieved, and the environmental information is compared with the detection conditions. If the environmental information meets the water attachment detection conditions, the detection mode is determined to be the early warning detection mode. When the environmental information meets the icing detection conditions, the detection mode is set to tracking detection mode.

[0008] Optionally, in one possible implementation of the first aspect, the step of acquiring early warning images from each wind turbine and determining the coating early warning area of ​​the wind turbine blades based on the early warning images includes: Control the drone to collect images of the wind turbines and obtain early warning images for each wind turbine; Based on the early warning image, the water adhering to the wind turbine blades is identified, and the morphology of the adhering water is determined, including water droplet morphology and water film morphology. When the form of the adhering water is determined to be water droplets, the outline area of ​​the corresponding adhering water is obtained as the normal area of ​​the coating. When the form of the attached water is determined to be a water film, the outline area of ​​the corresponding attached water is extracted as the coating warning area.

[0009] Optionally, in one possible implementation of the first aspect, generating the early warning maintenance queue based on the coating early warning area includes: Calculate the warning area of ​​the coating warning zone on each wind turbine, obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the warning area, and obtain the warning maintenance queue.

[0010] Optionally, in one possible implementation of the first aspect, the step of acquiring multiple tracking image groups from the wind turbine according to a cyclic acquisition strategy includes: The wind turbines in the wind farm are divided into multiple wind turbine inspection groups, and the corresponding UAVs for each wind turbine inspection group are determined. The drone is controlled to acquire images of the wind turbines in the corresponding wind turbine detection group at multiple time points, resulting in tracking image groups corresponding to each time point.

[0011] Optionally, in one possible implementation of the first aspect, determining the coating failure areas of the wind turbine blades and the failure levels of each coating failure area based on the tracking image set includes: Based on the tracking image group, newly added icing areas at each time point are extracted as coating failure areas; Based on the coating failure areas, a failure area group corresponding to each time point is constructed, and the failure area group is sorted in ascending order based on the time point to obtain the failure sequence; The failure level of the coating failure area in each failure area group is determined based on the failure sequence.

[0012] Optionally, in one possible implementation of the first aspect, generating the failure maintenance queue based on the coating failure area and failure level includes: Determine the failure area of ​​the coating failure region corresponding to each failure level on the wind turbine; Retrieve the preset reference coefficients corresponding to each failure level, calculate the product of each failure area and the corresponding preset reference coefficient, and obtain the failure value; The operation and maintenance value of the wind turbine is obtained by summing up the multiple failure values ​​corresponding to the wind turbine. Obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the maintenance value, and obtain the failure maintenance queue.

[0013] A second aspect of the present invention provides a wind farm operation and maintenance scheduling system, comprising: The judgment module is used to determine the detection mode based on the environmental information of the wind farm, and the detection mode includes an early warning detection mode and a tracking detection mode. The early warning module is used to acquire images of each wind turbine to obtain early warning images when the detection mode is early warning detection mode, determine the coating early warning area of ​​the wind turbine blades based on the early warning images, and generate an early warning maintenance queue based on the coating early warning area. The failure module is used to determine that when the detection mode is the tracking detection mode, it acquires multiple tracking image groups of the wind turbine according to the cyclic acquisition strategy, determines the coating failure area of ​​the wind turbine blade and the failure level of each coating failure area based on the tracking image group, and generates a failure operation and maintenance queue according to the coating failure area and failure level.

[0014] A third aspect of the present invention provides a computer device comprising: a memory, a processor, and a computer program, the computer program being stored in the memory, and the processor executing the computer program to perform the methods described in the first aspect of the present invention and various possible methods related to the first aspect.

[0015] A fourth aspect of the present invention provides a storage medium storing a computer program, which, when executed by a processor, is used to implement the first aspect of the present invention and various methods possibly involved in the first aspect.

[0016] The beneficial effects of this invention are as follows: 1. This invention determines the detection mode based on the environmental information of the wind farm. When the environmental information meets specific water adhesion detection conditions or icing detection conditions, it triggers an early warning detection mode and a tracking detection mode respectively, achieving precise location of partial failures in the anti-icing coating of wind turbine blades. This invention obtains the early warning area of ​​the wind turbine blade coating and generates an early warning maintenance queue through the early warning detection mode, and obtains the coating failure area and its failure level of the wind turbine blade coating and generates a failure maintenance queue through the tracking detection mode. This avoids the shortcomings of existing technologies that can only make post-event overall icing judgments without being able to trace the specific coating failure location. It enables the location and investigation of hidden dangers and damaged parts of the anti-icing coating surface under different environmental conditions, avoiding the waste of maintenance resources caused by blindly replacing the entire coating.

[0017] 2. This invention controls a drone to acquire images of wind turbines in an early warning detection mode, obtaining early warning images. Based on these images, it identifies water adhering to the turbine blades and determines whether the water droplets or film form. This invention transforms the physical changes in the hydrophobic properties of the coating surface into objective characteristics of the adhering water morphology, avoiding the high cost and inefficiency of traditional manual instrument-carrying detection. This invention extracts the outline region of the adhering water in the form of a water film as the coating early warning region, calculates the early warning area of ​​the coating early warning region on each wind turbine, and sorts the equipment numbers in descending order based on the early warning area to obtain an early warning maintenance queue, achieving efficient maintenance scheduling.

[0018] 3. In the tracking and detection mode, this invention divides wind turbines into multiple turbine detection groups, controls drones to acquire images at multiple time points to obtain tracking image groups, and extracts newly added icing areas at each time point as coating failure areas based on the tracking image groups, eliminating interference from the shooting time span caused by the large-scale drone patrol. This invention sorts the constructed failure area groups in ascending order based on time points to obtain a failure sequence, and determines the failure level of the coating failure areas based on the failure sequence, realizing the use of the icing time of local areas as the basis for assessing the degree of damage to the anti-icing coating. This invention further utilizes a preset benchmark coefficient corresponding to the failure level to obtain the operation and maintenance value of each wind turbine, and finally outputs a failure operation and maintenance queue containing specific equipment numbers and priority order. Attached Figure Description

[0019] Figure 1 A flowchart of a wind farm operation and maintenance scheduling method provided by the present invention; Figure 2 This is a schematic diagram of the newly added icing area at the first time point in this invention; Figure 3 This is a schematic diagram of the newly added icing area at the second time point in this invention; Figure 4This is a schematic diagram of the structure of a wind farm operation and maintenance scheduling system provided by the present invention; Figure 5 This is a schematic diagram of the hardware structure of a computer device provided by the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0021] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0022] It should be understood that in the various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0023] It should be understood that in this invention, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0024] It should be understood that in this invention, "multiple" refers to two or more. "And / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "Contains A, B, and C", "Contains A, B, and C" means that all three A, B, and C are contained; "Contains A, B, or C" means that one of A, B, and C is contained; "Contains A, B, and / or C" means that any one, two, or three of A, B, and C are contained.

[0025] It should be understood that in this invention, "B corresponding to A", "B corresponding to A", "A and B correspond", or "B and A correspond" means that B is associated with A, and B can be determined based on A. Determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information. Matching A and B is defined as a similarity between A and B that is greater than or equal to a preset threshold.

[0026] Depending on the context, "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection."

[0027] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0028] This invention provides a wind farm operation and maintenance scheduling method, such as... Figure 1 As shown, it includes: S1, determine the detection mode based on the environmental information of the wind farm, the detection mode includes early warning detection mode and tracking detection mode.

[0029] It should be noted that the anti-icing coating on wind turbine blades is usually transparent and close in color to the primer, making damage difficult to detect with the naked eye or conventional cameras. This invention employs two detection schemes, utilizing the morphological behavior of water and ice on the anti-icing coating to expose coating problems. This invention includes both early warning detection and tracking detection modes because icing environments are not frequent. Relying solely on detection during icing would only allow for post-accident verification and remediation. Therefore, when water accumulation detection conditions are met, early warning detection provides a preventative and initial detection method, allowing for proactive inspection upon detection of anomalies. Conversely, when icing detection conditions are met, tracking detection is activated to monitor the icing area and promptly identify the damaged portion of the coating. This invention achieves both early warning and prevention of anti-icing coating issues and subsequent problem localization.

[0030] Among them, environmental information refers to the current real-time meteorological parameters and equipment operating status parameters of the wind farm, such as temperature, humidity and wind turbine speed; detection mode refers to the specific inspection strategy matched and triggered based on the current real-time environmental information of the wind farm for diagnosing abnormalities of the anti-icing coating.

[0031] In some embodiments, step S1 (determining the detection mode based on the environmental information of the wind farm, wherein the detection mode includes an early warning detection mode and a tracking detection mode) includes S11-S13: S11, Obtain preset detection conditions, including water attachment detection conditions and freezing detection conditions.

[0032] It should be noted that this solution uses water or ice to expose localized failures in the anti-icing coating of wind turbine blades, so coating defects can only be detected under appropriate natural conditions. This invention pre-sets two detection conditions: one to measure whether adhering water is generated, and the other to measure whether ice will form, ensuring that the corresponding detection mode is triggered only when specific conditions are met, thus avoiding ineffective inspections.

[0033] It is understood that the present invention pre-sets detection conditions, including water attachment detection conditions and freezing detection conditions.

[0034] Among them, the detection conditions refer to the pre-configured combination of parameter thresholds used to trigger a specific detection mode; the water attachment detection conditions refer to the parameter indicators used to determine whether the current environment can generate water (dew) on the blade surface, which usually include the standard air humidity, a specific ambient temperature range, and a low fan speed. For example, the humidity can be set to be greater than or equal to 80%, the temperature to be greater than 0 degrees and less than or equal to 15 degrees, and the fan speed to be less than or equal to 3 rpm (or the fan can be set to the off state directly); the icing detection conditions refer to the parameter indicators used to determine whether the current environment has reached the point where water can freeze into ice, which usually means that the ambient temperature drops below zero degrees.

[0035] S12, retrieve the environmental information of the wind farm, compare the environmental information with the detection conditions, and determine that the detection mode is the early warning detection mode when the environmental information meets the water attachment detection conditions.

[0036] It is understandable that the system receives environmental information in real time and compares it with the detection conditions. When the environmental information is confirmed to meet the water-attached detection conditions (i.e., humidity, temperature, and fan speed are within the set range), the detection mode is set to the early warning detection mode.

[0037] Among them, the early warning detection mode refers to the inspection strategy of identifying coating warning areas in advance by recognizing the form of water adhering to the wind turbine blades.

[0038] S13, when the environmental information meets the icing detection conditions, determine the detection mode as the tracking detection mode.

[0039] Understandably, when the environmental information is confirmed to meet the icing detection conditions (i.e., the temperature has dropped below zero), the detection mode will be set to tracking detection mode.

[0040] The tracking and detection mode refers to an inspection strategy that uses tracking and identification of icing areas on wind turbine blades to accurately locate coating failure areas and determine their failure levels. Icing areas refer to the icing locations on the surface of wind turbine blades where ice adheres.

[0041] S2, when the detection mode is determined to be the early warning detection mode, images are acquired from each wind turbine to obtain early warning images, the coating early warning area of ​​the wind turbine blades is determined based on the early warning images, and an early warning maintenance queue is generated based on the coating early warning area.

[0042] It's important to note that frigid, icy environments are not the norm for wind farms. Relying solely on icy areas for coating inspection would result in a prolonged period of inconsistency in the inspection of anti-icing coatings. This step, when the conditions for water adhesion detection are met, performs initial preventative inspection of the wind turbine blades to identify warning areas in the coating. If the anti-icing coating is intact, water will adhere to it and form regular droplets; however, if the coating is locally damaged or peeling off, the water will spread out and form a film.

[0043] Among them, wind turbine blades refer to the blades of wind turbine generators, which are the physical carriers of anti-icing coatings.

[0044] In some embodiments, step S2 (acquiring images of each wind turbine to obtain warning images, and determining the coating warning area of ​​the wind turbine blades based on the warning images) includes S21-S24: S21, control the drone to collect images of the wind turbines and obtain early warning images corresponding to each wind turbine.

[0045] Understandably, when executing the early warning detection mode, the drone will be controlled to fly to the location of each wind turbine, and the camera on the drone will be controlled to collect images of the wind turbine. The high-definition images of each wind turbine collected will be used as early warning images.

[0046] Among them, the early warning image refers to the image of the wind turbine collected in the early warning detection mode.

[0047] It is worth mentioning that multiple drones can be deployed simultaneously to perform data collection tasks, depending on the number of wind turbines in the wind farm.

[0048] S22, based on the early warning image, identify the water adhering to the wind turbine blades and determine the morphology of the water adhering to the blades, including water droplet morphology and water film morphology.

[0049] It's important to note that anti-icing coatings are essentially hydrophobic coatings, designed to reduce the affinity between water and the blade surface. If the coating is intact, water adhering to it will form relatively rounded droplets due to the hydrophobic effect. However, if the coating is locally damaged, thinned, or peels off, that area will lose its hydrophobicity, causing the adhering water to spread out and form a puddle (i.e., a water film). This step identifies the morphology of the adhering water, transforming differences in coating quality that are difficult to detect visually into differences in water morphology, thus enabling preliminary testing of the anti-icing coating's performance.

[0050] Understandably, the warning image obtained in step S21 is retrieved and input into a pre-trained image recognition model (such as a deep learning semantic segmentation model). The image recognition model first filters out the background and leaf base in the warning image and identifies the water adhering to the leaf surface. Subsequently, feature extraction and classification calculations are performed on the edge features and geometry of these adhering waters. The extracted feature data is compared and analyzed with preset morphological classification standards to determine whether each adhering water appears as a water droplet or a water film.

[0051] Among them, attached water refers to liquid water (such as dew) that condenses or adheres to the surface of the wind turbine blades; morphology refers to the aggregation state and geometric outline of the attached water on the surface of the wind turbine blades; water droplet morphology refers to the regular, rounded water droplets formed by the hydrophobic repulsion of the attached water due to the hydrophobic repulsion of the anti-icing coating; water film morphology refers to the irregular water layer or puddle formed by the attached water spreading out and flat on the blade surface due to the failure of the local waterproof coating and the loss of hydrophobic repulsion.

[0052] S23, when the form of the attached water is determined to be water droplets, the outline area of ​​the corresponding attached water is obtained as the normal area of ​​the coating.

[0053] Understandably, for water droplets that are identified as adhering to the surface, an edge detection algorithm is called to extract their closed pixel boundaries, obtain the contour area of ​​the adhering water, and mark this contour area as a normal coating area (the contour area can be mapped to the actual 3D model or digital coordinate system of the wind turbine blade), indicating that the anti-icing coating at that location is in a healthy state.

[0054] Among them, the outline area refers to the outer edge of the attached water extracted in the early warning image; the normal coating area refers to the local area of ​​the wind turbine blade where the surface anti-icing coating has good hydrophobic properties and does not require maintenance.

[0055] S24, when the form of the attached water is determined to be a water film, the outline area of ​​the corresponding attached water is extracted as the coating warning area.

[0056] Understandably, for water adhering to the form of a water film, an edge detection algorithm is called to obtain the outline area of ​​the water adhering to the form, and this outline area is marked as a coating warning area (the outline area can be mapped to the actual three-dimensional model or digital coordinate system of the wind turbine blade), indicating that the anti-icing coating at this location needs to be inspected.

[0057] Among them, the coating early warning area refers to the local area of ​​the wind turbine blade where the surface anti-icing coating is identified as potentially damaged and requires inspection and maintenance based on the early warning detection mode.

[0058] In some embodiments, step S2 (generating an early warning maintenance queue based on the coating early warning area) includes S25: S25, calculate the warning area of ​​the warning zone on the coating of each wind turbine, obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the warning area, and obtain the warning maintenance queue.

[0059] It should be noted that wind farms typically contain a large number of wind turbines. If multiple wind turbines are found to have coating warning areas, maintenance personnel cannot address all warning areas simultaneously. This step prioritizes all wind turbines based on the warning area of ​​their coating warning zones, forming a warning maintenance queue. This ensures that limited human and material resources are prioritized for wind turbines with larger coating warning areas and a higher risk of subsequent icing.

[0060] Understandably, after obtaining the coating warning areas extracted in step S24, the total area of ​​all coating warning areas appearing on each wind turbine is first calculated using a statistical algorithm to obtain the warning area for each wind turbine. Then, the device numbers of these wind turbines with coating warning areas are retrieved, and all device numbers are sorted in descending order based on their warning areas. Finally, the device number with the largest warning area is placed at the beginning, and the one with the smallest area is placed at the end. This sorted list of numbers is then encapsulated to generate a warning maintenance queue and sent to the receiving terminal of the maintenance team.

[0061] Among them, the warning area refers to the total pixel area of ​​all coating warning areas on a single wind turbine; the equipment number refers to the unique digital ID or code used to uniquely identify each wind turbine in the wind farm; and the warning maintenance queue refers to an ordered set of data containing equipment numbers generated from the warning areas of each wind turbine.

[0062] S3, when the detection mode is determined to be the tracking detection mode, multiple tracking image groups are obtained by acquiring images of the wind turbine according to the cyclic acquisition strategy. Based on the tracking image groups, the coating failure area of ​​the wind turbine blade and the failure level of each coating failure area are determined, and a failure maintenance queue is generated according to the coating failure area and failure level.

[0063] It should be noted that damage to the anti-icing coating often occurs locally (e.g., due to bird strikes or localized peeling). In extremely cold climates, once a localized area of ​​damage occurs in the anti-icing coating, it will freeze first, eventually covering the entire blade with ice. Current technology can usually only take photos of large areas of icing on the wind turbine blades afterward, thus concluding that the blades are icy, but it cannot pinpoint which area of ​​the coating failed first. This step involves controlling a drone to monitor the wind turbine inspection team multiple times over time, under icing detection conditions, to obtain the icing process of the wind turbine blades. The earlier the icing area appears, the more severe the corresponding anti-icing coating failure. Therefore, this solution determines the failure level of the anti-icing coating based on the timing of icing.

[0064] Among them, the cyclic acquisition strategy refers to the scheduling scheme in which wind turbines are divided into several groups, and drones are used to acquire images of all wind turbines in the group at different time points; the coating failure area refers to the local area of ​​the wind turbine blade where the surface anti-icing coating is damaged and needs repair, as determined by the tracking and detection mode; the failure level refers to the level index used to quantify the severity of damage to the coating failure area.

[0065] In some embodiments, step S3 (acquiring multiple tracking image groups from the wind turbine according to a cyclic acquisition strategy) includes S31-S32: S31, divide the wind turbines in the wind farm into multiple wind turbine detection groups, and determine the UAV corresponding to each wind turbine detection group.

[0066] It's important to note that large wind farms often contain dozens or even hundreds of wind turbines. If only one drone were used to inspect all the turbines, it could take several hours to photograph them from the first to the last. In this situation, the large time difference between photographing different turbines would make it impossible to accurately capture the icing status of the turbines at the same time. This step involves dividing the wind turbine cluster into multiple smaller grid groups and equipping each group with a dedicated drone for parallel operation. This significantly reduces the time span of a single inspection, ensuring that the photographing of all turbines within the same group can be completed within a smaller time window. This keeps the time difference between photographing all turbines within the same group within a short error range, making the subsequently extracted icing areas from the same batch comparable and referential in terms of time series, thus providing a more accurate basis for determining the failure level.

[0067] Understandably, the process begins by obtaining the total number of wind turbines within the wind farm, their geographical coordinates, and the number of available drones. Next, using clustering algorithms or grid partitioning rules, wind turbines that are close together are grouped into the same set, thus generating multiple turbine detection groups (e.g., 10 wind turbines per group). After partitioning, each turbine detection group is uniquely bound to and assigned a corresponding drone, and the coordinates and flight paths of all wind turbines within that group are sent to that drone.

[0068] Among them, the wind turbine inspection group refers to an inspection unit composed of a portion of the wind turbines in the wind farm, which is used to shorten the inspection cycle.

[0069] S32, control the drone to collect images of the wind turbines in the corresponding wind turbine detection group at multiple time points, and obtain the tracking image group corresponding to each time point.

[0070] It's important to note that icing is a gradual process; not all damaged areas of the coating will freeze at a single moment. Therefore, this step involves the drone performing a complete scan of the entire group at a specific time (e.g., 10:00). This scan is completed within a shorter time window (e.g., 10:00-10:05), followed by another complete scan at a different time (e.g., 10:30). This scanning strategy is used to sequentially track and detect multiple time points. Subsequently, regardless of which wind turbine or wind turbine inspection group is involved, images acquired at a specific time point are grouped together into an image set for that time point.

[0071] Understandably, when the first time point is reached (e.g., 10:00), the drone is controlled to take off and sequentially photograph the first wind turbine in the corresponding wind turbine inspection group up to the last wind turbine. The images collected from each wind turbine inspection group are then packaged into a tracking image group corresponding to the first time point. When the second time point is reached (e.g., 10:30), the drone is again controlled to photograph the first wind turbine in the corresponding wind turbine inspection group up to the last wind turbine, completing the second scan of each wind turbine inspection group. All the collected images are then packaged into a tracking image group corresponding to the second time point. This process is repeated, controlling the drone to continuously perform inspections at multiple set time points, ultimately resulting in multiple tracking image groups corresponding to the inspection time points.

[0072] Here, "time point" refers to the moment when the drones are used to collect images of the wind turbine inspection group; "tracking image group" refers to the set of images collected by all drones from their respective wind turbine inspection groups at a certain time point.

[0073] In some embodiments, step S3 (determining the coating failure areas of the wind turbine blades and the failure levels of each coating failure area based on the tracking image set) includes S33-S35: S33, based on the tracking image group, extract the newly added icing areas at each time point as the coating failure areas.

[0074] It's important to note that icing is a continuous process that spreads over time. Extracting only all icing areas at a single point in time would confuse previously iced areas with newly formed ones, making it impossible to accurately identify damaged coating areas. This step extracts only the newly formed icing areas relative to the previous time point, ensuring that each extracted coating failure area corresponds to its actual icing time, thus allowing for accurate failure level classification.

[0075] Understandably, when analyzing the tracking image group at the current time point (e.g., 10:30), the tracking image group at the previous time point (e.g., 10:00) will be used as a reference. Using image difference algorithms or connected component comparison techniques, the icing areas that already existed at the previous time point will be filtered out, and the newly appearing icing areas (i.e., newly added icing areas) in this batch of images at the current time point will be extracted.

[0076] In addition, when analyzing the tracking image group at the first time point, the icing area in the tracking image group is directly extracted as the newly added icing area.

[0077] Among them, the newly added icing area refers to the outline of the icing part extracted from the tracking image group at a specific time point, which is unique to that time point.

[0078] like Figure 2 and Figure 3 As shown, Figure 2 This is a schematic diagram showing newly formed icing areas on a wind turbine at the first time point. Figure 2 New icing areas in Figure 2 The striped frame (in the image) appears at the first time point; Figure 3 This is a schematic diagram showing the newly formed icing area on the same wind turbine at the second time point. Figure 3 A difference appears in the middle. Figure 2 The newly added icing area corresponding to the second time point ( Figure 3 (the striped circle in the middle), that is Figure 2 The newly added icing areas in the data cannot be used as... Figure 3 The newly added icing area.

[0079] S34. Construct a failure region group corresponding to each time point based on the coating failure region, and sort the failure region group in ascending order based on the time point to obtain the failure sequence.

[0080] It should be noted that, as time progresses naturally, this invention dynamically extracts newly formed icing areas corresponding to each time point. Furthermore, the more severely damaged the anti-icing coating, the shorter its resistance to extreme cold, and the sooner it will freeze. Therefore, this step packages all coating failure areas extracted at the same time point into failure area groups, and arranges all failure area groups in ascending order according to the progression of time. This is done to organize the originally chaotic localized icing phenomena into a clearer timeline (i.e., failure sequence). On this timeline, the earlier the failure area group appears, the more severe the coating damage corresponding to the failed coating areas within it.

[0081] Understandably, the coating failure areas extracted at each time point are stored in a corresponding group. For example, all coating failure areas extracted at the first time point are stored in failure area group A, those extracted at the second time point are stored in failure area group B, and so on, resulting in failure area groups for each time point. Then, based on the order of these time points, all failure area groups are sorted in ascending order (e.g., the failure area group corresponding to 10:00 is first, and the failure area group corresponding to 10:30 is second), forming a failure sequence that directly reflects the degree of coating damage based on time.

[0082] Among them, the failure area group refers to the collection formed by gathering all the coating failure areas extracted at a certain time point; the failure sequence refers to the sequence formed by sorting the failure area groups corresponding to the time points in ascending order according to the progression of time points.

[0083] S35, determine the failure level of the coating failure area in each failure area group based on the failure sequence.

[0084] It should be noted that the degree of damage to the anti-icing coating is also objectively reflected in the length of time it can resist freezing. The area that freezes earliest indicates the most severe damage to the coating, while the area that freezes latest indicates the least damage.

[0085] Understandably, the failure sequence generated in step S34 is retrieved, and failure levels are assigned to the coating failure areas in each failure area group. The failure levels decrease sequentially according to the order of the failure area groups in the failure sequence. Specifically, the process begins by traversing from the very beginning of the failure sequence (i.e., the failure area group corresponding to the first time point). For the failure area group at the top of the sequence, the coating failure areas within it are determined to have the worst anti-icing capability, and therefore, all coating failure areas within that group are assigned the highest failure level (e.g., Level 1 failure). Subsequently, the next failure area group in the failure sequence (e.g., the failure area group corresponding to the second time point) is read, and all coating failure areas within that group are assigned a secondary failure level (e.g., Level 2 failure), and so on. Based on the failure sequence, the corresponding failure levels can be matched to dozens or even hundreds of coating failure areas across the entire site relatively efficiently and accurately.

[0086] In some embodiments, step S3 (generating a failure maintenance queue based on the coating failure area and failure level) includes S36-S39: S36, determine the failure area of ​​the coating failure region corresponding to each failure level on the wind turbine.

[0087] Understandably, the process involves extracting all coating failure areas on a wind turbine marked as having the highest failure level (e.g., Level 1 failure), and then calculating the total area (failure area) of the coating failure areas corresponding to that highest failure level using an integral algorithm or contour pixel statistics. Similarly, the total area (failure area) of the coating failure areas for the next lower failure level (e.g., Level 2 failure) is calculated. This process continues until the failure area corresponding to each failure level on the wind turbine is determined.

[0088] The failure area refers to the total area of ​​the coating failure zone at a specific failure level on a single wind turbine, as calculated from the data.

[0089] S37, retrieve the preset reference coefficients corresponding to each failure level, calculate the product of each failure area and the corresponding preset reference coefficient, and obtain the failure value.

[0090] Understandably, a preset reference coefficient matching each failure level is retrieved; for example, a first-level failure level corresponds to a first-level reference coefficient, and a second-level failure level corresponds to a second-level reference coefficient. Then, the failure area corresponding to each failure level extracted from the wind turbine is multiplied by the corresponding preset reference coefficient to obtain multiple failure values. For example, the failure area corresponding to a first-level failure level is multiplied by the first-level reference coefficient to obtain the failure value corresponding to the first-level failure level; the failure area corresponding to a second-level failure level is multiplied by the second-level reference coefficient to obtain the failure value corresponding to the second-level failure level.

[0091] Among them, the preset benchmark coefficient refers to the pre-configured mathematical weighting parameter corresponding to each failure level. Generally, the higher the failure level, the larger the corresponding preset benchmark coefficient. The failure value refers to the quantitative value that represents the degree of damage and danger of a failure level, calculated by multiplying the failure area of ​​a certain failure level by the corresponding preset benchmark coefficient.

[0092] S38 sums up the multiple failure values ​​corresponding to the wind turbine to obtain the operation and maintenance value of the wind turbine.

[0093] Understandably, for a given wind turbine, the failure values ​​corresponding to all failure levels are extracted, and an accumulation and summation algorithm is performed on all failure values. The sum is then used as the operation and maintenance value for that wind turbine. This process is repeated to obtain the operation and maintenance value for each wind turbine.

[0094] Among them, the operation and maintenance value refers to the total quantitative score obtained by summing the failure values ​​of the wind turbine, which is used to assess the overall damage and danger of the anti-icing coating of a single wind turbine (the higher the operation and maintenance value, the more dangerous the state of the corresponding wind turbine).

[0095] S39, obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the maintenance value, and obtain the failure maintenance queue.

[0096] Understandably, the process involves retrieving the device numbers of all wind turbines in the site, sorting them in descending order based on the maintenance values ​​calculated in step S38, and finally encapsulating this sorted list of device numbers into a failure maintenance queue, which is then pushed to the maintenance management platform or mobile terminal. In this failure maintenance queue, the device number with the highest maintenance value is placed at the front, with those having lower values ​​placed at the back.

[0097] Among them, the failure maintenance queue refers to an ordered set of data containing equipment numbers generated based on the maintenance values ​​of each wind turbine.

[0098] See Figure 4 This is a schematic diagram of a wind farm operation and maintenance scheduling system provided in an embodiment of the present invention. The system includes: The judgment module is used to determine the detection mode based on the environmental information of the wind farm, and the detection mode includes an early warning detection mode and a tracking detection mode. The early warning module is used to acquire images of each wind turbine to obtain early warning images when the detection mode is early warning detection mode, determine the coating early warning area of ​​the wind turbine blades based on the early warning images, and generate an early warning maintenance queue based on the coating early warning area. The failure module is used to determine that when the detection mode is the tracking detection mode, it acquires multiple tracking image groups of the wind turbine according to the cyclic acquisition strategy, determines the coating failure area of ​​the wind turbine blade and the failure level of each coating failure area based on the tracking image group, and generates a failure operation and maintenance queue according to the coating failure area and failure level.

[0099] See Figure 5 This is a schematic diagram of the hardware structure of a computer device 40 provided in an embodiment of the present invention. The computer device 40 includes: a processor 41, a memory 42, and a computer program; wherein... The memory 42 is used to store the computer program, and the memory may also be flash memory. The computer program is, for example, an application program or functional module that implements the above method.

[0100] The processor 41 is configured to execute the computer program stored in the memory to implement the various steps performed by the device in the above method. For details, please refer to the relevant descriptions in the preceding method embodiments.

[0101] Alternatively, the memory 42 can be either standalone or integrated with the processor 41.

[0102] When the memory 42 is a device independent of the processor 41, the device may further include: Bus 43 is used to connect the memory 42 and the processor 41.

[0103] The present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, is used to implement the methods provided in the various embodiments described above.

[0104] The readable storage medium can be a computer storage medium or a communication medium. A communication medium includes any medium that facilitates the transfer of computer programs from one location to another. A computer storage medium can be any available medium accessible to a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application-Specific Integrated Circuit (ASIC). Alternatively, the ASIC can be located in a user equipment. Of course, the processor and the readable storage medium can also exist as discrete components in a communication device. The readable storage medium can be a read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0105] The present invention also provides a program product including executable instructions stored in a readable storage medium. At least one processor of the device can read the executable instructions from the readable storage medium, and the at least one processor executes the executable instructions to cause the device to implement the methods provided in the various embodiments described above.

[0106] In the embodiments of the above-described device, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly manifested as execution by a hardware processor, or execution by a combination of hardware and software modules within the processor.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wind farm operation and maintenance scheduling method, characterized in that, include: The detection mode is determined based on the environmental information of the wind farm, and the detection mode includes an early warning detection mode and a tracking detection mode. When the detection mode is determined to be the early warning detection mode, images are acquired from each wind turbine to obtain early warning images. The early warning area of ​​the wind turbine blade coating is determined based on the early warning images, and an early warning maintenance queue is generated based on the coating early warning area. When the detection mode is determined to be the tracking detection mode, multiple tracking image groups are obtained by acquiring images of the wind turbine according to the cyclic acquisition strategy. Based on the tracking image groups, the coating failure areas of the wind turbine blades and the failure levels of each coating failure area are determined, and a failure maintenance queue is generated according to the coating failure areas and failure levels.

2. The method according to claim 1, characterized in that, The detection mode is determined based on the environmental information of the wind farm. The detection mode includes an early warning detection mode and a tracking detection mode, including: Obtain preset detection conditions, including water attachment detection conditions and freezing detection conditions; The environmental information of the wind farm is retrieved, and the environmental information is compared with the detection conditions. If the environmental information meets the water attachment detection conditions, the detection mode is determined to be the early warning detection mode. When the environmental information meets the icing detection conditions, the detection mode is set to tracking detection mode.

3. The method according to claim 1, characterized in that, The process of acquiring early warning images from each wind turbine and determining the early warning area for the coating of the wind turbine blades based on these images includes: Control the drone to collect images of the wind turbines and obtain early warning images for each wind turbine; Based on the early warning image, the water adhering to the wind turbine blades is identified, and the morphology of the adhering water is determined, including water droplet morphology and water film morphology. When the form of the adhering water is determined to be water droplets, the outline area of ​​the corresponding adhering water is obtained as the normal area of ​​the coating. When the form of the attached water is determined to be a water film, the outline area of ​​the corresponding attached water is extracted as the coating warning area.

4. The method according to claim 3, characterized in that, The generation of the early warning operation and maintenance queue based on the coating early warning area includes: Calculate the warning area of ​​the coating warning zone on each wind turbine, obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the warning area, and obtain the warning maintenance queue.

5. The method according to claim 1, characterized in that, The process of acquiring images of the wind turbine according to a cyclic acquisition strategy to obtain multiple tracking image groups includes: The wind turbines in the wind farm are divided into multiple wind turbine inspection groups, and the corresponding UAVs for each wind turbine inspection group are determined. The drone is controlled to acquire images of the wind turbines in the corresponding wind turbine detection group at multiple time points, resulting in tracking image groups corresponding to each time point.

6. The method according to claim 5, characterized in that, The determination of the coating failure areas of the wind turbine blades and the failure level of each coating failure area based on the tracking image group includes: Based on the tracking image group, newly added icing areas at each time point are extracted as coating failure areas; Based on the coating failure areas, a failure area group corresponding to each time point is constructed, and the failure area group is sorted in ascending order based on the time point to obtain the failure sequence; The failure level of the coating failure area in each failure area group is determined based on the failure sequence.

7. The method according to claim 6, characterized in that, The step of generating a failure maintenance queue based on the coating failure area and failure level includes: Determine the failure area of ​​the coating failure region corresponding to each failure level on the wind turbine; Retrieve the preset reference coefficients corresponding to each failure level, calculate the product of each failure area and the corresponding preset reference coefficient, and obtain the failure value; The operation and maintenance value of the wind turbine is obtained by summing up the multiple failure values ​​corresponding to the wind turbine. Obtain the equipment number of each wind turbine, sort the equipment numbers in descending order based on the maintenance value, and obtain the failure maintenance queue.

8. A wind farm operation and maintenance dispatch system, characterized in that, include: The judgment module is used to determine the detection mode based on the environmental information of the wind farm, and the detection mode includes an early warning detection mode and a tracking detection mode. The early warning module is used to acquire images of each wind turbine to obtain early warning images when the detection mode is early warning detection mode, determine the coating early warning area of ​​the wind turbine blades based on the early warning images, and generate an early warning maintenance queue based on the coating early warning area. The failure module is used to determine that when the detection mode is the tracking detection mode, it acquires multiple tracking image groups of the wind turbine according to the cyclic acquisition strategy, determines the coating failure area of ​​the wind turbine blade and the failure level of each coating failure area based on the tracking image group, and generates a failure operation and maintenance queue according to the coating failure area and failure level.

9. A computer device, characterized in that, include: The method comprises a memory, a processor, and a computer program, the computer program being stored in the memory, and the processor executing the computer program to perform the method according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, is used to implement the method described in any one of claims 1 to 7.

Images