Wind turbine generator state evaluation method and system based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of wind power, and particularly discloses a wind turbine generator state evaluation method and system based on an unmanned aerial vehicle, and the method comprises the steps: obtaining the image data of a wind turbine generator, and segmenting the image data of the wind turbine generator to obtain the image data of a wind turbine generator; determining blade image data according to the fan image data, and determining a blade defect type according to the blade image data; and calculating a unit characteristic value of the wind turbine generator according to the blade defect type, and evaluating the state of the wind turbine generator according to the unit characteristic value of the wind turbine generator. According to the method, the defect type of the fan blade can be accurately identified according to the image data of the wind turbine generator, so that the state monitoring of the wind turbine generator is realized, and the normal operation of the wind turbine generator is ensured.

Description

Technical Field

[0001] The present application relates to the field of wind power technology, and more specifically, to a method and system for evaluating the status of a wind turbine based on an unmanned aerial vehicle. Background Art

[0002] In recent years, wind energy has become increasingly prominent in the world's energy structure, and wind power will gradually become the third largest conventional energy source after thermal power and hydropower. As my country's large-scale wind turbine construction plans have been launched one after another and the warranty period of most wind turbines currently in operation has gradually exceeded or is about to exceed, the current situation of high failure rate and high operation and maintenance costs has attracted more and more attention.

[0003] Since wind turbines work in harsh natural environments for a long time and are affected by normal and extreme temperatures, solar radiation, rainfall, snow, salt spray, dust, terrain contours and other factors, the insulation strength, fatigue strength and operating performance of each component will inevitably decline gradually with changes in the operating environment and operating time. If the status of the wind turbine cannot be evaluated in a timely manner, failures are likely to occur. Summary of the invention

[0004] The present invention provides a method and system for evaluating the status of a wind turbine generator set based on an unmanned aerial vehicle, which is used to solve the problem of how to timely monitor the status of a wind turbine generator set in the prior art, including:

[0005] Acquire wind turbine image data, segment the wind turbine image data, and obtain wind turbine image data;

[0006] Determine blade image data according to the fan image data, and determine the blade defect type according to the blade image data;

[0007] The unit characteristic value of the wind turbine is calculated according to the blade defect type, and the state of the wind turbine is evaluated according to the unit characteristic value of the wind turbine.

[0008] Furthermore, the segmenting of the wind turbine image data includes:

[0009] Grayscale processing is performed on the image data of the wind turbine generator set to obtain a grayscale image of the wind turbine generator set;

[0010] Obtain the grayscale value of each pixel in the grayscale image of the wind turbine, perform clustering on each pixel according to the grayscale value, and determine the initial seed point according to the clustering result;

[0011] Calculate the similarity between the initial seed point and each pixel in the preset area, and set the pixel with a similarity less than a first preset threshold as a new seed point;

[0012] Continue to detect the remaining pixels based on the new seed point until the area can no longer grow, and obtain the wind turbine image data.

[0013] Furthermore, clustering each pixel point according to the gray value includes:

[0014] The k value is determined according to the number of pixels in the grayscale image of the wind turbine, and k initial cluster centers are randomly selected;

[0015] Calculate the Euclidean distance from each pixel to the initial cluster center, and divide each pixel into the corresponding cluster according to the Euclidean distance from each pixel to the initial cluster center;

[0016] Calculate the grayscale average of all pixels in each cluster, and update the cluster center according to the grayscale average of all pixels in each cluster;

[0017] The above steps are iterated repeatedly until the cluster center no longer changes, and k clusters of the grayscale image of the wind turbine are obtained.

[0018] Further, determining the initial seed point according to the clustering result includes:

[0019] Obtain the cluster center value of the target cluster in the grayscale image of the wind turbine generator set, and calculate the absolute value of the difference between the target cluster and the cluster center values ​​of its adjacent clusters;

[0020] The average absolute value of the difference between the target cluster and its adjacent clusters is calculated, and the cluster center of the target cluster whose average value is greater than a second preset threshold is set as the initial seed point.

[0021] Furthermore, the calculating of the similarity between the initial seed point and each pixel point in its preset area includes:

[0022] The similarity of each pixel is calculated according to the similarity calculation formula, and the similarity calculation formula is specifically:

[0023]

[0024] Among them, S ij is the similarity between the i-th seed point and the j-th pixel point in its preset area, D ij is the Euclidean distance between the i-th seed point and the j-th pixel point in its preset area, G i is the gray value of the i-th seed point, G j is the gray value of the j-th pixel, C max is the maximum gray value of the cluster to which the i-th seed point belongs, C min is the minimum grayscale value of the cluster to which the i seed point belongs.

[0025] Further, determining the blade image data according to the wind turbine image data includes:

[0026] Acquire historical wind turbine image data and corresponding blade image data collected by the UAV, and pre-process the historical wind turbine image data and the corresponding blade image data;

[0027] Establishing a training sample set based on the preprocessed historical fan image data and the corresponding blade image data, and establishing an initial blade recognition model based on the training sample set;

[0028] The initial leaf recognition model is trained according to the training sample set to obtain a leaf recognition model;

[0029] The current fan image data is input into the trained blade recognition model to obtain the blade image data corresponding to the current fan image data.

[0030] Further, determining the blade defect type according to the blade image data includes:

[0031] Perform edge detection on the blade image data based on the canny edge detection algorithm, and determine the blade defect area according to the edge detection results;

[0032] Obtaining the maximum length line segment and the maximum width line segment of the blade defect area, connecting the endpoints of the maximum length line segment and the maximum width line segment of the blade defect area in sequence, and obtaining a defect fitting graph;

[0033] A blade defect type library is obtained, and the defect fitting graph is matched with the defect type in the blade defect type library to obtain the defect type corresponding to the defect fitting graph.

[0034] Further, the calculating of the wind turbine unit characteristic value according to the blade defect type includes:

[0035] Obtaining the grayscale values ​​of the pixels of the defect fitting graph, and calculating the average grayscale value of the defect fitting graph according to the grayscale values ​​of the pixels of the defect fitting graph;

[0036] Calculate the ratio of the area value of the defect fitting graph to the average gray value to obtain the state characteristic value of the defect fitting graph;

[0037] Assign a preset state weight to each defect fitting graph according to the defect type, and multiply the state characteristic value of the defect fitting graph by the corresponding preset state weight to obtain a corrected state characteristic value;

[0038] The sum of the state characteristic values ​​after correction of all defect fitting graphs in the blade image data is calculated to obtain the unit characteristic value.

[0039] Further, the evaluating the state of the wind turbine generator set according to the wind turbine generator set characteristic value includes:

[0040] Obtaining a preset standard unit characteristic value, and calculating the difference between the unit characteristic value and the standard unit characteristic value;

[0041] Determine whether the difference between the characteristic value of the unit and the characteristic value of the standard unit is greater than a third preset threshold value, and if the difference between the characteristic value of the unit and the characteristic value of the standard unit is greater than the third preset threshold value, determine that the wind turbine is in an abnormal state;

[0042] If the difference between the characteristic value of the unit and the characteristic value of the standard unit is less than the third preset threshold, it is determined that the wind turbine unit is in a normal state.

[0043] In order to achieve the above object, the present invention also provides a wind turbine status assessment system based on a drone, comprising:

[0044] An acquisition module is used to acquire wind turbine image data, segment the wind turbine image data, and obtain wind turbine image data;

[0045] An identification module, used to determine blade image data according to the fan image data, and determine the blade defect type according to the blade image data;

[0046] The evaluation module is used to calculate the unit characteristic value of the wind turbine according to the blade defect type, and evaluate the state of the wind turbine according to the unit characteristic value of the wind turbine.

[0047] The beneficial effects of the present invention are:

[0048] By applying the above technical scheme, the present invention collects image data of the wind turbine set, accurately segments the image data of the wind turbine set, obtains accurate blade image data, and then identifies the defect type of the blade image data, thereby realizing timely monitoring of the status of the wind turbine set and ensuring the normal operation of the wind turbine set. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 A schematic flow chart of a method for evaluating a wind turbine status based on a drone according to an embodiment of the present invention is shown;

[0051] Figure 2 The figure shows the overall structure of a wind turbine status assessment system based on a drone according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0053] The present application embodiment provides a method for evaluating the status of a wind turbine generator system based on a drone, such as Figure 1 As shown, including:

[0054] S101, acquiring wind turbine image data, segmenting the wind turbine image data, and obtaining wind turbine image data;

[0055] In some embodiments of the present application, the segmentation of wind turbine image data includes: grayscale processing of the wind turbine image data to obtain a grayscale image of the wind turbine; obtaining the grayscale value of each pixel in the grayscale image of the wind turbine, clustering each pixel according to the grayscale value, and determining an initial seed point according to the clustering result; calculating the similarity between the initial seed point and each pixel within its preset range, and setting the pixel whose similarity is less than a first preset threshold as a new seed point; continuing to detect the remaining pixels according to the new seed point until the area can no longer grow, thereby obtaining wind turbine image data.

[0056] In this embodiment, the wind turbine image data is collected by a drone, and the wind turbine image in the wind turbine image data is segmented based on a region growing algorithm, thereby achieving accurate segmentation of the wind turbine image data.

[0057] In some embodiments of the present application, the clustering processing of each pixel point according to the grayscale value includes: determining the k value according to the number of pixels in the grayscale image of the wind turbine generator set, and randomly selecting k initial clustering centers; calculating the Euclidean distance of each pixel point to the initial clustering center, and dividing each pixel point into a corresponding cluster according to the Euclidean distance of each pixel point to the initial clustering center; calculating the grayscale average value of all pixels in each cluster, and updating the clustering center according to the grayscale average value of all pixels in each cluster; repeating the above steps until the clustering center no longer changes, and obtaining k clustering clusters of the grayscale image of the wind turbine generator set.

[0058] In this embodiment, the grayscale image of the wind turbine is clustered based on the k-means clustering algorithm, and the grayscale image of the wind turbine is divided into k clusters. The k value is determined by the number of pixels in the grayscale image of the wind turbine. The larger the number of pixels, the higher the corresponding k value.

[0059] In some embodiments of the present application, determining the initial seed point based on the clustering results includes: obtaining the cluster center value of the target cluster cluster in the grayscale image of the wind turbine group, and calculating the absolute value of the difference between the cluster center values ​​of the target cluster cluster and its adjacent cluster clusters; statistically calculating the average value of the absolute value of the difference between the cluster center values ​​of the target cluster cluster and its adjacent cluster clusters, and setting the cluster center of the target cluster cluster whose average value is greater than a second preset threshold as the initial seed point.

[0060] In this embodiment, the initial seed point is set by the cluster center grayscale values ​​of the target cluster and the adjacent clusters in the wind turbine generator grayscale image, so as to facilitate the subsequent regional growth of the wind turbine generator grayscale image.

[0061] In some embodiments of the present application, the calculating of the similarity between the initial seed point and each pixel point in its preset area includes: calculating the similarity of each pixel point according to a similarity calculation formula, wherein the similarity calculation formula is specifically:

[0062]

[0063] Among them, S ij is the similarity between the i-th seed point and the j-th pixel point in its preset area, D ij is the Euclidean distance between the i-th seed point and the j-th pixel point in its preset area, G i is the gray value of the i-th seed point, G j is the gray value of the j-th pixel, C max is the maximum gray value of the cluster to which the i-th seed point belongs, C min is the minimum grayscale value of the cluster to which the i seed point belongs.

[0064] In this embodiment, the similarity between two points is calculated by the Euclidean distance between the seed point and the domain pixel, the grayscale difference and the difference between the maximum and minimum values ​​in the corresponding cluster.

[0065] In some embodiments of the present application, determining blade image data based on wind turbine image data includes: acquiring historical wind turbine image data and corresponding blade image data collected by a drone, and preprocessing the historical wind turbine image data and the corresponding blade image data; establishing a training sample set based on the preprocessed historical wind turbine image data and the corresponding blade image data, and establishing an initial blade recognition model based on the training sample set; training the initial blade recognition model based on the training sample set to obtain a blade recognition model; and inputting current wind turbine image data into the trained blade recognition model to obtain blade image data corresponding to the current wind turbine image data.

[0066] In this embodiment, the blade recognition model is trained by using the historical wind turbine image data and the corresponding blade image data collected by the drone, thereby achieving accurate recognition of the blade image data through the blade recognition model.

[0067] S102, determining blade image data according to the wind turbine image data, and determining a blade defect type according to the blade image data;

[0068] In some embodiments of the present application, determining the blade defect type based on blade image data includes: performing edge detection on the blade image data based on a canny edge detection algorithm, and determining the blade defect area based on the edge detection result; obtaining the maximum length line segment and the maximum width line segment of the blade defect area, and connecting the endpoints of the maximum length line segment and the maximum width line segment of the blade defect area in sequence to obtain a defect fitting graph; obtaining a blade defect type library, matching the defect fitting graph with the defect type in the blade defect type library, and obtaining the defect type corresponding to the defect fitting graph.

[0069] In this embodiment, the canny edge detection algorithm is used to detect the remaining edge areas in the blade image data except the blade edge, so as to obtain the blade defect area, and the maximum length line segment of each blade defect area is connected in sequence with the endpoints of the maximum width line segment, and a graph is fitted for each blade defect area. The fitted defect fitting graph is matched with the blade defect type library, and then the defect type corresponding to the defect fitting graph is obtained.

[0070] S103, calculating a unit characteristic value of the wind turbine generator set according to the blade defect type, and evaluating the state of the wind turbine generator set according to the unit characteristic value of the wind turbine generator set.

[0071] In some embodiments of the present application, the calculation of the unit characteristic value of the wind turbine set according to the blade defect type includes: obtaining the pixel grayscale value of the defect fitting graphic, and calculating the average grayscale value of the defect fitting graphic according to the pixel grayscale value of the defect fitting graphic; calculating the ratio of the area value of the defect fitting graphic to the average grayscale value to obtain the state characteristic value of the defect fitting graphic; assigning a preset state weight to each defect fitting graphic according to the defect type, multiplying the state characteristic value of the defect fitting graphic by the corresponding preset state weight to obtain a corrected state characteristic value; and calculating the sum of the corrected state characteristic values ​​of all defect fitting graphics in the blade image data to obtain the unit characteristic value.

[0072] In this embodiment, the state characteristic value of the defect fitting graph is obtained by calculating the ratio of the area value of the defect fitting graph to the average gray value. The defect types of the defect fitting graph specifically include: holes, fractures, rust, peeling, missing blocks and other defects. A preset state weight is assigned to each defect fitting graph according to the defect type. The more serious the defect type, the higher the corresponding preset state weight. The state characteristic value of the defect fitting graph is multiplied by the corresponding preset state weight to obtain the corrected state characteristic value, and then the unit characteristic value is calculated.

[0073] In some embodiments of the present application, the state of the wind turbine set is evaluated based on the unit characteristic value of the wind turbine set, including: obtaining a preset standard unit characteristic value, and calculating the difference between the unit characteristic value and the standard unit characteristic value; judging whether the difference between the unit characteristic value and the standard unit characteristic value is greater than a third preset threshold value, if the difference between the unit characteristic value and the standard unit characteristic value is greater than the third preset threshold value, judging that the wind turbine set is in an abnormal state; if the difference between the unit characteristic value and the standard unit characteristic value is less than the third preset threshold value, judging that the wind turbine set is in a normal state.

[0074] In this embodiment, the state of the wind turbine generator set is evaluated by the difference between the characteristic value of the wind turbine generator set and the characteristic value of the standard wind turbine generator set, so that the state of the wind turbine generator set can be monitored in a timely manner and the normal operation of the wind turbine generator set can be ensured.

[0075] Based on the same technical concept, such as Figure 2 As shown, the present invention also provides a wind turbine status assessment system based on a UAV, including: an acquisition module, used to acquire wind turbine image data, segment the wind turbine image data, and obtain wind turbine image data; an identification module, used to determine blade image data according to the wind turbine image data, and determine the blade defect type according to the blade image data; an assessment module, used to calculate the unit characteristic value of the wind turbine according to the blade defect type, and assess the status of the wind turbine according to the unit characteristic value of the wind turbine.

[0076] By applying the above technical scheme, the present invention obtains wind turbine image data, segments the wind turbine image data, and obtains fan image data; determines blade image data according to the fan image data, and determines the blade defect type according to the blade image data; calculates the unit characteristic value of the wind turbine according to the blade defect type, and evaluates the state of the wind turbine according to the unit characteristic value of the wind turbine. The present invention can accurately identify the defect type of the wind turbine blade based on the image data of the wind turbine, thereby realizing the state monitoring of the wind turbine and ensuring the normal operation of the unit.

[0077] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0078] Those skilled in the art will appreciate that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A wind turbine status assessment method based on drone, characterized in that: The method comprises: Acquire wind turbine image data, segment the wind turbine image data, and obtain wind turbine image data; Determine blade image data according to the fan image data, and determine the blade defect type according to the blade image data; The unit characteristic value of the wind turbine is calculated according to the blade defect type, and the state of the wind turbine is evaluated according to the unit characteristic value of the wind turbine.

2. The wind turbine status assessment method based on drone according to claim 1 is characterized in that: The segmenting of the wind turbine image data comprises: Grayscale processing is performed on the image data of the wind turbine generator set to obtain a grayscale image of the wind turbine generator set; Obtain the grayscale value of each pixel in the grayscale image of the wind turbine, perform clustering on each pixel according to the grayscale value, and determine the initial seed point according to the clustering result; Calculate the similarity between the initial seed point and each pixel in the preset area, and set the pixel with a similarity less than a first preset threshold as a new seed point; Continue to detect the remaining pixels based on the new seed point until the area can no longer grow, and obtain the wind turbine image data.

3. The wind turbine status assessment method based on drone according to claim 2 is characterized in that: The clustering of each pixel point according to the gray value includes: The k value is determined according to the number of pixels in the grayscale image of the wind turbine, and k initial cluster centers are randomly selected; Calculate the Euclidean distance from each pixel to the initial cluster center, and divide each pixel into the corresponding cluster according to the Euclidean distance from each pixel to the initial cluster center; Calculate the grayscale average of all pixels in each cluster, and update the cluster center according to the grayscale average of all pixels in each cluster; The above steps are iterated repeatedly until the cluster center no longer changes, and k clusters of the grayscale image of the wind turbine are obtained.

4. The wind turbine status assessment method based on UAV according to claim 3 is characterized in that: The step of determining the initial seed point according to the clustering result includes: Obtain the cluster center value of the target cluster in the grayscale image of the wind turbine generator set, and calculate the absolute value of the difference between the target cluster and the cluster center values ​​of its adjacent clusters; The average absolute value of the difference between the target cluster and its adjacent clusters is calculated, and the cluster center of the target cluster whose average value is greater than a second preset threshold is set as the initial seed point.

5. The wind turbine status assessment method based on UAV according to claim 4 is characterized in that: The calculation of the similarity between the initial seed point and each pixel point in its preset area includes: The similarity of each pixel is calculated according to the similarity calculation formula, and the similarity calculation formula is specifically: Among them, S ij is the similarity between the i-th seed point and the j-th pixel point in its preset area, D ij is the Euclidean distance between the i-th seed point and the j-th pixel point in its preset area, G i is the gray value of the i-th seed point, G j is the gray value of the j-th pixel, C max is the maximum gray value of the cluster to which the i-th seed point belongs, C min is the minimum grayscale value of the cluster to which the i seed point belongs.

6. The wind turbine status assessment method based on UAV according to claim 5 is characterized in that: Determining blade image data according to wind turbine image data includes: Acquire historical wind turbine image data and corresponding blade image data collected by the UAV, and pre-process the historical wind turbine image data and the corresponding blade image data; Establishing a training sample set based on the preprocessed historical fan image data and the corresponding blade image data, and establishing an initial blade recognition model based on the training sample set; The initial leaf recognition model is trained according to the training sample set to obtain a leaf recognition model; The current fan image data is input into the trained blade recognition model to obtain the blade image data corresponding to the current fan image data.

7. The wind turbine status assessment method based on UAV according to claim 6 is characterized in that: Determining the blade defect type according to the blade image data includes: Perform edge detection on the blade image data based on the canny edge detection algorithm, and determine the blade defect area according to the edge detection results; Obtaining the maximum length line segment and the maximum width line segment of the blade defect area, connecting the endpoints of the maximum length line segment and the maximum width line segment of the blade defect area in sequence, and obtaining a defect fitting graph; A blade defect type library is obtained, and the defect fitting graph is matched with the defect type in the blade defect type library to obtain the defect type corresponding to the defect fitting graph.

8. The method for evaluating the status of a wind turbine generator system based on a drone according to claim 7, characterized in that: The step of calculating the characteristic value of the wind turbine generator set according to the blade defect type includes: Obtaining the grayscale values ​​of the pixels of the defect fitting graph, and calculating the average grayscale value of the defect fitting graph according to the grayscale values ​​of the pixels of the defect fitting graph; Calculate the ratio of the area value of the defect fitting graph to the average gray value to obtain the state characteristic value of the defect fitting graph; Assign a preset state weight to each defect fitting graph according to the defect type, and multiply the state characteristic value of the defect fitting graph by the corresponding preset state weight to obtain a corrected state characteristic value; The sum of the state characteristic values ​​after correction of all defect fitting graphics in the blade image data is calculated to obtain the unit characteristic value.

9. The wind turbine status assessment method based on UAV according to claim 8 is characterized in that: The step of evaluating the state of the wind turbine generator set according to the characteristic value of the wind turbine generator set comprises: Obtaining a preset standard unit characteristic value, and calculating the difference between the unit characteristic value and the standard unit characteristic value; Determine whether the difference between the characteristic value of the unit and the characteristic value of the standard unit is greater than a third preset threshold value, and if the difference between the characteristic value of the unit and the characteristic value of the standard unit is greater than the third preset threshold value, determine that the wind turbine is in an abnormal state; If the difference between the characteristic value of the unit and the characteristic value of the standard unit is less than the third preset threshold, it is determined that the wind turbine unit is in a normal state.

10. A wind turbine status assessment system based on drones, characterized in that: include: An acquisition module is used to acquire wind turbine image data, segment the wind turbine image data, and obtain wind turbine image data; An identification module, used to determine blade image data according to the fan image data, and determine the blade defect type according to the blade image data; The evaluation module is used to calculate the unit characteristic value of the wind turbine according to the blade defect type, and evaluate the state of the wind turbine according to the unit characteristic value of the wind turbine.

Images