Windmill blade repair support device and windmill blade repair support method

Abstract

To provide a windmill blade repair support device or the like from a damage image of a windmill blade.SOLUTION: A windmill blade repair support device 100 includes: an input part 110 for inputting facility information related to a wind turbine generator system and a photographed image of a blade 73; a damage portion extraction part 121 for extracting a damage image showing an image of a damaged portion from the photographed image of the blade 73; a repair method database 150 previously accumulating a plurality of damage images as reference damage images, and storing by associating a repair method for the reference damage images; a similar image retrieval part 122 for retrieving the most similar reference damage image from the reference damage images accumulated in the repair method database 150 based on the extracted damage image: and a repair method information generation part 120 for extracting a repair method in association with the retrieved reference damage image from the repair method database 150 to combine with the damage image, and generating repair method information for the damage specified by the damage image.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a wind turbine blade repair support device and a wind turbine blade repair support method.

Background Art

[0002] Patent Document 1 discloses a structure display device that displays divided images obtained by dividing a structure into a plurality of imaging ranges, assigns identification information to each of the divided images, and stores one or more pieces of management information in association with the identification information. And a control unit that receives an input of extraction conditions combining the identification information or the identification information and the management information, extracts the divided images that match the extraction conditions from the storage unit, and displays them on a display unit. A structure display device is described.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the structure display device described in Patent Document 1, the main problem was to efficiently create management information useful for screen display to organize a large number of captured images obtained in the inspection of wind power generation facilities and efficiently grasp the state of the blades. However, Patent Document 1 only stops at organizing and displaying a large number of captured images obtained in the inspection, and when there are damaged parts such as cracks in the captured images, it is not disclosed at all to manage and display the association between the damaged parts and their repair methods.

[0005] The present invention has been made in view of such circumstances, and an object thereof is to provide a wind turbine blade repair support device and a wind turbine blade repair support method that propose a repair method corresponding to a damaged portion from a damaged image of a wind turbine blade.

Means for Solving the Problems

[0006] In order to solve the above problems, a wind turbine blade repair support device of the present invention is a wind turbine blade repair support device that supports the repair of a damaged portion of a wind turbine blade of a wind power generation facility, and includes facility information regarding the wind power generation facility and an input unit that inputs a captured image of the wind turbine blade, a damaged location extraction unit that extracts a damaged image showing an image of the damaged portion from the captured image of the wind turbine blade, a repair method database that stores a plurality of damaged images in advance as reference damaged images and associates and stores a repair method for the reference damaged images, and a similar image search unit that searches for the most similar reference damaged image from the reference damaged images stored in the repair method database based on the extracted damaged image, and a repair method information generation unit that extracts the repair method associated with the searched reference damaged image from the repair method database, combines it with the damaged image, and generates repair method information for the damage specified by the damaged image.

Effects of the Invention

[0007] According to the present invention, it is possible to provide a wind turbine blade repair support device and a wind turbine blade repair support method that propose a repair method corresponding to a damaged portion from a damaged image of a wind turbine blade.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16A

Figure 16B

Figure 16C

Figure 17A

Figure 17B

Figure 17C

MODE FOR CARRYING OUT THE INVENTION

[0009] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. (Embodiment) FIG. 1 is a diagram showing the configuration of a windmill blade repair support apparatus according to an embodiment of the present invention. As shown in FIG. 1, the windmill blade repair support apparatus 100 includes an input unit 110, a repair method information generation unit 120, a repair content recording unit 130, a master recording database 140, a repair method database 150, and a display unit 160 (output unit).

[0010] <Input unit 110> The input unit 110 inputs facility information regarding the wind power generation facility and captured images of the wind turbine blades 73 (hereinafter referred to as blades 73). The facility information includes the name of the site where the captured wind power generation facility is installed, the unit number of the wind power generation facility, and the blade number. The input unit 110 receives the captured image, its capture ID, the capture date and time, and information regarding the captured blade position. The blade position is the position from the blade root to the tip, specified at equal intervals as No.1, No.2, ··· (to be described later in FIG. 8). The input unit 110 receives facility information regarding the wind power generation facility. For example, the site name, unit number, and blade number. FIG. 2 is a table 10 showing the captured image and facility information input to the input unit 110. In the captured image and facility information table 10, the capture ID (ID for calling the captured image), the capture date and time, the facility information (site name · unit number · blade number), and the captured blade position are recorded in association. For example, within the site of a certain wind power plant X, there are 3 wind turbines (blades A - C in FIG. 12 to be described later), and each of the 3 blades is provided for each wind turbine. In the captured image and facility information table 10 in this case, the site name: A, unit number: Unit 1, blade number: 1, blade position: 4 (shooting direction B - 4 in FIGS. 12 and 13 to be described later).

[0011] <Repair method information generation unit 120> The repair method information generation unit 120 extracts the repair method associated with the retrieved reference damage image from the repair method database 150 and combines it with the damage image to generate repair method information for the damage in the damage image. When the repair method information generation unit 120 generates repair method information for the damage identified by the damage image, it associates the damage image used in the generation with the repair method extracted from the repair method database 150 and adds it to the repair method database 150.

[0012] In addition, when the repair method information generation unit 120 cannot extract the repair method associated with the reference damage image from the repair method database 150 of the imaged wind power generation facility, since it is arranged in a wind power generation facility different from the imaged wind power generation facility, the repair method associated with the reference damage image is extracted. Note that, in the present embodiment, the configuration in which the wind turbine blade repair support device 100 includes the repair method database 150 is shown, but the repair method database 150 may exist on the cloud or the like. That is, the configuration in which the wind turbine blade repair support device 100 itself does not include the repair method database 150 may be adopted, such as referring to a database on the cloud.

[0013] The repair method information generation unit 120 includes a damage location extraction unit 121 and a similar image search unit 122 for damage images. · Damage location extraction function The damage location extraction unit 121 extracts a damage image showing an image of the damaged part from the captured image of the wind turbine blade 73. When there are a plurality of images of the damaged parts in the captured image of the wind turbine blade 73, the damage location extraction unit 121 separates and cuts out the images of each damaged part.

[0014] FIG. 3 is a diagram for explaining the damage location extraction function of the damage location extraction unit 121. As shown in FIG. 3, when damage locations 1 and 2 are found in the captured image 20a, the damage location extraction unit 121 cuts out damage images 20b and 20c (see the dashed line enclosures in FIG. 3) including the damage locations 1 and 2, and saves them as damage image 20b○○.jpg and damage image 20c△△.jpg. Note that the damaged location is a part where the difference value is large by comparison with an image taken in the past. In addition, damage to the wind turbine blade has characteristics in the damage shape in the case of lightning strike or the like. Therefore, the pattern of the damaged location stored in the repair method database 150 is read out and collated with the pattern of the captured image, so that the damaged location can be found. According to the pattern matching method, the damaged location and dirt or the like can be identified.

[0015] · Similar image search function for damage images Based on the extracted damage image, the similar image search unit 122 searches for the most similar reference damage image from among the reference damage images stored in the repair method database 150. The similar image search unit 122 has a function of searching the repair method database 150 to check if there is a reference damage image similar to the damage image. Here, when searching for a reference damage image similar to the damage image, it may be searched from the repair method database 150 of the same site as the imaged wind power generation facility. However, if it cannot be found even when searching from the repair method database 150 of the same site, the situation where no search result is found can be avoided by searching from the repair method database 150 of a different site from the imaged wind power generation facility.

[0016] <Repair content recording unit 130> When repair corresponding to the damage image is performed, the repair content recording unit 130 adds a repair record to the master record database 140. The repair content recording unit 130 associates the equipment information, damage image, and repair record and updates them to the master record database 140. Specifically, when repair corresponding to the damage image 〇〇.jpg is performed, the repair content recording unit 130 adds a repair record (repair company, repairer, repair date and time) to the master record database 140.

[0017] <Master record database 140> The master record database 140 records, in order from the left column, the shooting ID and shooting date and time, equipment information (site name, unit number, blade number), photographed blade position, damage image ID, damage part ID, damage type, damage degree, and repair record (repair company, repairer, repair date and time) (Figure 2).

[0018] Figure 4 is a diagram showing the structure (record) of the master record database 140. As shown in FIG. 4, the master recording database 140 records, in order from the left column, the shooting ID and shooting date and time, facility information 140a (site name, unit number, blade number), the blade position photographed, the damage image ID, the damage part ID, the damage type, the damage degree, and the repair record 140b (repair company, repair personnel, repair date and time).

[0019] <Repair method database 150> The repair method database 150 pre-accumulates a plurality of damage images as reference damage images and stores the repair methods associated with the reference damage images. The repair method database 150 is a database that records the association between past damage images and the repair methods used when repairing those damage images. The repair method database 150 records the association between the damage image 〇〇.jpg and the repair method "protective tape". By adding new damage images and repair methods, the repair method database 150 can grow and improve the extraction accuracy for new damage images.

[0020] FIG. 5 is a diagram showing the structure (records) of the repair method database 150. The repair method database 150 records the association between past damage images and the repair methods used when repairing those damage images. For example, in the case of the damage images ●●.jpg and 〇〇.jpg, the repair method "protective tape" is associated. In the case of the damage image ■■.jpg, the repair method "〇〇 paint" is associated. In the case of the damage image ★★.jpg, the repair method "polish the surface and attach a protective tape" is associated.

[0021] <Display unit 160> The display unit 160 is an output unit that outputs repair method information (repair method corresponding to the damaged part) for the damage specified by the damage image generated by the repair method information generation unit 120. For example, the display unit 160 displays the damage image 〇〇.jpg and the repair method "protective tape". Note that the display unit 160 is a form of the output unit, and includes forms of output other than display, for example, printing, and output to a dedicated signal line or network by wired / wireless communication in addition to printing.

[0022] <Shape of the windmill 70> Next, the shape of the windmill 70 of the wind power generation facility to be inspected will be described. FIG. 6 is a perspective view showing the windmill 70 to be inspected. The windmill 70 includes a nacelle 72 provided on a tower 71 and blades 73 (windmill blades) fixed to the nacelle 72. Each blade 73 rotates along the rotation direction θ by receiving wind power, and power generation is performed using this rotational force. First, an R axis is defined along the longitudinal direction of the blade 73 from the base of the nacelle 72 toward the tip of the blade 73.

[0023] FIG. 7 is a perspective view showing a coordinate system defined with the blade 73 as the center. The "R axis" is an axis along the longitudinal direction of the blade 73 as described in FIG. 6. Regardless of the direction in which the rotation direction θ faces (upward, horizontal, or downward), the R axis is defined following the rotation direction θ. The "R cross-section" is a plane perpendicular to the R axis. For convenience, one axis thereof is defined as the X axis and the other axis is defined as the Y axis.

[0024] The "imaging plane" defines from which direction of the R cross-section the blade 73 is imaged. For example, in FIG. 7, an imaging plane a for imaging the blade 73 with the angle of view directed in the positive Y-axis direction, an imaging plane b for imaging the blade 73 with the angle of view directed in the negative X-axis direction, an imaging plane c in the positive X-axis direction (not shown as it is hidden behind the imaging planes a and b), and an imaging plane d in the negative Y-axis direction (not shown as it is hidden behind the imaging planes a and b) are exemplified, for a total of four imaging planes. Note that four imaging planes may be provided around the blade as shown in FIG. 7, or may be provided without being limited to four, such as five. In any case, by imaging the blade 73 from a plurality of directions, un-imaged portions are prevented from occurring.

[0025] Figure 8 is a perspective view showing the movement of the flight route (F1 during forward movement, F2 during reverse movement) of the unmanned aerial vehicle along the R-axis of the blade 73. The unmanned aerial vehicle that has moved near the nacelle 72 moves to a reference position (r = 1) near the nacelle 72, which is the first end of the blade 73. This reference position is the starting point of the flight route F1. Note that two pieces of information (such as No.1, r = 1, etc.) are added to one divided image 33. First, the first additional information (No.1, No.2,... No.20) is unique identification information 31 for each image taken at the shooting position. That is, the identification information 31 can also be said to be information numbered according to the shooting order of the images.

[0026] On the other hand, the second additional information (r = 1, 2,... 10) is specific information about the shooting position along the R-axis from the first end (r = 1) of the blade 73 to the opposite second end (r = 10). The smaller the value of r, the closer the position is to the nacelle 72 side, and the larger the value of r, the closer the position is to the tip side of the blade 73. That is, even at the same shooting position r = 1, when shooting the shooting surface a according to the flight route F1, the identification information (No.1) is attached, and when shooting the shooting surface b according to the flight route F2, the identification information (No.20) is attached.

[0027] By attaching the shooting position information for each blade to the divided image 33, even when the drone moves back and forth, the entire blade can be correctly displayed on the screen. For example, by arranging and displaying two images (identification information 31 = No.1 and No.20) with the same shooting position r = 1, it is easy to manage the divided images with a display suitable for inspection.

[0028] The unmanned aerial vehicle that has reached the tip of the blade 73 similarly starts shooting along the flight route F2 during reverse movement with respect to a shooting surface b different from the shooting surface a that has been shot so far. That is, the unmanned aerial vehicle moves in the direction of returning from the second end (No.11 at the tip of the blade 73) far from the reference position to the first end (No.20 at the base of the nacelle 72) and shoots at predetermined equal intervals.

[0029] Next, the operation of the wind turbine blade repair support device 100 configured as described above will be described. FIG. 9 is a flowchart showing a wind turbine blade repair support process for cutting out a damaged image from the captured image input to the input unit 110 and displaying a repair method. First, in step S1, the input unit 110 inputs a captured image, a shooting ID, a shooting date and time, facility information (site name, unit number, blade number), and the blade position where the image was taken. Here, as shown in FIG. 2, the input unit 110 is input with a shooting ID of "0000004", a shooting date and time of "2022 / 01 / 10 / 10:01", and as facility information, a site name of "A" (blades A to C in FIG. 12 described later), a unit number of "1", a blade number of "1", and a blade position where the image was taken of "4" (shooting direction B-4 in FIGS. 12 and 13 described later).

[0030] In step S2, the damage location extraction unit 121 (FIG. 1) cuts out (extracts) a damaged image from the captured image. FIG. 10 is a diagram showing an example of a table 20 (damaged image table 20) in which the damage location extraction unit 121 has cut out damaged images from the captured image. FIG. 10 shows an example in which two damaged images (〇〇.jpg, △△.jpg) are cut out from one captured image (shooting ID "0000004" and shooting date and time "2022 / 01 / 10 / 10:01") shown in FIG. 2.

[0031] Hereinafter, the similar image search unit 122 (FIG. 1) executes the extraction process of the most similar reference damaged image (step S3) and the extraction process of the repair method associated with this reference damaged image (steps S4 to S6). · Extraction process of reference damaged image First, in step S3, the similar image search unit 122 extracts the most similar reference damaged image from the repair method database 150 (FIG. 5) using the damaged images (〇〇.jpg, △△.jpg) as search keys. A method for extracting the most similar reference damage image from the repair method database 150 (Fig. 5) will be described. The similar image search unit 122 selects and extracts information on the damage type and damage degree (level) determined after inspection (after imaging) and information on the state after repair (repair method, results of repair agents used, etc.) of the damaged part from past data. Specifically, the similar image search unit 122 extracts the most similar reference damage image (●●.jpg in Fig. 5) based on the following information: (1) damaged part, (2) damage type, (3) damage degree (level), (4) repair agent used, (5) repairer, and (6) repair completion date. Here, the similar image search unit 122 uses the damage images (〇〇.jpg, △△.jpg) as search keys and extracts the reference damage image (●●.jpg) most similar to the damage images (〇〇.jpg, △△.jpg).

[0032] · Extraction process of repair method Next, in step S4, the repair method information generation unit 120 (Fig. 1) determines whether it is possible to extract the repair method associated with the reference damage image from the repair method database at the same site as the imaged wind power generation facility. If it is possible to extract the repair method associated with the reference damage image from the repair method database 150 at the same site as the imaged wind power generation facility (S4: Yes), the process proceeds to step S5. If it is not possible to extract the repair method associated with the reference damage image (S4: No), the process proceeds to step S6.

[0033] In step S5, the similar image search unit 122 (Fig. 1) extracts the most similar reference damage image (●●.jpg in Fig. 5) from the repair method database 150 and also extracts the repair method associated with the reference damage image, and then proceeds to step S7.

[0034] Specifically, the similar image search unit 122 uses the damage image ID "damage image 〇〇.jpg" shown in Fig. 10 as a search key, determines that the damage type is "crack", and extracts the repair method "protective tape" associated with the most similar reference damage image from the repair method database 150 (Fig. 3) (Fig. 5).

[0035] In step S6, the similar image search unit 122 extracts the repair method associated with the reference damage image from the repair method database of a site different from the wind power generation facility. Here, when the similar image search unit 122 searches for a reference damage image similar to the damage image, it may search from the repair method database 150 of the same site as the imaged wind power generation facility (step S5). However, if it cannot be found even by searching from the repair method database 150 of the same site, by searching from the repair method database 150 of a site different from the imaged wind power generation facility (step S6), it is possible to avoid a situation where no search result is found.

[0036] In step S7, the repair method information generation unit 120 causes the display unit 160 to display the damage image (〇〇.jpg in FIG. 10) and the repair method ("protective tape" in FIG. 5).

[0037] In step S8, the repair method information generation unit 120 associates and records the damage image (〇〇.jpg in FIG. 10) and the repair method ("protective tape" in FIG. 5) in the repair method database 150. By adding the damage image and the repair method to the repair method database 150, the database can grow, and the extraction accuracy for new damage images can be improved. As described above, the repair method database 150 may be configured to be provided on the cloud outside the wind turbine blade repair support device 100.

[0038] In step S9, when the repair corresponding to the damage image (〇〇.jpg in FIG. 10) is performed, the repair content recording unit 130 updates (adds to) the master record database 140 based on the repair record (repair contractor, repairer, repair date and time), and ends the processing of this flow.

[0039] FIG. 11 is a diagram showing the repair record addition data 140C added to the master record database 140 in FIG. 4. As shown in FIG. 11, the supplementary repair record data 140C has the shooting ID "0000004" and shooting date and time "2022 / 01 / 10 / 10:01", site name "A", unit number "1", blade number "1", photographed blade position "4", damage image ID "〇〇.jpg", damage part ID "0005", damage type "crack", and the shooting ID "0000004" and shooting date and time "2022 / 01 / 10 / 10:01", site name "A", unit number "1", blade number "1", photographed blade position "4", damage image ID "△△.jpg", damage part ID "0006", damage type "crack", and is appended to the master record database 140 in FIG. 4.

[0040] [Embodiment] [Screen Example] First, a screen example of the wind turbine blade repair support device 100 will be described. FIG. 12 is a diagram showing overall images TIP captured in shooting directions B-1 to B-5 for blades A, B, and C of unit number n of the blade, and images each having a length of 5 to 25 m. The ● marks in FIG. 12 indicate that the damage has not been confirmed, and the thick frame indicates that there is damage. FIG. 13 is a diagram showing the damage levels Lv2 to Lv4 and unclassified display for each of the shooting directions B-1 to B-5 of blade B among the blade confirmation results in FIG. 12. Supplementary explanation will be given regarding the captured images. It is assumed that captured images 01 to 09 are stored in the screen example folder (not shown) of the storage unit (not shown) of the wind turbine blade repair support device 100. Note that 01 to 09 of the captures are an example.

[0041] Capture 01: Wind turbine master data, information for each wind turbine Capture 02: Automatic flight route setting, creating a route with master data and local information (nacelle angle) Capture 03: Image of the blade viewed from above Capture 04: Image of the blade viewed from the side Capture 05: Screen on which the captured data is automatically sorted and displayed (FIG. 12) Capture 06: Judgment screen. If there is a damaged part, set and record the degree of damage, etc. Capture 07: Comparison screen. Compare the same position with past inspection data and confirm the progress degree of the damaged part, etc. Capture 08: Task basic information registration screen. Register detailed information, person in charge, repair deadline, etc. for the damaged part. Capture 09: Task registration screen. Register the repair method, repair agent, etc. for the damaged part and give work instructions.

[0042] <Specific example of repair method> Next, a specific example of the repair method of the windmill blade repair support device 100 will be described. The specific example of the repair method includes 1. main damage types, 2. reasons for damage, and 3. repair methods. 1. Main damage types The main damage types are (1) wear, (2) crack, (3) paint peeling, (4) FEP (Fiber Reinforced Plastics) laminate peeling, (5) protective tape peeling, (6) through-hole due to lightning strike, (7) internal conductor damage due to lightning strike, (8) receptacle dropout, and (9) breakage. In the master record database 140 of FIGS. 4 and 10, "crack" and "crack" are exemplified as damage types. However, the damage types include the above (1) to (9), and the master record database 140 records the damage types (1) to (9).

[0043] 2. Reasons for damage The reasons for damage are (1), (2), (3), and (5) in the case of aging deterioration. Also, the reasons for damage are (6), (7), and (8) in the case of lightning strike influence, and (8) and (9) in the case of typhoon.

[0044] 3. Repair methods The repair methods include removal of the peeled / damaged part, grinding / cutting / drilling, FRP lamination, putty filling, heating, curing, molding, painting, and conductor connection.

[0045] <Specific example of changing the damage level judgment method> Next, a specific way to change the damage level judgment method according to the damaged part will be described. · Due to the structure of the blade, the components and materials vary from part to part, so the repair method differs depending on the damaged part and its condition. · As described above, there are various types of damage, and the repair procedures and types of repair agents also differ. · Since the repair is carried out outdoors, the handling of repair agents and repair methods that are affected by temperature and humidity are different. When the temperature is low, it is heated and cured. · The degree of damage also changes depending on the size (range) and depth (such as internal penetration) of the damage, which affects the level determination. · In addition, basically, the repair is carried out without lowering the blade to the ground, so it is necessary to consider how to access the damaged part at a high altitude as part of the repair method. · The access methods include installing boarding equipment at the tip of the crane boom (such as a sky box), suspending a gondola similar to building maintenance with a wire, and rope access where a person hangs down with a rope.

[0046] <Specific examples of repair methods> Next, a highly reliable repair technique will be described. Figure 14 is a diagram for explaining the wear process of the blade. The upper figure of Figure 14 schematically shows the state of the blade before damage. The blade has a core material, FRP, putty, and paint in order from the inside to the surface. The middle figure of Figure 14 shows the state where the paint and putty are worn by particle collision. In addition, as shown in the lower figure of Figure 14, the FRP deteriorates and cracks occur due to ultraviolet rays. As time passes from the middle figure of Figure 14 to the lower figure of Figure 14, the repair scale increases from small to large.

[0047] Figure 15 is a diagram showing an example of FRP damage. As shown in Figure 15, FRP damage 3 has occurred on the blade 73. Neglecting maintenance will lead to breakage, and costly repairs will affect the business viability. Regular inspections and appropriate maintenance are important. The wind turbine blade repair support device 100 can propose a repair method corresponding to the damaged part from the damage image of the blade.

[0048] Figs. 16A to 16C are diagrams for explaining an example of blade repair. Fig. 16A is a diagram showing the leading edge of the blade before repair. Fig. 16B is a diagram showing the leading edge of the blade after repair. Fig. 16C is a diagram showing the leading edge of the blade after repair.

[0049] Fig. 17A is a diagram for explaining the selection of a repair agent. As the blade wears, FRP, putty, paint, and protective tape are selected as repair agents. Also, as shown in the Venn diagram in the center of Fig. 17A, each repair agent is used in combination. Fig. 17B is a diagram for explaining an example of performing tests and verifications based on the repair agent selection and combination of repair agents shown in Fig. 17A. The upper diagram of Fig. 17B shows the wear resistance test results of materials (paint, or paint and putty) A to D when the erosion length [μm] is taken on the vertical axis and the projected particle weight [g] is taken on the horizontal axis. The middle diagram of Fig. 17B is an image diagram of the adhesion test. The lower diagram of Fig. 17B is an actual measurement diagram for workability verification. Fig. 17C is an example of blade repair application reflecting the test results of Fig. 17B. The upper diagram of Fig. 17C shows the repair work of an operator, and the lower diagram of Fig. 17C shows an example of repair with putty.

[0050] [Effect] As described above, the wind turbine blade repair support device 100 (Fig. 1) according to one embodiment is a wind turbine blade repair support device that supports the repair of a damaged portion of the blade 73 of a wind power generation facility, and includes an input unit 110 that inputs facility information regarding the wind power generation facility and a captured image of the blade 73, a damage location extraction unit 121 that extracts a damage image showing an image of the damaged portion from the captured image of the blade 73, a repair method database 150 that stores a plurality of damage images in advance as reference damage images and associates and stores a repair method for the reference damage images, a similar image search unit 122 that searches for the most similar reference damage image from among the reference damage images stored in the repair method database 150 based on the extracted damage image, and a repair method information generation unit 120 that extracts the repair method associated with the searched reference damage image from the repair method database 150, combines it with the damage image, and generates repair method information for the damage specified by the damage image.

[0051] By doing so, a repair method corresponding to the damaged part can be proposed from the damage image of the blade 73. For example, when there are damaged parts such as cracks in the captured image, the damaged part and its repair method can be linked and managed and displayed.

[0052] For example, when the screen on which the capture 05 (FIG. 12) showing the automatically sorted and displayed shooting data is input to the wind turbine blade repair support device 100 (FIG. 1), according to the damaged part and the damage state (damage cause, damage depth, spread of damage, number of damages), and further considering the components and materials for each part of the blade 73, an optimal repair method (FRP lamination, putty filling, protective tape, etc.) can be displayed to support administrators and the like.

[0053] Administrators and the like can take prompt and appropriate actions based on the support from the wind turbine blade repair support device 100, specifically, based on the damaged part and its repair method displayed on the display unit 160. Since the damaged part and its repair method are proposed, it leads to the promotion of maintenance and can prevent breakage and the like. Since it contributes to regular inspections and appropriate maintenance, it becomes possible to take countermeasures before the damage of the damaged part expands, and as a result, the operation cost of the wind power generation facility can be reduced.

[0054] The wind turbine blade repair support device 100 (FIG. 1) includes an output unit that outputs repair method information for the damage identified by the damage image generated by the repair method information generation unit 120.

[0055] By doing so, administrators and the like can take prompt and appropriate actions based on the damaged part and its repair method output from the wind turbine blade repair support device 100.

[0056] In the wind turbine blade repair support device 100 (FIG. 1), when there are a plurality of images of the damaged part in the captured image of the blade 73, the damaged part extraction unit 121 is characterized by separating and cutting out the images of each damaged part.

[0057] By doing so, it is possible to propose a repair method corresponding to each damaged part by separately cutting out the damaged parts from the captured image information.

[0058] In the wind turbine blade repair support device 100 (Fig. 1), when the repair method information generation unit 120 generates repair method information for the damage specified by the damage image, the damage image used for the generation is linked to the repair method extracted from the repair method database 150 and added to the repair method database 150.

[0059] By doing so, the repair method database 150 can grow and the extraction accuracy for new damage images can be improved.

[0060] In the wind turbine blade repair support device 100 (Fig. 1), when repair corresponding to the damage image is performed, it is provided with a repair content recording unit 130 that adds a repair record to the master record database 140. The repair content recording unit 130 links the equipment information, the damage image, and the repair record and updates the master record database 140.

[0061] By doing so, the master record database 140 can be grown and the search accuracy can be improved.

[0062] In the wind turbine blade repair support device 100 (Fig. 1), when the repair method information generation unit 120 cannot extract the repair method linked to the reference damage image from the repair method database 150 of the imaged wind power generation equipment, it extracts the repair method linked to the reference damage image from the repair method database 150 arranged in a wind power generation equipment different from the imaged wind power generation equipment.

[0063] By doing so, it is possible to extract a repair method based on the image information of another site and avoid a situation where no search result is found.

[0064] In addition, in this embodiment, as the structure to be displayed by the wind turbine blade repair support device 100, the blade 73 (Fig. 5) of the wind turbine 70 captured by the unmanned aircraft is exemplified. However, any device that inputs the captured image of the blade 73 may be used, and it is not necessary to use an unmanned aircraft. For example, an imaging device may be installed on a large crane.

[0065] The above-described embodiment examples have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Also, a part of the configuration of one embodiment example can be replaced with the configuration of another embodiment example, and the configuration of another embodiment example can be added to the configuration of one embodiment example. Further, it is possible to add, delete, or replace other configurations for a part of the configuration of each embodiment example.

Explanation of Reference Numerals

[0066] 10 Equipment information table (equipment information) 20 Damage image table (damage image) 70 Wind turbine 71 Tower 72 Nacelle 73 Blade (wind turbine blade) 100 Wind turbine blade repair support device 110 Input unit 120 Repair method information generation unit 121 Damage location extraction unit 122 Similar image search unit 130 Repair content recording unit 140 Master record database 140C Additional repair record data 150 Repair method database 160 Display unit (output unit) S1 Input step of inputting equipment information related to the wind power generation equipment and the captured image of the wind turbine blade S2 Extraction step of extracting a damage image showing an image of the damaged part from the captured image of the wind turbine blade Search step of searching for the most similar reference damage image from among the reference damage images stored in the repair method database based on the extracted damage image in S3 Generation step of extracting the repair method associated with the searched reference damage image from the repair method database, combining it with the damage image, and generating repair method information for the damage specified by the damage image Storage step of preliminarily storing a plurality of damage images as reference damage images and storing in the repair method database while associating the repair method for the reference damage image

Claims

1. A wind turbine blade repair support device for supporting the repair of damaged parts of a wind turbine blade of a wind power generation facility, an input unit for inputting equipment information related to the wind power generation facility and a captured image of the wind turbine blade, a damaged location extraction unit for extracting a damaged image showing an image of the damaged part from the captured image of the wind turbine blade, referring to a repair method database that stores a plurality of damaged images in advance as reference damaged images and associates and stores a repair method for the reference damaged images, and based on the extracted damaged image, searching for the most similar reference damaged image from the reference damaged images stored in the repair method database; a similar image search unit, a repair method information generation unit that extracts the repair method associated with the retrieved reference damaged image from the repair method database, combines it with the damaged image, and generates repair method information for the damage specified by the damaged image, comprising a wind turbine blade repair support device characterized by the above.

2. An output unit for outputting repair method information for the damage specified by the damaged image generated by the repair method information generation unit is provided. The wind turbine blade repair support device according to claim 1, characterized by the above.

3. The damaged location extraction unit separates and cuts out images of each damaged part when there are a plurality of images of the damaged part in the captured image of the wind turbine blade. The wind turbine blade repair support device according to claim 1, characterized by the above.

4. The equipment information includes the name of the site where the imaged wind power generation facility is installed, the unit number and blade number of the wind power generation facility. The wind turbine blade repair support device according to claim 1, characterized by the above.

5. When the repair method information generation unit generates repair method information for the damage specified by the damaged image, it associates the damaged image used in the generation with the repair method extracted from the repair method database and adds it to the repair method database. The wind turbine blade repair support device according to claim 1, characterized by the above.

6. When the repair corresponding to the damaged image is performed, a repair content recording unit for adding a repair record to a master record database is provided. The repair content recording unit updates the master record database by associating the equipment information, the damaged image, and the repair record. The wind turbine blade repair support device according to claim 1, characterized by the above.

7. When the repair method information generation unit cannot extract the repair method associated with the reference damage image from the repair method database of the imaged wind power generation facility, the repair method associated with the reference damage image is extracted from the repair method database arranged in a wind power generation facility different from the imaged wind power generation facility. The wind turbine blade repair support device according to claim 1, characterized in that.

8. A wind turbine blade repair support method for supporting the repair of a damaged part of a wind turbine blade of a wind power generation facility, An input step of inputting the facility information regarding the wind power generation facility and the captured image of the wind turbine blade, An extraction step of extracting a damage image showing an image of the damaged part from the captured image of the wind turbine blade, A plurality of damage images are preliminarily accumulated as reference damage images, and with reference to a repair method database that stores and associates repair methods for the reference damage images, based on the extracted damage image, the most similar reference damage image is searched for from among the reference damage images accumulated in the repair method database. A generation step of extracting the repair method associated with the searched reference damage image from the repair method database, combining it with the damage image, and generating repair method information for the damage specified by the damage image. The wind turbine blade repair support method, characterized in that.

Images