Detecting defects in wind turbine blades

Abstract

In a first aspect, a wind turbine blade inspection system for detecting defects in a wind turbine blade is provided. A wind turbine blade inspection system includes a directional light source, a diffused light source, an image capture device, and a controller. The controller is configured to analyze an image from the image capture device to detect a defect. In another aspect, a computer-implemented method for detecting defects in a wind turbine blade is provided. In yet another aspect, a computing system is provided that includes a processor configured to perform a method according to any example herein. In yet another aspect, a computing program is provided that includes instructions that, when executed by a processor, cause the processor to perform a method according to any example herein.

Description

[0001] The present disclosure relates to wind turbine blade inspection systems, methods, controllers, and computer programs for detecting defects in wind turbine blades. Background Art

[0002] Modern wind turbines are commonly used to supply power to the electrical grid. Such wind turbines typically include a rotor with a rotor hub and a plurality of wind turbine blades. The rotor rotates under the influence of the wind on the blades. The rotation of the rotor shaft drives the generator rotor, either directly ("direct drive") or through the use of a gearbox. The gearbox (if present), generator, and other systems are typically mounted in a nacelle at the top of the wind turbine tower.

[0003] Wind turbine blades are typically made of fiber-reinforced polymers or plastics (FRP), which are composite materials consisting of a polymer matrix reinforced with fibers. The fibers are usually glass or carbon and provide longitudinal stiffness and strength.

[0004] Wind turbine blades are typically manufactured by joining two blade shell parts made of fiber-reinforced polymers (e.g., glass or carbon fiber reinforced polymers). The two blade shell parts are first molded and then bonded together, for example, using an adhesive. For example, the pressure-side blade shell can be bonded to the suction-side blade shell via a joint line along the leading and trailing edges.

[0005] These blade shell parts can be molded using resin infusion technology or prepreg technology. In resin infusion technology, fibers are placed in a mold and then resin is injected into the mold cavity under pressure. The resin fills the volume between the cavities and then the resin is cured or hardened. Examples of resin infusion technology can be resin transfer molding (RTM) or vacuum assisted resin transfer molding (VARTM). In VARTM, the resin is injected under vacuum or pressure below atmospheric pressure.

[0006] Because blade manufacturing is a complex task, defects can occur during blade fabrication. Defects are imperfections or weak spots in the blade that can trigger failure of a wind turbine blade during operation. Different types of defects can occur during blade manufacturing. For example, fibers may become misaligned in the mold before or during curing. These fiber misalignments can result in wrinkles or steps in the blade shell, which can reduce the composite's compressive strength.

[0007] Fiber layers can also peel or delaminate, meaning they can separate due to a lack of fusion between the layers. These peeling or delamination defects can serve as starting points for crack growth. Voids and bubbles can also occur during blade manufacturing when air pockets become trapped in the material of the blade shell. During operation of the wind turbine blade, these voids and bubbles can create localized stress concentrations. Foreign material can also become trapped in the composite material. Another example of a defect can be dry areas where some areas of the blade shell lack resin.

[0008] Wind turbine blades can be inspected manually to detect manufacturing defects. Qualified operators are employed to perform visual inspections of wind turbine blades (e.g., blade shell components). For example, operators may visually inspect the blade shell components while they are placed in a mold. This type of inspection typically takes considerable time. Furthermore, this manual inspection relies on the operator's subjectivity and expertise. Furthermore, training operators to detect manufacturing defects is time-consuming.

[0009] The present disclosure provides examples of systems and methods that at least partially address some of the above-mentioned shortcomings. Summary of the Invention

[0010] In a first aspect, a wind turbine blade inspection system for detecting defects in a wind turbine blade is provided.

[0011] A wind turbine blade inspection system includes a directional light source for illuminating an inspection surface of a wind turbine blade at an acute angle relative to the inspection surface, and a diffuse light source for diffusely illuminating the inspection surface of the wind turbine blade. The wind turbine blade inspection system also includes an image capture device for capturing images of the inspection surface. The wind turbine blade inspection system also includes a controller for selectively activating the directional light source or the diffuse light source, receiving images of the inspection surface when illuminated by the directional light source and when illuminated by the diffuse light source from the image capture device, and analyzing the received images of the inspection surface to detect defects.

[0012] In the present disclosure, a directional light source is understood to be a light source that emits light and projects this light at an acute angle relative to the inspection surface.A directional light source is thus configured to emit a light beam that directly illuminates the inspection surface at an acute angle.

[0013] A light beam is light that propagates from a light source in essentially one direction. The beam angle is an angular expression showing how light is emitted from a light source and can be defined as the degree of width of the light emitted from the light source. As light spreads out, the intensity decreases. A smaller beam angle therefore provides concentrated light. The beam angle is the angle between relative points on the beam axis where the intensity drops to 50% of its maximum value. In the context of the present disclosure, the range of the beam angle of a directional light source is greater than 0° and less than 90°, specifically greater than 0° and less than 45°, and more specifically greater than 0° and less than 30°. These beam angles provide a guarantee that the directional light source emits light in a concentrated manner.

[0014] A directional light source illuminating an inspection surface is to be understood as a directional light source that emits light at an angle towards the inspection surface. In the context of the present disclosure, the axis of the light beam emitted by the directional or offset light source forms an acute angle with respect to the inspection surface, i.e. the angle of incidence to the inspection surface is an acute angle. The acute angle formed by the axis of the light beam and the inspection surface is lower than 70°, e.g. between 10° and 70°. Thus, the light beam may illuminate the inspection surface at an acute angle, which corresponds to the angle formed between the axis of the light beam (which is incident on the inspection surface) and the inspection surface. Due to the acute angle between the axis of the light beam and the inspection surface, the beam angle is less than 90°, e.g. greater than 0° and less than 45°, the directional light results in visible shadows, since the object is illuminated only from one direction at a certain angle and is obscured from another direction.

[0015] In this disclosure, the angle formed between the directed light and the inspection surface refers to the angle of the beam axis relative to the inspection surface.

[0016] In this disclosure, a diffuse light source should be understood as a light source that emits light in all directions. Unlike a directional light source, the light emitted from a diffuse light source is not concentrated in a specific area. Therefore, diffuse light can be considered indirect light. Light emitted from a diffuse light source is evenly distributed across surfaces perpendicular to the diffuse light source.

[0017] Therefore, the light emitted from a directional light source is substantially concentrated, i.e., the beam angle is greater than 0° but less than 90°, more specifically, greater than 0° but less than 45°, and more specifically, greater than 0° but less than 30°. In contrast, a diffuse light source propagates in all directions, i.e., the diffuse light source is evenly distributed over the inspection surface. Therefore, the beam angle of diffuse light is greater than that of directional light.

[0018] According to this aspect, defects in wind turbine blades can be automatically detected. This reduces the inspection time and the number of operators required to inspect wind turbine blades. Furthermore, the accuracy and consistency of inspections are improved. This improves the quality of wind turbine blades while reducing the number of false defects.

[0019] Furthermore, using two different light sources to illuminate the same surface improves the system's detection capabilities. Depending on the morphology of a wind turbine blade defect, detection accuracy can be improved by illuminating the inspection surface with only one of the light sources. This improves defect detection reliability.

[0020] Image capture devices can more easily detect some types of defects, such as wrinkles or steps, when illuminated by a directional light source (i.e., using a concentrated light beam). When illuminating a wrinkle or step, the directional light emitted by the directional light source creates a shadow. This shadow can be easily captured by the image capture device. However, when illuminated with diffuse light, wrinkles or step defects are not easily visible because they do not create a discernible shadow unless the wrinkle or step defect is sufficiently large.

[0021] Other types of defects, such as delaminations, voids, and bubbles, can be easily detected when illuminated with diffuse light. These defects can scatter the light projected by the diffuse light source to improve the identification of these types of defects.

[0022] Therefore, the present disclosure aims to improve the identification and detection of different types of defects in wind turbine blades by using two different types of light.

[0023] In another aspect, a computer-implemented method for detecting defects in a wind turbine blade is provided. The computer-implemented method includes activating a directional light source of a wind turbine blade inspection system to illuminate an inspection surface of the wind turbine blade at an acute angle relative to the inspection surface; and receiving, by a controller, a first image of the inspection surface illuminated by the directional light source. Furthermore, the computer-implemented method includes activating a diffuse light source of the wind turbine blade inspection system to diffusely illuminate the inspection surface, and receiving, by the controller, a second image of the inspection surface illuminated by the diffuse light source. The computer-implemented method also includes analyzing, by the controller, the first and second images of the inspection surface to detect defects in the inspection surface.

[0024] In yet another aspect, a controller or computing system is provided, comprising a processor configured to perform a method according to any example herein.

[0025] In yet another aspect, there is provided a computer program comprising instructions which, when executed by a processor, cause the processor to perform a method according to any example herein.

[0026] Advantages derived from these aspects may be similar to those mentioned with respect to the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Non-limiting examples of the present disclosure will now be described with reference to the accompanying drawings, in which: Figure 1shows a perspective view of a wind turbine according to one example; Figure 2 shows a perspective view of a wind turbine blade according to one example; Figure 3 Shown Figure 2 a cross-sectional view of a wind turbine blade; FIG4 a schematically illustrates a directional light source illuminating a defect in an inspection surface according to an example; Figures 4b and 4c schematically illustrate examples of diffuse light sources illuminating defects in an inspection surface; FIG5 a schematically illustrates a wind turbine blade inspection system for detecting defects in a wind turbine blade when a directional light source is illuminating an inspection surface according to an example of the present disclosure; FIG5 b schematically illustrates the wind turbine blade inspection system of FIG5 a when a diffuse light source is illuminating the inspection surface; 6a and 6b illustrate a front view and a side view, respectively, of a wind turbine blade inspection system according to an example of the present disclosure; Figure 7 schematically illustrates a wind turbine blade inspection system according to an example of the present disclosure; Figure 8 schematically illustrates a wind turbine blade inspection system according to an example of the present disclosure; Figure 9 is a block diagram of a computer-implemented method for detecting defects in a wind turbine blade according to an example of the present disclosure; Figure 10A 、 10B 10C and 10D are block diagrams of computer-implemented methods for detecting defects in wind turbine blade shells according to examples of the present disclosure; and Figure 11 Controllers and computing programs according to examples of the present disclosure are shown. DETAILED DESCRIPTION

[0028] In these figures, like reference numerals are used to denote matching elements.

[0029] Figure 1A perspective view of an example of a wind turbine 1 is shown. As shown, wind turbine 1 includes a tower 2 extending from a support surface 3, a nacelle 4 mounted on tower 2, and a rotor 5 coupled to nacelle 4. Rotor 5 includes a rotatable hub 6 and at least one wind turbine blade 7 coupled to and extending outward from rotor hub 6. For example, in the example shown, rotor 5 includes three wind turbine blades 7. However, in alternative embodiments, rotor 5 may include more or fewer than three blades 7. Each wind turbine blade 7 may be spaced apart from rotor hub 6 to facilitate rotating rotor 5, thereby enabling kinetic energy from the wind to be converted into usable mechanical energy and subsequently into electrical energy. For example, rotor hub 6 may be rotatably coupled to a generator positioned within or forming part of nacelle 4 to enable the generation of electrical energy.

[0030] Figure 2 An example of a wind turbine blade 7 is shown. The wind turbine blade 7 extends in a longitudinal or spanwise direction 37 from a blade root end 71 to a blade tip end 72. The blade 7 includes a blade root region or portion 50 closest to the rotor hub, a profiled or airfoil portion 52 furthest from the rotor hub, and a transition portion 51 between the blade root 50 and the airfoil portion 52. The blade 7 includes a leading edge 53 that faces in the direction of rotation of the blade 7 when mounted on the rotor hub, and a trailing edge 54 that faces in the opposite direction of the leading edge 53.

[0031] The airfoil member 52 has a shape designed to generate lift, while the blade root 50 has a circular or elliptical cross-section for structural reasons and to facilitate mounting of the blade to the rotor hub. The diameter or chord of the blade root 50 can be constant along the entire blade root 50. At the transition member 51, the profile gradually changes from the circular or elliptical cross-section of the blade root 50 to the airfoil profile of the airfoil member 52. The wind turbine blade 7 can be connected to the rotor hub via a blade root attachment member 55.

[0032] Wind turbine blade 7 includes a blade shell 73. Blade shell 73 includes an outer side or surface that defines the outer shape of the blade, such as the outer shape at the blade root and the outer shape at the airfoil. Blade shell 73 also includes an inner side or surface that defines the inner volume of the blade and faces a load-bearing structure (not shown). Blade shell 73 can be made of a fiber-reinforced polymer or plastic (e.g., fiberglass and / or carbon fiber).

[0033] The blade shell may be formed from a plurality of blade shell components. The plurality of blade shell components may be joined together to form the blade shell. The blade shell components may be formed according to any of the examples disclosed herein and then joined. Resin infusion techniques, such as RTM or VARTM, or prepreg techniques may be used to manufacture the blade shell components.

[0034] In some examples, the blade shell comprises a pressure side blade shell component and a suction side blade shell component. The pressure side blade shell component may be joined to the suction side blade shell component along a joining line along the leading edge 53 and the trailing edge 54. Each of these blade shell components may be manufactured in a mold and then joined together to define the entire blade shell of the wind turbine blade 7. A load-bearing structure may be arranged between the pressure side blade shell component and the suction side blade shell component.

[0035] Figure 3 Shown Figure 2 FIG2 is a cross-sectional view of a wind turbine blade. A suction side 57, or downwind side, and a pressure side 56, or upwind side, extend from a leading edge 53 to a trailing edge 54. Wind turbine blade 7 also includes a chord line 38 between leading edge 53 and trailing edge 54. Chord line 38 extends in an edgewise or chordwise direction. A flapwise direction 39 is substantially perpendicular to chord line 38.

[0036] The wind turbine blade 7 includes a blade structure that provides stiffness to the wind turbine blade. The blade structure of this example includes a blade shell 73 and a load-bearing structure. In other examples, the blade structure may also include a plurality of structural ribs arranged along the length of the blade. In this example, the load-bearing structure includes shear webs, such as a leading edge shear web 43 and a trailing edge shear web 44. A cavity 42 is defined between the leading edge shear web 43 and the trailing edge shear web 44. The cavity 42 may extend along the entire length in the span direction. The load-bearing structure of this figure also includes a pressure side spar cap 74 arranged at the pressure side 56 and a suction side spar cap 76 at the suction side 57. In some examples, the shear webs 43 and 44 may be spar boxes having spar sides (such as a trailing edge spar side and a leading edge spar side).

[0037] 4a schematically shows a directional light source 110 illuminating a defect in an inspection surface 200 according to an example. The defect in this figure is a wrinkle 210. The directional light source 110 emits directional light 112 towards the inspection surface 200.

[0038] Beam angle 111 may be greater than 0° but less than 90°. Directional light 112 is thus relatively concentrated. In this example, beam angle 111 is approximately 20°. In some examples, beam angle 111 may be greater than 0° but less than 30°.

[0039] Directed light 112 is emitted so as to form an acute angle 113 relative to inspection surface 200. Angle 113 is measured between the inspection surface and beam axis 114. Beam axis 114 is the axis of a cone defined by directed light 112 emitted by directional light source 110. When illuminating wrinkles 210 of inspection surface 200, angle 113 falls below 70° to form shadow region 201. Shadow region 201 is formed on the opposite side of directional light source 110 from wrinkles 210.

[0040] Therefore, the image capture device 130 arranged above the inspection surface can acquire an image of the shadow area 201 .

[0041] Figures 4b and 4c schematically illustrate examples of diffuse light sources illuminating defects in an inspection surface.

[0042] In FIG4b , the diffuse light source 120 is arranged concentrically with the image capture device 130. In FIG4c , the diffuse light source 120 is arranged adjacent to the image capture device 130. The diffuse light source 120 of FIG4c includes a pair of diffuse light source units 125a and 125b. These diffuse light source units 125a and 125b are arranged on opposite sides of the image capture device 130.

[0043] In these figures, diffuse light source 120 emits diffuse light 121 in different directions. In these figures, beam angle 111 is greater than 120°. In some examples, beam angle 111 can be greater than 180°. In some examples, diffuse light source 120 can emit light around 360° (i.e., in all directions). In these examples, beam angle 111 can be considered to be 360°. Diffuse light source 120 is arranged perpendicular to inspection surface 200.

[0044] When the diffuse light 121 is emitted in different directions, a shadow generated by one ray or one direction of the diffuse light is cleared by other rays of the diffuse light 121 .

[0045] Therefore, no sharp shadows are formed on the inspection surface. A substantially uniform illumination of the inspection surface is thus achieved. Therefore, the beam angle generated by the directional light source is smaller than the beam angle generated by the diffuse light source. Thus, the directionally emitted light forms a concentrated beam.

[0046] Figures 5a and 5b schematically illustrate a wind turbine blade inspection system 100 for detecting defects in a wind turbine blade, according to an example. In Figure 5a, a directional light source 110 directionally illuminates an inspection surface 200 of a wind turbine blade 7, and in Figure 5b, a diffuse light source 120 diffusely illuminates the inspection surface 200. In this example, the inspection surface 200 is a blade shell 73, in particular, an inner surface of the blade shell 73.

[0047] The inspection surface 200 in these figures includes two different types of defects: wrinkles 210 and bubbles 220 .

[0048] 5a, a directional light source 110 emits directional light 112 towards an inspection surface 200. The directional light 112 comprises a beam angle 111 of less than 90°. An angle 113 from the centre line of the light source to the inspection surface is up to 70 degrees.

[0049] Examples of directional light sources 110 may be LEDs (light emitting diodes) and lasers. Some directional light sources may include a reflector to control or adjust the light beam. When a reflector is used, an additional light source may be used as a directional light source, such as an incandescent lamp or a fluorescent lamp. The reflector may concentrate the light beam of the incandescent lamp or the light beam of the fluorescent lamp. As a result, the light beam of the incandescent lamp or the light beam of the fluorescent lamp may be focused by the reflector so that the incandescent lamp or the fluorescent lamp directly illuminates the inspection surface. For example, an incandescent lamp, a fluorescent lamp or an LED may be provided with a reflector to limit the light to a beam angle of 90 degrees, specifically to a beam angle of 45°, and more specifically to a beam angle of 30°. A reflector or lens arrangement may be provided around or in front of the light emitter to concentrate the beam angle 111. The directional light source 110 may emit directional light of any suitable wavelength.

[0050] Directed light 112 of FIG5 a illuminates wrinkles 210 and bubbles 220 located within inspection surface 200. Direct illumination of wrinkles 210 by directional light source 110 generates shadows 201. However, no shadows are generated by illuminating bubbles 220. It should be understood that other types of wind turbine defects, such as steps, may also produce shadows when illuminated by directional light source 110.

[0051] Directional light source 110 directs directional light 112 at an angle. The axis of the light beam forms an acute angle 113 with the inspection surface, for example, between 70° and 1°, specifically between 70° and 10°. Thus, the axis of the light beam is not perpendicular to the inspection surface. This makes the shadows created by defects more visible.

[0052] In contrast to the directional light 112 of FIG5a , FIG5b shows a diffuse light source 120 that emits diffuse light 121. The diffuse light 121 of FIG4b is emitted in all directions. In this example, the diffuse light source 120 emits diffuse light 121 at a 360° angle. In other examples, the diffuse light 121 may be emitted at a beam angle greater than 90° (e.g., greater than 180°).

[0053] Because the diffuse light 121 is not focused in a single direction, the shadow formed by one ray of light is illuminated by other rays of light. Therefore, a well-defined shadow cannot be formed. In this sense, when the wrinkles 210 are illuminated with diffuse light projected in multiple directions, no shadow is formed. The diffuse light 121 is normally reflected by the surface in a substantially uniform manner. However, the air pockets or bubbles 220 scatter the diffuse light 121 projected onto the bubbles 220. Therefore, the light reflected by the bubbles 220 is different from the light reflected by other parts of the surface of the inspection surface 200. Other types of defects, such as delamination, can also scatter the diffuse light 121. The diffuse light 121 can also be effectively used to identify these other types of defects. The diffuse light source 120 can emit diffuse light of any suitable wavelength.

[0054] Image capture device 130 is configured to capture an image of inspection surface 200. In some examples, image capture device 130 may include a digital camera, such as an optical digital camera and / or a video camera. In some examples, image capture device 130 may include an infrared camera. Image capture device 130 may capture light of any suitable wavelength. In some examples, the image capture device may capture wavelengths of visible light falling between 400 nm and 700 nm. In some examples, the image capture device may capture the infrared spectrum from 700 nm to 1200 nm.

[0055] Image capture device 130 may acquire or capture an image of inspection surface 200 when illuminated by directed light 112, as depicted in Figure 5a. Thus, image capture device 130 may capture shadow 201 generated by directed light facing wrinkle 210 in this first image.

[0056] Image capture device 130 can also capture an image of inspection surface 200 when illuminated with diffuse light 121, as shown in FIG5b. Because the light reflected by bubble 220 is different from the light reflected by the surrounding surface, bubble 220 can be identified in the second image captured by image capture device 130.

[0057] Wind turbine blade inspection system 100 also includes a controller 140. Controller 140 may control the operation of wind turbine blade inspection system 100. Controller 140 in these figures is communicatively coupled to directional light source 110, to diffuse light source 120, and to image capture device 130.

[0058] The controller 140 is configured to selectively activate the directional light source 110 or the diffuse light source 120 to illuminate the inspection surface 200 of the wind turbine blade 7. The controller 140 is also configured to receive images of the inspection surface 200 from the image capture device 130 when illuminated by the directional light source 110 and when illuminated by the diffuse light source 120. Thus, the controller 140 can receive a first image (illuminated by the directional light) and a second image (illuminated by the diffuse light).

[0059] In some examples, controller 140 can selectively instruct image capture device 130 to capture images of inspection surface 200 when illuminated by directional light source 110 and when illuminated by diffuse light source 120. For example, controller 140 can be configured to activate directional light source 110 and instruct image capture device 130 to capture a first image when inspection surface 200 is illuminated by directional light 112. Controller 140 can also be configured to deactivate or turn off directional light source 110 and activate or turn on diffuse light source 120. Controller 140 can then instruct image capture device 130 to obtain a second image of inspection surface 200 when inspection surface 200 is illuminated by diffuse light 121.

[0060] The controller 140 is also configured to analyze the received images of the inspection surface to detect defects. The analysis or processing of the images detects or identifies defects on the inspection surface 200.

[0061] In some examples, controller 140 may compare the image from image capture device 130 with a reference image. For example, controller 140 may compare the received image with a reference image that does not have defects. Differences between the received image and the reference image may indicate a defect in inspection surface 200.

[0062] In some examples, controller 140 may be configured to determine the type of defect if a defect is determined in inspection surface 200. For example, a defect identified in an image obtained using directional light 112 may indicate a wrinkle or a step. On the other hand, when controller 140 detects a difference between a reference image and an image obtained using diffuse light 121, the difference may indicate that the defect is at least one of a void, an air pocket, a delaminated area, and / or a delamination.

[0063] In some examples, multiple reference images may be stored in a reference image database. Controller 140 may compare the image of inspection surface 200 with the reference images in the reference image database, thereby improving defect recognition.

[0064] Reference images can include examples without defects, as well as images with defects. Therefore, the reference image database can include multiple images with different defects. For example, the reference image database can include a collection of images with wrinkle defects, a collection of images with step defects, a collection of images with void defects, a collection of images with bubble defects, and a collection of images with delamination or peeling defects. Comparing images captured by image capture device 130 with reference images depicting different defects can improve the identification of specific defects. Consequently, the accuracy of determining defect types can be increased.

[0065] The reference image database can be updated with images obtained during blade inspection. Furthermore, the reference image database can be updated manually or by using machine learning methods. The reference image database can be further updated with new wind turbine blade defects.

[0066] In some examples, analyzing images can include classifying them into images without defects and images with potential defects. Images with potential defects can then be further analyzed, for example, by comparison with multiple reference images. Categorizing images can thus reduce the data and time required for inspection. Image classification can employ statistical image processing and / or machine learning methods.

[0067] The controller 140 may be configured to analyze the image using a supervised model. Examples of supervised models may include convolutional neural networks (CNNs), support vector machines (SVMs), and / or decision trees. For example, the controller 140 may be configured to analyze the image using a convolutional neural network. Analyzing the image using a convolutional neural network may include classifying, locating, and / or segmenting the image. In some examples, classification, locating, and segmenting may be sequential tasks. In some examples, at least two of these tasks may be performed together.

[0068] Using deep learning models to analyze images can improve the accuracy and efficiency of blade defect recognition. It can also improve the determination of the blade defect type.

[0069] A deep learning algorithm can be used to train the convolutional neural network. A large amount of data can be considered to detect blade defects and / or their properties. The convolutional neural network can be trained using images of the inspection surface 200 including manually detected defects. Controller 140 can be configured to train the convolutional neural network using images of the inspection surface received from image capture device 130.

[0070] In some examples, controller 140 performs supervised training of a computer-implemented machine learning model using a training dataset that includes one or more images of inspection surface 200 and a label indicating the presence or absence of a defect in each image. Supervised training can also include, for each image, setting an output parameter of the machine learning model that corresponds to the label indicating the presence or absence of a defect.

[0071] In some examples, controller 140 is configured to classify images using a trained convolutional neural network. Controller 140 can therefore be configured to detect defects in images using the trained convolutional neural network. The output of the classification can be an image containing defects or an image without defects.

[0072] In some examples, controller 140 is configured to locate or determine the position of a defect within an image, for example, by using a trained convolutional neural network. In some examples, controller 140 is configured to combine classification and localization to detect defects and locate the defect within the image. In some examples, classification and localization can be performed together. The combination of classification and localization of defects in an image can be referred to as defect detection. Thus, controller 140 can be configured to detect and locate defects within an image using a trained convolutional neural network.

[0073] In some examples, using a convolutional neural network includes image segmentation. Image segmentation techniques separate or divide an image into regions. The controller 140 can therefore be configured to segment the image into image regions. Thus, regions with potential defects can be separated from other regions of the image. These regions can be divided or segmented into pixels. An example of an image segmentation technique can be a region-based convolutional neural network (R-CNN). In a region-based convolutional neural network, the input can include the entire image and the output can include pixels required for subsequent inspection and / or location. Another example of an image segmentation technique can be a region-based fully convolutional network (R-FCN).

[0074] Classification and / or localization techniques can then be used to address areas with potential defects. Thus, the classification and / or localization techniques focus on areas with potential defects. Because the entire image does not need to be analyzed, the data used to detect and / or identify defects can be reduced without compromising accuracy.

[0075] In some examples, wind turbine blade inspection system 100 includes multiple image capture devices 130. Multiple image capture devices 130 can capture images of a set of inspection surfaces 200. In some examples, each image capture device 130 can capture an image of one inspection surface 200 in the set of inspection surfaces. In some examples, several image capture devices 130 can capture images from a single inspection surface. The set of inspection surfaces can extend in a chord-wise direction 38 of wind turbine blade 7. In some examples, the set of inspection surfaces can extend from the leading edge to the trailing edge. Thus, the inspection surfaces can be arranged side by side from leading edge 53 to trailing edge 54. Providing multiple image capture devices 130 can allow for inspection of a surface of wind turbine blade 7, such as a portion of the inner surface of blade shell 73 extending in the chord-wise direction 38, without moving wind turbine blade inspection system 100.

[0076] In some examples, wind turbine blade inspection system 100 includes multiple directional light sources 110 and multiple diffuse light sources 120. This further improves illumination accuracy. The multiple light sources can be used to illuminate a collection of inspection surfaces. In some examples, each directional light source 110 and each diffuse light source 120 is associated with a single inspection surface 200. In other examples, a single inspection surface can be illuminated by several directional light sources 110 and / or several diffuse light sources 120.

[0077] In some examples, wind turbine blade inspection system 100 includes a support structure that supports directional light source(s) 110 , diffuse light source(s) 120 , and image capture device(s) 130 .

[0078] In some examples, the support structure may also support controller 140. Thus, controller 140 may be located at the same support structure. Thus, a wired connection may be used. Thus, the controller may move with the support structure, and thus with (one or more) directional light sources 110, (one or more) diffuse light sources 120, and (one or more) image capture devices 130. In other examples, the controller may be mounted independently of the support structure. In these examples, the controller may be located in a fixed location (e.g., in an adjacent area within a manufacturing plant), and the support structure may be moved along the wind turbine blade.

[0079] In some examples, the wind turbine blade inspection system 100 includes a transport system to move the wind turbine blade inspection system 100 along the spanwise direction 37 of the wind turbine blade 7. For example, the wind turbine blade inspection system 100 may be moved from the root portion 50 to a portion adjacent to the tip portion 72. Thus, the wind turbine blade inspection system 100 may be moved over an inner blade shell surface of a blade shell component (e.g., a suction side blade shell component or a pressure side blade shell component).

[0080] Figures 6a and 6b illustrate a front view and a side view, respectively, of a wind turbine blade inspection system 100 according to an example of the present disclosure. The wind turbine blade inspection system can be used to detect defects in the inner surface of a blade shell component. Therefore, the inspection surface of the wind turbine blade can be located on the inner surface of the blade shell component. The wind turbine blade inspection system 100 of this example includes a support structure 170 that supports a plurality of image capture devices 130, a plurality of directional light sources 110, and a plurality of diffuse light sources 120. The plurality of image capture devices 130, the plurality of directional light sources 110, and the diffuse light sources 120 are arranged along a transverse direction 101.

[0081] The wind turbine blade inspection system 100 also includes a controller 140. The controller 140 can be configured to selectively activate one or more directional light sources from the plurality of directional light sources or one or more diffuse light sources from the plurality of diffuse light sources. Depending on the position of the wind turbine blade inspection system 100 relative to the length of the wind turbine blade, the controller 140 can be configured to select a diffuse light source from the plurality of diffuse light sources and / or a directional light source from the plurality of directional light sources to be activated.

[0082] Controller 140 may also be configured to receive multiple images from multiple image capture devices. In some examples, each pair of images (one obtained with directional light and the other obtained with diffuse light) may be captured from a different inspection surface. In other examples, two or more image capture devices may capture a pair of images from a single inspection surface. In some examples, controller 140 may instruct one or more of the multiple image capture devices to capture images from a collection of inspection surfaces.

[0083] The plurality of directional light sources 110 and the plurality of image capture devices 130 can be arranged at different positions in the longitudinal direction 103. In this example, the plurality of directional light sources 110 are arranged behind the plurality of image capture devices 130 in the longitudinal direction 103. The plurality of directional light sources 110 can be configured to direct the directional light forward so as to illuminate an inspection surface located below the plurality of image capture devices 130. Thus, the directional light can be angled relative to the inspection surface to enhance detection of some types of defects, such as wrinkles and / or steps.

[0084] In other examples, multiple directional light sources 110 can be arranged in front of multiple image capture devices 130. In these examples, multiple directional light sources 110 can direct light backward to form an angle with the inspection surface.

[0085] In this example, the support structure 170 includes a center frame 150 extending in a vertical direction 102 from a center frame lower portion 153 to a center frame upper portion 154. One or more columns may extend in the vertical direction 102. The center frame 150 in these figures includes a first side column 151 and a second side column 152 arranged at opposite sides of the center frame 150. A transverse bar may connect the first side column 151 to the second side column 152.

[0086] The center frame upper portion 154 in these figures includes a rear upper transverse bar 155 connecting the first side post 151 to the second side post 152. A first side upper longitudinal bar 157 and a second side upper longitudinal bar 158 extend forward from the first side post 151 and from the second side post 152, respectively, in the longitudinal direction 103. A front upper transverse bar 156 connects the front ends of the first side upper longitudinal bar 157 and the second side upper longitudinal bar 158. Therefore, the front upper transverse bar 156 is spaced a distance apart from the rear upper transverse bar 155 in the longitudinal direction 103.

[0087] In these figures, the center frame 150 includes first and second longitudinal struts 142 extending from lower portions of the first and second side posts 151, 152, respectively, to a front upper transverse bar 156. The center frame 150 in these figures also includes a front lower transverse bar 143 that connects the center portion of the first longitudinal strut to the center portion of the second longitudinal strut 142. The front lower transverse bar 143 can be arranged between the front upper transverse bar 156 and the rear upper transverse bar 155 along the longitudinal direction 103. Therefore, the front upper transverse bar 156 is arranged forward of the front lower transverse bar 143.

[0088] In these figures, support structure 170 includes a first side wing 181 and a second side wing 182. Side wings 181 and 182 are connected to opposite sides of center frame upper portion 154. Side wings 181 and 182 extend a certain length in transverse direction 101. In this example, side wings 181 and 182 respectively include a first side transverse bar 183 and a second side transverse bar 186 extending in transverse direction 101. In Figures 5a and 5b, first side transverse bar 183 is connected to front upper transverse bar 156 and extends outward in transverse direction 101 to a first side transverse bar end 185. Similarly, second side transverse bar 186 is connected to front upper transverse bar 156 and extends outward to a second side transverse bar end 188. In this example, first side transverse bar 183 and second side transverse bar 186 are spaced apart from first side posts 151 and second side posts 152 in longitudinal direction 103.

[0089] The first side upper support rod 184 can extend in an inclined manner from the first side transverse rod end 185 to the first side column 151. Similarly, the second side upper support rod 187 can be provided on the second side wing 182.

[0090] The support structure 170 in these figures supports a plurality of image capture devices 130a, 130b, 130c, 130d, 130e, and 130f. These image capture devices 130 are arranged along the transverse direction 101 of the wind turbine blade inspection system 100. Thus, these image capture devices 130 can acquire a plurality of images arranged in the transverse direction 101. The example in these figures includes six image capture devices; however, other suitable numbers of image capture devices are also possible.

[0091] In these figures, image capture devices 130e and 130f are supported by the center frame lower portion 153. In particular, image capture devices 130e and 130f are connected to the front lower transverse bar 143. Image capture devices 130e and 130f are arranged adjacent the vertical axis of the wind turbine blade inspection system 100.

[0092] In these figures, image capture devices 130b and 130c are positioned at the center frame upper portion 154, specifically at the front upper transverse bar 156. Furthermore, image capture devices 130a and 130d are positioned at the first and second transverse bar ends 185 and 188, respectively. Image capture devices 130a, 130b, and 130e are generally oriented toward the first side, while image capture devices 130c, 130d, and 130f are generally oriented toward the second side. Image capture devices 130e and 130f are positioned behind image capture devices 130a, 130b, 130c, and 130d in the longitudinal direction 103. This longitudinal offset allows for an increased surface area to be inspected.

[0093] The bracket can connect the image capture device 130 to the support structure 170. In some examples, the bracket fixedly connects the image capture device 130 to the support structure 170. In other examples, the bracket rotatably connects the corresponding image capture device 130 to the support structure. Thus, the image capture device can be oriented to a desired angle.

[0094] In some examples, the orientation of the image capture device can be fixed for inspecting the entire inner surface of a blade shell component for a given wind turbine blade shape. In other examples, the orientation of the image capture device can be adjusted depending on the location of the wind turbine blade inspection relative to the length of the wind turbine blade. For example, when wind turbine blade inspection system 100 is at root 50, the orientation of image capture devices 130a and 130d can be different than when wind turbine blade inspection system 100 is at the region with the maximum chord. Orienting image capture device 130 based on the spanwise location of wind turbine blade inspection system 100 can improve the accuracy of images obtained from the inspection surface. In some examples, a controller can adjust the orientation of the image capture device. Additionally or alternatively, orientation of the image capture device can be performed manually.

[0095] In some examples, first wing 181 and / or second wing 182 can be movable in vertical direction 102 relative to center frame 150. Additionally or alternatively, first lateral transverse bar 183 and / or second lateral transverse bar 186 can be extendable. Thus, the position of some image capture devices can be adjusted to different wind turbine blade shapes. Controller 140 can be configured to move one or more of the plurality of image capture devices by actuating support structure 170, for example, by moving first wing 181 and second lateral wing 182 and / or by extending or retracting first lateral transverse bar 183 and / or second lateral transverse bar 186. Additionally or alternatively, the image capture device can be manually positioned by moving support structure 170.

[0096] The wind turbine blade inspection system 100 in these figures includes a plurality of diffuse light sources 120. Specifically, the example in these figures includes six diffuse light sources 120a, 120b, 120c, 120d, 120e, and 120f. However, in other examples, a different number of diffuse light sources may be provided. The diffuse light sources may be capable of diffusely illuminating an inspection surface to be captured by the plurality of image capture devices 130.

[0097] The plurality of diffuse light sources in these figures are arranged along the transverse direction 101. Diffuse light sources 120a and 120d are arranged at the first side transverse bar end 185 and the second side transverse bar end 188, respectively. Diffuse light sources 120b and 120c are arranged at the center frame upper portion 154, specifically at the front upper transverse bar 156. Diffuse light sources 120e and 120f of this example are arranged at the center frame lower portion 153, specifically at the front lower transverse bar 143.

[0098] In this example, each diffuse light source 120 is associated with an image capture device 130. For example, diffuse light source 120a is associated with image capture device 130a. The diffuse light sources 120 in these figures are arranged adjacent to or around the corresponding image capture device 130. This can avoid interference and undesirable shadows.

[0099] The diffuse light sources 120 can be connected to the support structure 170 by corresponding brackets of the image capture devices 130. For example, a single bracket can connect the diffuse light source 120d and the image capture device 130d to the second wing 182. Thus, the diffuse light source(s) 120 can be oriented as explained with respect to the image capture device(s) 130.

[0100] The support structure 170 in these figures includes a first side articulated arm 161 and a second side articulated arm 162. Each articulated arm can support one or more directional light sources from a plurality of directional light sources. In Figures 5a and 5b, the first side articulated arm 161 and the second side articulated arm 162 are arranged on opposite sides of the central frame 150. The first side articulated arm 161 and the second side articulated arm 162 in these figures extend substantially in the transverse direction 101. In these figures, the articulated arms 161 and 162 are rotatably connected to the central frame 150.

[0101] In Figures 6a and 6b, a first side articulated arm 161 supports directional light sources 110a and 110b, and a second side articulated arm 162 supports directional light sources 110c and 110d. In these figures, each of these arms 161 and 162 is rotatably connected to the center frame lower portion 153, and in particular, to a corresponding side of the center frame lower portion 153. The articulated arms 161 and 162 of this example are arranged at the rear of the center frame 150.

[0102] In these figures, the first side articulated arm 161 includes a first side inner rod 163 and a first side outer rod 164 connected to each other by a first side rotary joint 165. Therefore, the first side outer rod 164 can rotate about the rotary joint 165 to form an angle relative to the first side inner rod 163. The first side actuator can move the first side outer rod 164 relative to the first side inner rod 163. In Figures 6a and 6b, the directional light source 110a is connected to the first side outer rod 164, and the directional light source 110b is connected to the first side inner rod 163. Therefore, the positions of the directional light sources 110a and 110b can be adjusted by moving the first side inner rod 163 and / or the first side outer rod 164.

[0103] Similar to the first side articulated arm 161, the second side articulated arm 162 in these figures includes a second side inner rod 166 and a second side outer rod 167 connected by a second side rotation joint 168. In Figures 6a and 6b, the directional light source 110c is connected to the second side inner rod 166, and the directional light source 110d is connected to the second side outer rod 167. A second side actuator 169 can move the second side outer rod 167 relative to the second side inner rod 166.

[0104] In some examples, multiple directional light sources can be rotatably connected to corresponding rods 163, 164, 166 and 167. Directional light sources 110a, 110b, 110c and 110d can rotate around the longitudinal axis of corresponding rods 163, 164, 166 and 167. The directional light can be adjusted in the span direction 37 of the wind turbine blade 7. Therefore, the angle between the directional light and the inspection surface can be adjusted. This can increase the ability to detect some types of defects (such as wrinkles and / or steps). The multiple directional lights in these figures are arranged behind multiple image capture devices. The directional lights of Figures 6a and 6b are configured to direct light forward in an oblique manner. For example, the axis of the light beam can form an angle between 45° and 10° with the inspection surface.

[0105] In some examples, the controller 140 can be configured to instruct the connecting element to rotate the diffuse light source about the longitudinal axis of the corresponding rod. Additionally or alternatively, the rotation can be performed manually.

[0106] When multiple directional light sources are connected to articulated arms 161 and 162, the position of the directional light sources can be adjusted to the specific shape of a wind turbine blade shell component. In this example, first side outer rod 164 and second side outer rod 167 are extendable. This provides additional adjustment to the shape of the wind turbine blade shell component. Consequently, a substantially fixed distance between the directional light sources and the inspection surface can be maintained. This improves the accuracy of defect identification in large wind turbine blades. For example, when inspecting the inner surface of a wind turbine blade shell component from root to tip, a distance between the directional light sources and the inspection surface can be maintained between 20 cm and 60 cm.

[0107] In some examples, a controller can control the position of first and second articulated arms 161 and 162. For example, the controller can actuate actuators to move first and second articulated arms 161 and 162 relative to center frame 150. Furthermore, the controller can control the rotation of outer rods 164 and 167 relative to inner rods 163 and 166. First and second side actuators 169 can move outer rods 164 and 167 relative to inner rods 163 and 166. First and / or second side actuators 169 can include hydraulic actuators. Alternatively or additionally, an operator can move the rods to a specific position.

[0108] The wind turbine blade inspection system 100 of this example includes a transport system 190 for moving or displacing the wind turbine blade inspection system along the spanwise direction 37 of the wind turbine blade. The transport system 190 of this example is connected to the center frame lower portion 153. In this example, the transport system 190 includes a plurality of wheels. These wheels are rotatable on the surface of the blade shell 73 to displace the wind turbine blade inspection system 100 along the spanwise direction 37.

[0109] In other examples, the transport system 190 may include one or more longitudinal guides extending along the length of the wind turbine blade. For example, one longitudinal guide may be adjacent to the trailing edge of the blade, and another longitudinal guide may be adjacent to the leading edge. A drive mechanism may move the wind turbine blade inspection system over the longitudinal guides.

[0110] In some examples, the transport system 190 includes a power system, such as an electric motor. The controller 140 may control the power system to power the transport system 190 to move the wind turbine blade inspection system 100. In other examples, the transport system 190 may be driven by an operator.

[0111] In some examples, conveyor system 190 includes a speed sensor to determine the speed of wind turbine blade inspection system 100 as it moves along spanwise direction 37. The speed sensor can measure the rotation of the wheels. A controller can receive the speed and control the conveyor system to maintain the speed within certain speed limits. For example, the controller can control the operation of a power system of conveyor system 190 to control the speed of wind turbine blade inspection system 100.

[0112] In these figures, the lower center frame portion 153 includes a platform for accommodating the controller 140. Thus, the controller 140 of this example can be moved with the support structure 170. In this example, the controller 140 is embedded in a computer. In other examples, the controller 140 can be, for example, a smartphone or a server.

[0113] The wind turbine blade inspection system 100 of this example includes a user interface device 145, such as a monitor. The controller 140 may output data regarding the detection and / or determination of defects to the user interface device 145. The user interface device 145 may then display the data.

[0114] In some examples, wind turbine blade inspection system 100 may include a positioning sensor to position wind turbine blade inspection system 100. For example, the positioning sensor may provide a position of wind turbine blade inspection system 100 relative to the length of wind turbine blade 7. Controller 140 may obtain the position of wind turbine blade inspection system 100 from the positioning sensor. Based on the position of wind turbine blade inspection system 100, controller 140 may also be configured to locate a defect if a defect is detected in inspection surface 200.

[0115] In this example, wind turbine blade inspection system 100 includes a plurality of distance sensors 195a, 195b, 195c, and 195d. These distance sensors can be used to determine the distance between a directional light source and an inspection surface. In this example, each directional light source 110a, 110b, 110c, and 110d is associated with a distance sensor 195a, 195b, 195c, and 195d. In Figures 6a and 6b, first side outer rod 164 supports distance sensor 195a, first side inner rod 165 supports distance sensor 195b, second side inner rod 166 supports distance sensor 195c, and second side outer rod 167 supports distance sensor 195d. The use of these distance sensors can improve the consistency and repeatability of illuminating a wind turbine blade surface with a directional light source.

[0116] In this example, the distance sensor is a LiDAR sensor. In other examples, ultrasonic sensors, capacitive sensors, infrared sensors, or other types of proximity sensors may also be used.

[0117] In some examples, controller 140 can obtain a distance between the directional light source and the inspection surface from a distance sensor. Based on the obtained distance, controller 140 can instruct the corresponding articulated arm to position the directional light source at a predetermined distance, such as within certain limits. For example, the controller can be configured to ensure that the light source is at a distance of between 20 cm and 60 cm from the inner surface of a wind turbine blade shell component (e.g., a suction side shell component or a pressure side shell component).

[0118] In some examples, wind turbine blade inspection system 100 may include tilt and / or proximity sensors. The tilt sensors may provide a directional light source and / or diffuse light source and / or the tilt of the image capture device. A controller may receive data from these tilt sensors to modify its orientation. Proximity sensors may be used to detect objects in the path of the wind turbine blade inspection system.

[0119] The tilt sensor can sense the tilt of wind turbine blade inspection system 100 in both longitudinal direction 103 and transverse direction 101. Deviations from the expected tilt can then be detected by controller 140. This deviation can then be corrected, for example, by actuating an articulated arm. The output of the tilt sensor can be used to detect the spanwise position of the wind turbine blade inspection system. The configuration of the wind turbine blade inspection system can then be adapted to the spanwise position.

[0120] Figure 7 A wind turbine blade inspection system 100 according to an example of the present disclosure is schematically shown. The wind turbine blade inspection system 100 of this figure is similar to the wind turbine blade inspection system 100 depicted in Figures 6a and 6b. However, the plurality of directional light sources 110 further includes a directional light source 110e. Directional light source 110e is disposed at the lower portion 153 of the central frame. Directional light source 110e can thus illuminate an inspection surface disposed in front of mobile system 190. Directional light source 110e can further improve the directional illumination of the inspection surface.

[0121] Figure 8 A wind turbine blade inspection system 100 according to an example of the present disclosure is schematically illustrated. In this figure, wind turbine blade inspection system 100 is inspecting blade defects in the inner surface of a blade shell component. The blade shell component in this figure is a suction-side shell component. In other examples, the blade shell component may be a pressure-side shell component.

[0122] The wind turbine blade inspection system 100 can detect defects in a surface extending from the leading edge 53 to the trailing edge 54. The wind turbine blade inspection system 100 can inspect a set of inspection surfaces 200a, 200b, 200c, 200c, 200d, and 200e. This set of inspection surfaces 200a, 200b, 200c, 200c, 200d, and 200e extends from the leading edge 53 to the trailing edge 54 in the chordwise direction 38. The inspection surfaces are arranged adjacent to each other to cover the surface extending from the leading edge 53 to the trailing edge 54. The inspection surface may include 4m 2 and 0.5m 2 The surface between.

[0123] Directional light sources 110a, 110b, 110c, and 110d are mounted on articulated arms 161 and 162. In this example, the directional light sources are substantially misaligned relative to the set of inspection surfaces in spanwise direction 37. The set of inspection surfaces is positioned downstream of the directional light sources. The directional light sources of this example can direct the directional light forward so that the axis of the light beam forms an acute angle with the corresponding inspection surface.

[0124] In this example, directional light source 110a is configured to directionally illuminate inspection surface 200a. Directional light source 110b can directionally illuminate inspection surfaces 200b and 200c. Directional light source 110c can directionally illuminate inspection surfaces 200d and 200e. In this example, directional light source 110d is not active due to the shape of the wind turbine blade shell component of this example.

[0125] Thus, the position of the directional light sources can be adjusted to the shape of the blade shell component. The tilt of the directional light sources can thus be adapted to the shape of the blade shell component. The directional light sources 110a, 110b, and 110c can follow the inner contour of the blade shell component at this longitudinal position. The distance between the directional light sources and the inner surface of the blade shell component can also be adjusted within certain limits. This can improve the consistency of the inspection.

[0126] In this figure, each diffuse light source 120a, 120b, 120c, 120d, 120e, and 120f is associated with an image capture device 130a, 130b, 130c, 130d, 130e, and 130f. The diffuse light sources in these figures are arranged around the corresponding image capture device. This can even out the intensity of the diffuse light received by the inspection device and reflected by the image capture device. The diffuse light sources and image capture devices are arranged substantially above the collection of inspection surfaces.

[0127] In this example, image capture device 130a can capture images from at least inspection surface 200a, image capture device 130b can capture images from at least inspection surface 200c, image capture device 130e can capture images from at least inspection surface 200b, image capture device 130c can capture images from at least inspection surface 200d, and image capture device 130f can capture images from at least inspection surface 200d. In this example, image capture device 130d is deactivated. Therefore, controller 140 can control the activation of the image capture devices.

[0128] In some examples, the images captured by the image capture devices partially overlap. For example, there can be at least 25% overlap between the images. For example, the images captured by image capture device 130b and image capture device 130c overlap by at least 25%. These images can be processed to generate multiple chord-wise views of the inspection surface.

[0129] The distance between the image capture device and the corresponding inspection surface can be within certain limits. For example, the distance can be between 0.3 meters and 3 meters.

[0130] The wind turbine blade inspection system 100 of this figure can be moved along the span direction 37 of the wind turbine blade shell component by a conveyor system 190. Thus, the wind turbine blade inspection system 100 can be positioned at different longitudinal positions of the blade shell component. Thus, a single wind turbine blade inspection system 100 can be used to inspect the entire wind turbine blade shell component.

[0131] In some examples, the wind turbine blade inspection system can be positioned at a first position along the spanwise direction 37. One or more inspection surfaces can then be illuminated by one or more directional light sources. When illuminated by the directional light sources, an image capture device can acquire images from the inspection surfaces. The directional light sources can then be deactivated, and one or more diffuse light sources can be activated. The image capture device can then capture images from the inspection surfaces while illuminated by the diffuse light sources. A controller can then analyze the images obtained from the image capture device. In this example, the inspection surface is first illuminated by the directional light sources and then by the diffuse light sources. However, in other examples, the inspection surface is first illuminated by the diffuse light sources and then by the directional light sources.

[0132] Wind turbine blade inspection system 100 can then be moved to a second position to inspect a second set of inspection surfaces 200 extending along the chord-wise direction 38 of the wind turbine blade shell component at the second position. Wind turbine blade inspection system 100 can then repeatedly move the wind turbine blade inspection system to the forward position and inspect the set of inspection surfaces 200 extending in the chord-wise direction 38 at the forward position. Thus, wind turbine blade inspection system 100 can inspect the wind turbine blade shell component in a single pass. While moving along spanwise direction 37, wind turbine blade inspection system 100 can adjust to the shape of the inner surface of the wind turbine blade shell component. For example, first side arm 161 and second side arm 162 can adopt different configurations depending on their position along spanwise direction 37, i.e., their longitudinal position relative to the length of the wind turbine blade.

[0133] In some examples, wind turbine blade inspection system 100 can be moved along spanwise direction 37 of a wind turbine blade shell, for example, from blade root 50 to blade tip 72, and can acquire images at different locations while being illuminated by directional light source 110. Thus, multiple images can be acquired in a first scan while being illuminated by directional light source 110. Then, wind turbine blade inspection system 100 can be moved again along spanwise direction 37 and can acquire images while being illuminated by diffuse light source 120. Multiple images can be captured in a second scan while being illuminated by diffuse light source 120. In other examples, images can be acquired while being illuminated by diffuse light source 120 in the first scan and by directional light source 110 in the second scan.

[0134] In some examples, wind turbine blade inspection system 100 can be moved spanwise at a substantially constant speed. The exposure time of the image capture device and the illumination intensity of the light source can be adjusted to minimize undesirable movement of the light source and image capture device while maintaining acceptable image quality. Therefore, the speed can be determined while taking into account the exposure time and illumination intensity.

[0135] In some examples, wind turbine blade inspection system 100 may inspect wind turbine blade shell components from blade root 50 to blade tip 37. Additionally or alternatively, wind turbine blade inspection system 100 may inspect wind turbine blade shell components from blade tip 37 to blade root.

[0136] In some examples, the wind turbine blade inspection 100 can inspect a portion of a wind turbine blade shell component in one manner and then inspect the portion of the wind turbine blade shell component in an opposite manner. For example, the wind turbine blade inspection system can move from the blade root 50 to the middle portion of the blade and then move in the opposite direction toward the blade root 50.

[0137] Figure 9 3 is a block diagram of a computer-implemented method for detecting defects in a wind turbine blade according to an example of the present disclosure. A wind turbine blade inspection system 100 according to an example herein can be used in a computer-implemented method 300. Method 300 can be used to detect blade defects in the inner surface of a blade shell component (e.g., a suction shell component or a pressure shell component). The inner surface of the blade shell component can be inspected while the blade shell component is in a mold after molding, such as by resin infusion or prepreg techniques.

[0138] At block 310, activation of the directional light source 110 of the wind turbine blade inspection system 100 is indicated to illuminate the inspection surface 200 of the wind turbine blade 7. The controller 140 may control the directional light source 110 to selectively turn on and off.

[0139] Method 300 also includes receiving, by controller 140, a first image of inspection surface 200 illuminated by the directional light source, as indicated at block 320. Method 300 may also include instructing image capture device 130 to capture the first image of inspection surface 200 when illuminated by directional light source 110.

[0140] At block 330 , the diffuse light source 120 of the wind turbine blade inspection system 100 is activated to diffusely illuminate the inspection surface 200 . The controller 140 may selectively activate and deactivate the diffuse light source 120 .

[0141] At block 340 , a second image of inspection surface 200 illuminated by directional light source 110 is represented as being received by controller 140 . In some examples, controller 140 may instruct image capture device 130 to capture the second image of inspection surface 200 when illuminated by diffuse light source 120 .

[0142] In some examples, controller 140 may activate diffuse light source 120 after turning off directional light source 110. In other examples, controller 140 may first activate diffuse light source 120 and receive the second image, and then activate directional light source 110 to receive the first image.

[0143] The method 300 also includes analyzing, by the controller 140 , the first image and the second image of the inspection surface 200 to detect defects in the inspection surface 200 , as indicated at block 350 .

[0144] Controller 140 can analyze the images according to any of the examples herein. For example, method 300 can include analyzing the first and second images of inspection surface 200 using a convolutional neural network. The convolutional neural network can be according to any of the examples herein. For example, using the convolutional neural network can include classifying, localizing, and / or segmenting the images.

[0145] In some examples, the first image and the second image can be used to train a convolutional neural network. According to examples herein, these images can be used to train a convolutional neural network.

[0146] Method 300 may include determining the type of defect if a defect is detected in inspection surface 200. Determining the type of defect may be performed according to any of the examples described herein. As previously described, the first and second images may be compared to a reference image, and / or a convolutional neural network may be used to determine the type of defect.

[0147] In some examples, method 300 further includes determining a location of the defect if a defect is detected in inspection surface 200 . Positioning sensors may be used to determine the position of wind turbine blade inspection system 100 along spanwise direction 37 of the wind turbine blade. A controller may receive the position of wind turbine blade inspection system 100 from the positioning sensors. The location of the defect may then be determined. Convolutional neural networks may also be used to locate the defect.

[0148] Analysis of the image can also be used to determine the location of defects. The controller can estimate the location of the defect by converting or correlating the pixels of the image into an estimate of the location of the defect in the blade shell. This correlation can also be used to determine the shape and / or size of the defect in the blade shell. The correlation can include converting the pixels of the image into millimeters. Furthermore, these pixel-to-millimeters conversions can be pre-programmed for each image capture device and at each location on the blade shell, based on the CAD outline of the blade shell and the position of the image capture device relative to the surface.

[0149] Determining the location and / or shape and / or size of a defect location identified in the inspection surface can be used to assess the severity of the defect. Less severe defects may be allowable or repaired. If a defect is determined to be repairable, the controller can output the defect size and location for subsequent repair tasks. For example, the controller can generate composite or stitched images of the blade shell component from images captured using an image capture device. These composite images can be compared with a geometric model of the blade shell component. For example, these composite images can be superimposed on a CAD model of the blade shell component to generate an inspection report.

[0150] This allows for mapping defects on the surface of the blade shell component. For example, a defect heat map can be generated. This can improve the detection of defects in wind turbine blades. These heat maps can be generated for different process parameters and / or for different wind turbine blade molds. These different heat maps can then be compared to optimize process parameters to reduce defects.

[0151] In some examples, wind turbine blade inspection system 100 includes a plurality of directional light sources 110 and a plurality of diffuse light sources 120. These light sources 110 and 120 can be used to illuminate a set of inspection surfaces 200. The set of inspection surfaces is arranged at a longitudinal position relative to the length of wind turbine blade 7. Thus, the set of inspection surfaces can extend in a chordwise direction 38. The inspection surfaces in the set of inspection surfaces can extend edge-to-edge in the chordwise direction from the trailing edge to the leading edge.

[0152] Method 300 may include activating a plurality of directional light sources 110 of wind turbine blade inspection system 7 to illuminate corresponding inspection surfaces in a set of inspection surfaces. Method 300 may also include receiving, by controller 140, a set of images of the set of inspection surfaces 200 illuminated by the plurality of directional light sources 110. In some examples, each inspection surface in the set of inspection surfaces is illuminated by a directional light source from the plurality of directional light sources 110. In other examples, one or more directional light sources from the plurality of directional light sources 110 may illuminate several inspection surfaces in the set of inspection surfaces 200.

[0153] Method 300 may include activating a plurality of diffuse light sources 120 of wind turbine blade inspection system 100 to diffusely illuminate corresponding inspection surfaces in the set of inspection surfaces 200, and receiving, by controller 140, a set of second images of the set of inspection surfaces illuminated by the plurality of diffuse light sources 120. In some examples, each diffuse light source may illuminate one inspection surface in the set of inspection surfaces. In other examples, one diffuse light source may illuminate several inspection surfaces, or one inspection surface may be diffusely illuminated by several diffuse light sources.

[0154] The plurality of image capture devices 130 may be activated to acquire a first set of images and a second set of images. These images may then be analyzed by the controller 140 to detect defects in the set of inspection surfaces 200.

[0155] Additionally, method 300 may include determining a position of wind turbine blade inspection system 100. For example, a positioning sensor may be used to determine a longitudinal position of the wind turbine blade inspection system relative to the longitudinal length of a wind turbine blade or blade shell component. Based on the determined position, the method may further include instructing wind turbine blade inspection system 100 to move the plurality of directional light sources 110 to a predetermined configuration. The position of directional light sources 110 may thus be adapted to the shape of the blade shell component. The plurality of directional light sources 110 may include different predetermined configurations based on longitudinal position along spanwise direction 37. For example, at a first position corresponding to 20% of the length of the wind turbine blade, the directional light sources 110 may be arranged in a first predetermined configuration, and at a second position corresponding to 60% of the length of the wind turbine blade, the directional light sources 110 may be arranged in a second predetermined configuration.

[0156] In some examples, instructing wind turbine blade inspection system 100 to move plurality of directional light sources 110 may include actuating first articulated arm 161 and second articulated arm 162. First articulated arm 161 and second articulated arm 162 may be actuated according to any example herein.

[0157] In some examples, method 300 may receive geometric data regarding the size and / or shape of a blade shell component to be inspected. This data may be a CAD geometry of the blade shell component. The spanwise movement path of the wind turbine blade inspection system and the position of the directional light source relative to the inner blade shell may be predefined prior to inspecting the blade shell component.

[0158] In some examples, the method may include obtaining the type or model of blade to be inspected. In some examples, the type of blade may be received from a user interface device. In some examples, the controller may receive dimensional data regarding the wind turbine blade, for example, from an image capture device. This dimensional data may be compared with a dimensional database to determine the type of blade. Once the blade type is obtained, the controller may obtain a configuration of the wind turbine blade inspection device. The controller may obtain a position of the image capture device, such as the height of the image capture device from the inspection surface and / or the position of the image capture device in a transverse and / or longitudinal direction.

[0159] In some examples, the method may include generating composite images based on images received from an image capture device. These composite images may represent a region of the blade shell component extending from a leading edge to a trailing edge. Partially overlapping the images acquired by the image capture device may improve the generation of the composite images.

[0160] In some examples, the method includes representing the identified defect on a composite image. This may include detecting the defect size and location according to any example herein. The composite image with the identified defect may be superimposed on a CAD model of the blade shell component. The method may further include generating data including the location and type of the defect. The data may include an inspection report.

[0161] In some examples, the method may include generating defect heat maps for different blade shell components. These heat maps may be generated for different process parameters and / or for different wind turbine blade molds. These different heat maps may then be compared to optimize process parameters to reduce defects.

[0162] In some examples, the method 300 includes repeatedly activating the directional light source(s) 110 and receiving a first image of the inspection surface 200 or a set of first images of the set of inspection surfaces for a plurality of inspection surfaces 200 or a plurality of sets of inspection surfaces 200 arranged at different longitudinal positions relative to the length of the wind turbine blade 7. The method 300 may also include repeatedly activating the diffuse light source(s) 120 and receiving a second image of the inspection surface 200 or a second set of images of the set of inspection surfaces for a plurality of inspection surfaces 200 or a plurality of sets of inspection surfaces 200 arranged at different longitudinal positions relative to the length of the wind turbine blade 7. Additionally, the method may include analyzing, by the controller 140, the first and second images of the plurality of inspection surfaces 200 arranged at different longitudinal positions relative to the length of the wind turbine blade 7 to detect defects in the plurality of inspection surfaces 200.

[0163] Figure 10Ais a block diagram of a computer-implemented method 400 for detecting defects in a wind turbine blade shell according to an example of the present disclosure. Blocks 310, 320, 330, 340, and 350 may be according to any example disclosed herein.

[0164] At block 410, geometric data of the blade shell component to be inspected is obtained. The geometric data may include the size and / or shape of the blade shell component to be inspected. The geometric data may include a CAD model of the blade shell component. Thus, the method may obtain the model or type of the wind turbine blade to be inspected.

[0165] At block 420, a position of the directional light source relative to the inner surface of the blade shell component to be inspected is determined based on the geometric data.The position of the directional light source or light sources may be determined prior to inspecting the wind turbine blade shell component.

[0166] In some examples, the method 400 may further include determining a movement path of the wind turbine blade inspection system along the spanwise direction of the blade shell component to be inspected based on the geometric data. Thus, the path may be determined before inspecting the wind turbine blade shell.

[0167] In some examples, method 400 may also include actuating a first articulated arm and a second articulated arm including the one or more directional light sources based on the determined position of the directional light sources.

[0168] Figure 10B is a block diagram of a computer-implemented method 500 for detecting defects in a wind turbine blade shell according to an example of the present disclosure.

[0169] At block 415 , a CAD model of the blade shell component is obtained. The CAD model may be an example of geometric data of the blade shell component.

[0170] At block 311, a plurality of directional light sources for wind turbine blade inspection is shown to illuminate corresponding inspection surfaces from a set of inspection surfaces of the wind turbine blade. The set of inspection surfaces is arranged at longitudinal positions relative to the length of the wind turbine blade. The positions of the plurality of directional light sources may be determined based on a CAD model of the blade shell component. According to any example herein, the plurality of directional light sources may illuminate the inspection surfaces.

[0171] The method 500 also includes receiving, by the controller, a set of first images of the set of inspection surfaces illuminated by the plurality of directional light sources. The image capture device can capture the set of first images according to any example herein.

[0172] At block 331 , a plurality of diffuse light sources of a wind turbine blade inspection system are represented for diffusely illuminating corresponding inspection surfaces of a set of inspection surfaces. The diffuse light sources may operate according to any example herein.

[0173] As shown at block 341 , the controller may receive a set of second images of a set of inspection surfaces illuminated by a plurality of diffuse light sources.

[0174] At block 351 , a set of first and second images is analyzed. The images may be analyzed according to any of the examples herein. A controller may detect defects contained in the images.

[0175] At block 560, the image containing the defect received from the image capture device is represented as being superimposed on the CAD model. The image containing the defect may be compared to the CAD model to illustrate the location of the defect within the blade shell component.

[0176] The method 500 further includes generating defect data, as shown at block 570. The generated defect data may include the location and type of the defect. Additionally, the defect data may include the shape and / or size of the defect.

[0177] In some examples, method 500 may also include generating a map of defects on the surface of the blade shell component. In some examples, method 500 may also include generating defect heat maps for different blade shell components based on the defect data. These defect heat maps may be used to compare different blade shell components. Thus, the manufacturing of the blade shell components may be adjusted to reduce the number and severity of defects.

[0178] In some examples, method 500 may also include: Figure 10A The steps described in .

[0179] Figure 10C is a block diagram of a computer-implemented method 600 for detecting defects in a wind turbine blade shell according to an example of the present disclosure. The method 600 includes blocks 420, 310, 320, 330, 340, and 350 according to any example herein.

[0180] At block 610, pixels are retrieved from the first image and the second image. The pixels may be retrieved according to any suitable method.

[0181] Based on the geometry of the blade shell component, the pixels can be converted into dimensions, for example into millimeters, as shown at block 620. Using this correlation, the shape and / or size of the defect can be determined. Furthermore, the location of the defect can be determined. This conversion can be used to overlay the image containing the defect onto the CAD model of the blade shell component.

[0182] The method 600 may also include any steps of any method herein.For example, a plurality of directional light sources may be used to illuminate several areas of the blade shell component.

[0183] Figure 11 1 represents a controller and computing program according to examples of the present disclosure. Controller 140, or computing system, includes a processor 131 that performs operations on data, for example, to detect defects in a wind turbine blade. Processor 131 is configured to perform a method for detecting defects in a wind turbine blade according to examples described herein. Processor 131 can execute a computing program 132 including instructions 133 that cause processor 131 to detect defects in a wind turbine blade according to examples described herein. Controller 140 can be a computer, smartphone, tablet, or server.

[0184] In some examples, processor 131 may be a dedicated processor for detecting defects in wind turbines. In other examples, processor 131 may also control other manufacturing operations.

[0185] The computer program 132 may be embodied on a storage medium (eg, CD-ROM, DVD, USB drive, computer memory, or read-only memory) or carried on a carrier signal (eg, on an electrical or optical carrier signal).

[0186] The computer program may be in the form of source code, object code, a code intermediate source and object code, such as in partially compiled form, or in any other form suitable for implementing the method for detecting defects in wind turbine blades according to the present disclosure. The carrier may be any entity or device capable of carrying a computer program.

[0187] For example, the carrier may comprise a storage medium such as a ROM, for example a CD ROM or a semiconductor ROM, or a magnetic recording medium, for example a hard disk. Furthermore, the carrier may be a transmissible carrier such as an electric or optical signal, which may be transmitted via an electric or optical cable or by radio or other means.

[0188] For reasons of completeness, various aspects of the disclosure are set out in the following numbered clauses: Item 1: A wind turbine blade inspection system for detecting defects in a wind turbine blade, comprising: a directional light source for directionally illuminating an inspection surface of a wind turbine blade; a diffuse light source for diffusely illuminating the inspection surface of the wind turbine blade; an image capture device for capturing an image of the inspection surface; a controller for: selectively activating the directional light source or the diffuse light source; receiving from the image capture device an image of the inspection surface when illuminated by the directional light source and when illuminated by the diffuse light source; and analyzing the received images of the inspection surface to detect defects.

[0189] Clause 2: The wind turbine blade inspection system of clause 1, wherein the controller is configured to selectively instruct the image capture device to capture images of the inspection surface when illuminated by the directional light source and when illuminated by the diffuse light source.

[0190] Clause 3: The wind turbine blade inspection system of any of clauses 1-2, wherein analyzing the image of the inspection surface comprises using a convolutional neural network.

[0191] Clause 4: The wind turbine blade inspection system of clause 3, wherein the controller is configured to train the convolutional neural network using received images of the inspection surface.

[0192] Clause 5: The wind turbine blade inspection system of any of clauses 1-4, wherein the controller is configured to determine a type of defect if a defect is detected in the inspection surface.

[0193] Clause 6: A wind turbine blade inspection system according to any of clauses 1-5, comprising a positioning sensor for determining a position of the wind turbine blade inspection system, and wherein the controller is configured to: obtain the position of the wind turbine blade inspection system from the positioning sensor; and if a defect is detected in the inspection surface, locate the defect.

[0194] Clause 7: The wind turbine blade inspection system according to any one of clauses 1 to 6, comprising a transport system to move the wind turbine blade inspection system along the span direction of the wind turbine blade.

[0195] Clause 8: A wind turbine blade inspection system according to any of clauses 1-7, comprising: a plurality of image capture devices for capturing a set of inspection surfaces extending in a chordwise direction of the wind turbine blade; a plurality of directional light sources; and a plurality of diffuse light sources.

[0196] Clause 9: The wind turbine blade inspection system of clause 8, comprising a support structure supporting the plurality of directional light sources, the plurality of diffuse light sources, and the plurality of image capture devices, wherein the support structure comprises a first side articulated arm and a second side articulated arm, wherein each of the articulated arms supports one or more of the plurality of directional light sources.

[0197] Clause 10: A wind turbine blade inspection system according to any one of clauses 1 to 9, wherein the controller is configured to: obtain geometric data of a blade shell component to be inspected; determine a position of the directional light source relative to an inner surface of the blade shell component to be inspected based on the geometric data; and optionally, determine a movement path of the wind turbine blade inspection system along the span direction of the blade shell component to be inspected based on the geometric data.

[0198] Clause 11: The wind turbine blade inspection system of any of clauses 1-10, wherein the controller is configured to: obtain a model of the wind turbine blade to be inspected; and determine a position of the directional light source.

[0199] Clause 12: A wind turbine blade inspection system according to any of clauses 1-10, wherein the controller is configured to: obtain a CAD model of the blade shell component; superimpose an image containing a defect received from the image capture device on the CAD file; and generate data including a location and type of the defect.

[0200] Item 13: A computer-implemented method for detecting defects in a wind turbine blade, comprising: activating a directional light source of a wind turbine blade inspection system to directionally illuminate an inspection surface of the wind turbine blade; receiving, by a controller, a first image of the inspection surface illuminated by the directional light source; activating a diffuse light source of the wind turbine blade inspection system to diffusely illuminate the inspection surface; receiving, by the controller, a second image of the inspection surface illuminated by the diffuse light source; and analyzing, by the controller, the first image and the second image of the inspection surface to detect defects in the inspection surface.

[0201] Clause 14: The computer-implemented method of clause 13, comprising: instructing the image capture device to capture an image of the inspection surface when illuminated by the directional light source; and instructing the image capture device to capture an image of the inspection surface when illuminated by the diffuse light source.

[0202] Clause 15: The computer-implemented method of any of clauses 13-14, wherein analyzing the first image and the second image of the inspection surface comprises using a convolutional neural network.

[0203] Clause 16: The computer-implemented method of clause 15, comprising training the convolutional neural network with the first image and the second image.

[0204] Clause 17: The computer-implemented method of any of clauses 13-16, comprising determining a type of defect if a defect is detected in the inspection surface.

[0205] Clause 18: The computer-implemented method of any of clauses 13-17, comprising locating the defect if a defect is detected in the inspection surface.

[0206] Clause 19: A computer-implemented method according to any one of clauses 13-18, comprising: repeatedly activating the directional light source and receiving the first image of the inspection surface for a plurality of inspection surfaces arranged at different longitudinal positions relative to the length of the wind turbine blade; repeatedly activating the diffuse light source and receiving the second image of the inspection surface for the plurality of inspection surfaces arranged at different longitudinal positions relative to the length of the wind turbine blade; and analyzing, by the controller, the first image and the second image of the plurality of inspection surfaces arranged at different longitudinal positions relative to the length of the wind turbine blade to detect defects in the plurality of inspection surfaces.

[0207] Clause 20: A computer-implemented method according to any one of clauses 13-19, comprising: activating a plurality of directional light sources of the wind turbine blade inspection system to directionally illuminate corresponding inspection surfaces in a set of inspection surfaces of the wind turbine blade, wherein the set of inspection surfaces is arranged at a longitudinal position relative to the length of the wind turbine blade; receiving, by the controller, a set of first images of the set of inspection surfaces illuminated by the plurality of directional light sources; activating a plurality of diffuse light sources of the wind turbine blade inspection system to diffusely illuminate the corresponding inspection surfaces in the set of inspection surfaces; receiving, by the controller, a set of second images of the set of inspection surfaces illuminated by the plurality of diffuse light sources; and analyzing, by the controller, the set of first images and the set of second images of the inspection surfaces to detect defects in the set of inspection surfaces.

[0208] Clause 21: The computer-implemented method of clause 20, comprising: determining a position of the wind turbine blade inspection system; and instructing the wind turbine blade inspection system to move the plurality of directional light sources to a predetermined configuration based on the determined position.

[0209] Clause 22: A computer-implemented method according to any one of clauses 13-21, comprising: obtaining geometric data of a blade shell component to be inspected; determining a position of the directional light source relative to an inner surface of the blade shell component to be inspected based on the geometric data; and optionally, determining a movement path of the wind turbine blade inspection system along the spanwise direction of the blade shell component to be inspected based on the geometric data.

[0210] Clause 23: The computer-implemented method of Clause 22, comprising actuating a first articulated arm and a second articulated arm comprising the one or more directional light sources based on the determined positions of the directional light sources.

[0211] Clause 24: The computer-implemented method of any of clauses 13-23, comprising: obtaining a model of a wind turbine blade to be inspected; and determining a position of a directional light source.

[0212] Clause 25: A computer-implemented method according to any of clauses 13-24, comprising: obtaining a CAD model of the blade shell component; superimposing an image containing a defect received from the image capture device on the CAD model; and generating defect data including a location and type of the defect.

[0213] Clause 26: The computer-implemented method of clause 25, comprising: generating defect heat maps for different blade shell components based on the generated defect data; and comparing the defect heat maps of the different blade shell components.

[0214] Clause 27: A controller comprising a processor configured to perform the method of any of clauses 13-26.

[0215] Clause 28: A computer program comprising instructions which, when executed by a processor, cause the processor to perform the method of any one of clauses 13-26.

[0216] This written description uses examples to disclose the invention, including preferred embodiments, and also to enable any person skilled in the art to practice the invention, including making and using any devices or systems and performing any incorporated methods. The patentable scope of the invention is defined by the claims and may include other examples that occur to those skilled in the art. These other examples are intended to fall within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements that do not differ substantially from the literal language of the claims. Aspects from the various embodiments described, and other known equivalents for each such aspect, can be mixed and matched by one of ordinary skill in the art to construct additional embodiments and techniques in accordance with the principles of the present application. If reference numerals relating to the drawings are placed in parentheses in the claims, they are used solely to attempt to increase the intelligibility of the claims and should not be construed as limiting the scope of the claims.

Claims

1. A wind turbine blade inspection system (100) for detecting defects in a wind turbine blade (7), comprising: a directional light source (110) for illuminating an inspection surface (200) of a wind turbine blade (7) at an acute angle (113) relative to the inspection surface (200); a diffuse light source (120) for diffusely illuminating the inspection surface (200) of the wind turbine blade (7); an image capture device (130) for capturing an image of the inspection surface (200); A controller (140) configured to: selectively activating the directional light source (110) or the diffuse light source (120); receiving, from the image capture device (130), images of the inspection surface (200) when illuminated by the directional light source (110) and when illuminated by the diffuse light source (120); as well as The received image of the inspection surface (200) is analyzed to detect defects.

2. The wind turbine blade inspection system (100) of claim 1, wherein analyzing the image of the inspection surface (200) comprises using a convolutional neural network.

3. The wind turbine blade inspection system (100) according to any one of claims 1-2, wherein the controller (140) is configured to determine a type of defect if a defect is detected in the inspection surface (200).

4. The wind turbine blade inspection system (100) according to any one of claims 1 to 3, comprising a positioning sensor for determining a position of the wind turbine blade inspection system (100), and wherein the controller (140) is configured to: obtain the position of the wind turbine blade inspection system (100) from the positioning sensor; and If a defect is detected in the inspection surface (200), the defect is located.

5. The wind turbine blade inspection system (100) according to any one of claims 1 to 4, comprising a transport system (190) for moving the wind turbine blade inspection system (100) along the span direction (37) of the wind turbine blade (7).

6. The wind turbine blade inspection system (100) according to any one of claims 1 to 5, comprising: a plurality of image capture devices (130) for capturing a set of inspection surfaces (200) extending in a chordwise direction (38) of the wind turbine blade (7); a plurality of directional light sources (110); as well as A plurality of diffuse light sources (120).

7. The wind turbine blade inspection system (100) according to claim 6, comprising a support structure (170) supporting the plurality of directional light sources (110), the plurality of diffuse light sources (120), and the plurality of image capture devices (130), wherein the support structure (150) comprises a first side articulated arm (161) and a second side articulated arm (162), wherein each of the articulated arms (161, 162) supports one or more directional light sources of the plurality of directional light sources (110).

8. A computer-implemented method (300) for detecting defects in a wind turbine blade (7), comprising: activating (310) a directional light source (110) of a wind turbine blade inspection system (100) to illuminate an inspection surface (200) of a wind turbine blade (7) at an acute angle (113) relative to the inspection surface (200), receiving (320) by a controller (140) a first image of the inspection surface (200) illuminated by the directional light source (110); activating (330) a diffuse light source (120) of the wind turbine blade inspection system (200) to diffusely illuminate the inspection surface (200); receiving (340) by the controller (140) a second image of the inspection surface (200) illuminated by the diffuse light source (120); as well as The first image and the second image of the inspection surface (200) are analyzed (350) by the controller (140) to detect defects in the inspection surface (200).

9. The computer-implemented method (300) of claim 8, wherein analyzing (350) the first image and the second image of the inspection surface (200) comprises using a convolutional neural network.

10. The computer-implemented method (300) of claim 9, comprising training the convolutional neural network with the first image and the second image.

11. The computer-implemented method (300) according to any one of claims 8 to 10, comprising: Repeating the activation of the directional light source (110) and receiving the first image of the inspection surface (200) for a plurality of inspection surfaces (200) arranged at different longitudinal positions relative to the length of the wind turbine blade (7); Repeating activating the diffuse light source (120) and receiving the second image of the inspection surface (200) for a plurality of inspection surfaces (200) arranged at different longitudinal positions relative to the length of the wind turbine blade (7); as well as The first and second images of a plurality of inspection surfaces (200) arranged at different longitudinal positions relative to the length of the wind turbine blade (7) are analyzed by the controller (140) to detect defects in the plurality of inspection surfaces (200).

12. The computer-implemented method (300) according to any one of claims 8 to 11, comprising: activating a plurality of directional light sources (110) of the wind turbine blade inspection system (100) to illuminate corresponding inspection surfaces of a set of inspection surfaces (200) of the wind turbine blade (7), wherein the set of inspection surfaces (200) are arranged at longitudinal positions relative to the length of the wind turbine blade (7); Receiving, by the controller (140), a set of first images of the set of inspection surfaces (200) illuminated by the plurality of directional light sources (110); activating a plurality of diffuse light sources (120) of the wind turbine blade inspection system (100) to diffusely illuminate corresponding inspection surfaces in the set of inspection surfaces (200); receiving, by the controller (140), a set of second images of the set of inspection surfaces (200) illuminated by the plurality of diffuse light sources (120); as well as The set of the first and second images of the inspection surface (200) is analyzed by the controller (140) to detect defects in the set of inspection surfaces (200).

13. The computer-implemented method (300) of claim 12, comprising: determining a position of the wind turbine blade inspection system (100); as well as Based on the determined positions, the wind turbine blade inspection system (100) is instructed to move the plurality of directional light sources (110) to a predetermined configuration.

14. A controller (140) comprising a processor (131) configured to perform the method according to any one of claims 8-13.

15. A computer program (131) comprising instructions (133) which, when said program (131) is executed by a processor (131), cause said processor (131) to perform the method according to any one of claims 8 to 13.

Images