Blade defect detection method, device and equipment and storage medium

Abstract

The invention discloses a blade defect detection method, device and equipment and a storage medium, and relates to the technical field of wind power blades. The method is applied to a defect detection platform, the defect detection platform comprises a control panel, and the defect detection platform is in communication connection with an inspection robot. The method comprises the steps of generating a control instruction in response to a first input of a user on a control panel; and a control instruction is sent to an inspection robot located in the target blade cavity, so that the inspection robot carries out image acquisition on the internal area of the target blade cavity. And receiving an internal image of the target blade cavity sent by the inspection robot. Identifying blade cavity defect feature information in the target blade cavity internal image through a blade cavity defect identification model to obtain a target blade cavity defect identification result; the blade cavity defect identification model is obtained by training based on a blade cavity internal image sample and a corresponding cavity defect label sample. According to the embodiment of the invention, the efficiency and accuracy of blade cavity internal defect detection can be improved.

Description

Technical Field

[0001] The present application belongs to the technical field of wind turbine blades, and in particular, relates to a blade defect detection method, device, equipment, storage medium and computer program product. Background Art

[0002] As the difficulty and complexity of large-scale blade manufacturing increases, wind turbine blades will produce various types of defects such as delamination, glue deficiency, and glue resistance in the inner cavity of the blades during the production process, significantly increasing the risk of blade damage or even failure during wind turbine operation.

[0003] The existing automatic defect detection methods for wind turbine blades are mainly for the detection of blade appearance, mostly with the help of mobile robots, drones and other equipment, focusing on machine equipment development and algorithm optimization for image stitching and positioning, and can only identify external damage defects such as cracks, lightning strikes, carbonization, etc. The identification of internal defects in blade cavities is mostly manual detection, which is inefficient and not very accurate. Summary of the invention

[0004] The embodiments of the present application provide a blade defect detection method, device, equipment, storage medium and computer program product, which can improve the efficiency and accuracy of blade cavity internal defect detection.

[0005] In a first aspect, an embodiment of the present application provides a blade defect detection method, which is applied to a defect detection platform, wherein the defect detection platform includes a control panel, and the defect detection platform is communicatively connected to an inspection robot; the method includes:

[0006] generating a control instruction in response to a first input by a user on the control panel;

[0007] Sending a control instruction to an inspection robot located in the target blade cavity so that the inspection robot collects images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity;

[0008] Receiving the internal image of the target blade cavity sent by the inspection robot;

[0009] The blade cavity defect recognition model is used to identify the blade cavity defect feature information in the target blade cavity internal image to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0010] In a second aspect, an embodiment of the present application provides a blade defect detection method, which is applied to an inspection robot, and the method includes:

[0011] Receive control instructions sent by the control panel on the defect detection platform;

[0012] In response to the control instruction, an image of an internal area of ​​the target blade cavity is collected in the target blade cavity to obtain an internal image of the target blade cavity;

[0013] The target blade cavity internal image is sent to the defect detection platform so that the defect detection platform can identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0014] In a third aspect, an embodiment of the present application provides a blade defect detection device, which is applied to a defect detection platform. The defect detection platform includes a control panel, and the defect detection platform is communicatively connected with an inspection robot. The device includes:

[0015] A receiving module, configured to generate a control instruction in response to a first input from a user on the control panel;

[0016] A sending module is used to send a control instruction to the inspection robot located in the target blade cavity, so that the inspection robot can collect images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity;

[0017] A receiving module, used for receiving the internal image of the target blade cavity sent by the inspection robot;

[0018] The recognition module is used to identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0019] In a fourth aspect, an embodiment of the present application provides a blade defect detection device, which is applied to an inspection robot, and the device includes:

[0020] A receiving module, used for receiving control instructions sent by a control panel on the defect detection platform;

[0021] An acquisition module is used to acquire an image of an internal area of ​​a target blade cavity in response to a control instruction, so as to obtain an internal image of the target blade cavity;

[0022] The sending module is used to send the target blade cavity internal image to the defect detection platform, so that the defect detection platform can identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0023] In a fifth aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the blade defect detection method of the first aspect is implemented.

[0024] In a sixth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the blade defect detection method of the first aspect is implemented.

[0025] In a seventh aspect, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the blade defect detection method of the first aspect.

[0026] The blade defect detection method, device, equipment, storage medium and computer program product of the embodiments of the present application receive the target blade cavity internal image sent by the inspection robot through the defect detection platform, and identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiments of the present application can improve the efficiency and accuracy of blade cavity internal defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0028] Figure 1 is a flow chart of a blade defect detection method provided by an embodiment of the present application;

[0029] Figure 2 is a schematic diagram of a defective sample block image provided by an embodiment of the present application;

[0030] Figure 3 is a flow chart of a blade defect detection method provided by another embodiment of the present application;

[0031] Figure 4 is a schematic diagram of the structure of an inspection robot provided by an embodiment of the present application;

[0032] Figure 5is a structural schematic diagram of a blade defect detection device provided by an embodiment of the present application;

[0033] Figure 6 is a structural schematic diagram of a blade defect detection device provided by another embodiment of the present application;

[0034] Figure 7 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0035] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0036] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or equipment including the elements.

[0037] As the difficulty and complexity of large-scale blade manufacturing increases, wind turbine blades will produce various types of defects such as delamination, glue deficiency, and glue resistance in the inner cavity of the blades during the production process, significantly increasing the risk of blade damage or even failure during wind turbine operation.

[0038] At present, defect identification inside the blade cavity mainly relies on manual visual inspection. The defect detection environment inside the cavity is harsh, and some areas are inaccessible to personnel. Generally, an endoscope cart is used to assist in manual inspection and defect identification, and management personnel are required to review the inspection results of the inspectors. This is highly dependent on the skills, experience and sense of responsibility of the personnel. The inspection work of quality inspectors is high-intensity, and the review workload of management personnel is large. The continuity and effectiveness of the inspection cannot be guaranteed, and the accuracy of the inspection results is not high.

[0039] The existing automatic defect detection methods for wind turbine blades are mainly for the detection of blade appearance, mostly with the help of mobile robots, drones and other equipment, focusing on machine equipment development and algorithm optimization for image stitching and positioning, and can only identify external damage defects such as cracks, lightning strikes, carbonization, etc. It is impossible to identify defects inside the blade cavity.

[0040] In order to solve the problems of the prior art, the embodiments of the present application provide a blade defect detection method, device, equipment, storage medium and computer program product. The target blade cavity internal image sent by the inspection robot is received through the defect detection platform, and the blade cavity defect feature information in the target blade cavity internal image is identified through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiments of the present application can improve the efficiency and accuracy of blade cavity internal defect detection.

[0041] The following first introduces the blade defect detection method provided in the embodiment of the present application.

[0042] Figure 1 FIG. 1 is a flow chart of a blade defect detection method provided by an embodiment of the present application. Figure 1 As shown, the method is applied to a defect detection platform, the defect detection platform may include a control panel; the defect detection platform is communicatively connected with the inspection robot. The method may include the following steps: S101 to S104.

[0043] S101: In response to a first input from a user on a control panel, a control instruction is generated.

[0044] The control panel is used to receive a first input from a user to generate a control instruction. The first input represents the user's control requirement for the inspection robot. The control instruction is used to control the inspection robot to perform an action corresponding to the control instruction.

[0045] S102: Sending a control instruction to an inspection robot located in the target blade cavity, so that the inspection robot collects images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity.

[0046] In some embodiments, sending a control instruction to the inspection robot located in the target blade cavity may be achieved through a first cable connecting the defect detection platform and the inspection robot.

[0047] The first cable can be connected to the control panel for communication. The first cable includes a video transmission line, a control full link line, and a power supply line. The video transmission line is used to transmit the internal image of the target blade cavity collected by the inspection robot to the defect detection platform in real time, and display it in real time through the control panel. The control full link line is used to transmit the control instructions received by the control panel to the inspection robot.

[0048] S103: Receive the internal image of the target blade cavity sent by the inspection robot.

[0049] In some embodiments, receiving the target blade cavity internal image sent by the inspection robot may be achieved through a second cable connected between the defect detection platform and the inspection robot.

[0050] The second cable may include a network transmission line, which is used to transmit the internal image of the target blade cavity collected by the inspection robot to the defect detection platform, so that the defect detection platform can perform defect detection based on the internal image of the target blade cavity.

[0051] S104: identifying blade cavity defect feature information in the target blade cavity internal image through a blade cavity defect recognition model to obtain a target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on blade cavity internal image samples and corresponding cavity defect label samples.

[0052] Among them, the blade cavity defect recognition model can be used to identify defects inside the blade cavity.

[0053] The target blade cavity defect recognition result indicates whether there are defects inside the target blade cavity and the defect information. The defect information may include the defect category and defect size. The defect category may include delamination, glue deficiency, glue resistance and other categories.

[0054] In some embodiments, a blade cavity defect recognition model is obtained based on training of blade cavity internal image samples and corresponding cavity defect label samples. Multiple preset target detection models can be trained using the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples until the training stop condition is met to obtain multiple trained target detection models. Then, the multiple trained target detection models are respectively detected using blade cavity internal image test samples to obtain detection results of the multiple trained target detection models, wherein the detection results at least include a defect detection rate. Finally, the target detection model with the highest defect detection rate is determined among the multiple trained target detection models.

[0055] In the embodiment of the present application, the target blade cavity internal image sent by the inspection robot is received through the defect detection platform, and the blade cavity defect feature information in the target blade cavity internal image is identified through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiment of the present application can improve the efficiency and accuracy of blade cavity internal defect detection.

[0056] In some embodiments, before S104, the blade defect detection method may further include:

[0057] Acquire initial blade cavity internal image samples corresponding to a plurality of defect categories respectively;

[0058] Perform sample expansion on the initial blade cavity internal image samples corresponding to each defect category to obtain expanded blade cavity internal image samples;

[0059] Based on the initial blade cavity internal image sample and the expanded blade cavity internal image sample, a blade cavity internal image sample is obtained;

[0060] Annotate the blade cavity internal image samples to obtain cavity defect label samples corresponding to the blade cavity internal image samples;

[0061] According to the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples, a preset target detection model is trained to obtain a blade cavity defect recognition model.

[0062] Due to the differences in blade models and manufacturing processes at each base, as well as the different locations and causes of defects, the same defect has different manifestations and frequencies, and the number of defects in each category is also different. The deep learning algorithm is data-driven, and training the blade cavity defect recognition model requires a large number of training samples.

[0063] To this end, by expanding the initial blade cavity internal image samples corresponding to each defect category based on the acquired initial blade cavity internal image samples, it is possible to ensure that the number of blade cavity internal image samples corresponding to each defect category is basically consistent, thereby ensuring that the trained blade cavity defect recognition model has high generalization ability and robustness.

[0064] As an example, obtaining initial blade cavity internal image samples corresponding to multiple defect categories can be achieved by, on the one hand, using an inspection robot to collect images inside the blade cavity. On the other hand, initial blade cavity internal image samples can also be obtained by making defective sample blocks. The defective sample block images can be as follows: Figure 2 The defect sample blocks are made by collecting defect sample block images of different sizes, different angles, and different lights, which can make up for the problem of insufficient number of image samples of a certain defect category or unclear image samples.

[0065] As an example, data enhancement technology may be used to expand the initial blade cavity internal image samples corresponding to each defect category to obtain expanded blade cavity internal image samples.

[0066] As an example, the labelimg labeling tool can be used to label the internal image samples of the blade cavity. After the labeling of the internal image samples of the blade cavity is completed, the labeled internal image samples of the blade cavity can be reviewed to ensure that there are no missed or mislabeled defects, and to ensure that the instance is consistent with the label content and the number of images is consistent with the number of labels.

[0067] In some embodiments, in order to efficiently implement data expansion of the initial blade cavity internal image samples to improve the robustness of the blade cavity defect recognition model, the above-mentioned sample expansion of the initial blade cavity internal image samples corresponding to each defect category to obtain the expanded blade cavity internal image samples may include:

[0068] The background image in the initial blade cavity internal image sample is randomly cropped, and the foreground image in the initial blade cavity internal image sample is rotated, translated, perspective transformed, light and dark changed, and mirrored to obtain an expanded blade cavity internal image sample.

[0069] The foreground image represents the image of defects in the blade cavity internal image sample, and the background image represents the image of the blade cavity internal image sample excluding the defects.

[0070] In some embodiments, the above-mentioned training of a preset target detection model based on the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples to obtain the blade cavity defect recognition model may include:

[0071] Using the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples, respectively training multiple preset target detection models until a training stop condition is met, thereby obtaining multiple trained target detection models;

[0072] Using the blade cavity internal image test samples to respectively detect the multiple trained target detection models, and obtain the detection results of the multiple trained target detection models; the detection results at least include the defect detection rate;

[0073] The target detection model with the highest defect detection rate is determined among multiple trained target detection models to obtain a blade cavity defect recognition model.

[0074] Among them, the preset target detection model can be a YOLO model, which can be used to detect the target object in the image and identify the position of the target object in the image, and has a higher detection speed than other models.

[0075] The multiple preset target detection models may be multiple different versions of YOLO models, such as YOLO v1 to YOLOv5.

[0076] The defect detection rate indicates that the defect recognition result obtained by inputting the internal image test sample of the blade cavity into the trained target detection model represents the probability of the existence of defects. When the defect detection rate is high, it can be considered that the accuracy of the target detection model is high; when the defect detection rate is low, it can be considered that the accuracy of the target detection model is low. Therefore, the blade cavity defect recognition model can be obtained by determining the target detection model with the highest defect detection rate among multiple trained target detection models.

[0077] By determining the target detection model with the highest defect detection rate among multiple trained target detection models, a blade cavity defect recognition model is obtained, which can improve the accuracy of the blade cavity defect recognition model.

[0078] As an example, the above-mentioned use of target inner cavity defect image samples to train multiple preset target detection models until the training stop condition is met to obtain multiple trained target detection models may include:

[0079] For each training sample, perform the following steps:

[0080] Inputting the internal image samples of the blade cavity into a plurality of preset target detection models respectively, and obtaining the blade cavity defect recognition results corresponding to the internal image samples of the blade cavity;

[0081] According to the blade cavity defect recognition results and the blade cavity defect label samples, respectively determine the loss function values ​​of multiple preset target detection models;

[0082] When multiple loss function values ​​do not meet the training stop conditions, the parameters of multiple preset target detection models are adjusted respectively to obtain multiple updated target detection models, and the blade cavity internal image samples are returned to input into the multiple preset target detection models respectively to obtain the blade cavity defect recognition results corresponding to the blade cavity internal image samples until the training stop conditions are met to obtain multiple trained target detection models.

[0083] In some embodiments, after identifying blade cavity defect feature information in the target blade cavity internal image by the blade cavity defect recognition model and obtaining the target blade cavity defect recognition result, the method further includes:

[0084] A blade cavity defect report of the target blade is generated according to the basic information of the target blade, the blade cavity defect recognition result corresponding to the target blade, the internal image of the target blade cavity, and the meter mark information corresponding to the internal image of the target blade cavity.

[0085] The basic information of the target blade may include the number and sequence number of the target blade and the system time when the target blade is detected.

[0086] The meter mark information indicates the progress position of the target blade cavity internal image in the inspection video taken by the inspection robot, and can be used to determine the actual position of the defect corresponding to the target blade cavity internal image inside the target blade cavity based on the progress position of the target blade cavity internal image in the inspection video.

[0087] The blade cavity defect report can be used for maintenance personnel to review, so as to determine the defect type, defect location, etc. of the target blade according to the blade cavity defect report, which can help maintenance personnel to quickly repair the target blade.

[0088] By generating a blade cavity defect report of the target blade based on the basic information of the target blade, the blade cavity defect identification result corresponding to the target blade, the internal image of the target blade cavity, and the meter mark information corresponding to the internal image of the target blade cavity, maintenance personnel can quickly repair the target blade, thereby improving production efficiency.

[0089] In some embodiments, after identifying blade cavity defect feature information in the target blade cavity internal image by the blade cavity defect recognition model and obtaining the target blade cavity defect recognition result, the method further includes:

[0090] When the target blade cavity defect recognition result indicates that the target blade cavity has defects, an alarm message is output.

[0091] The alarm information is used to remind the staff that there are defects in the target blade cavity, so that the staff can pay special attention to it or take corresponding measures.

[0092] By outputting an alarm message when the target blade cavity defect identification result indicates that the target blade cavity has defects, the staff can be reminded to pay attention to the defects inside the blade cavity in a timely manner, thereby avoiding the omission of defects due to negligence and reducing the manual burden.

[0093] As an example, when the target blade cavity defect recognition result indicates that there is a defect in the target blade cavity, an alarm message is output. This can be done by setting an alarm module on the defect detection platform. The alarm module can have a voice function. When the target blade cavity defect recognition result indicates that there is a defect in the target blade cavity, the alarm module can output a voice prompt of "Please note that there is a defect here" to achieve output of the alarm message.

[0094] Figure 3 FIG. 1 is a flow chart of a blade defect detection method provided by an embodiment of the present application. Figure 3 As shown in FIG. 1 , the method is applied to an inspection robot. The inspection robot can be Figure 4 As shown, it includes a crawler 401, a camera system 402, a rotating base 403, a driven wheel 404, an obstacle-crossing wheel 405, a driving wheel 406, and a cable interface 407. Among them, the crawler 401 is mainly used for forward, backward and steering functions, and has good maneuverability in a narrow space. The driven wheel 404, the obstacle-crossing wheel 405, and the driving wheel 406 are connected by gears to form a set of wheels. The two sets of wheels are fixedly and symmetrically arranged on both sides of the crawler 401 to ensure the flexibility of the inspection robot to move inside the blade cavity. The use of the inspection robot can greatly reduce the time that the inspectors watch the inspection screen for a long time. They only need to focus on controlling the inspection robot to crawl inside the blade cavity, which can reduce labor intensity.

[0095] The method may include the following steps: S301 to S303.

[0096] S301: receiving a control instruction sent by a control panel on a defect detection platform.

[0097] In some embodiments, receiving the control instruction sent by the control panel on the defect detection platform can be achieved through a first cable connecting the defect detection platform and the inspection robot.

[0098] The first cable can be connected to the control panel for communication. The first cable includes a control full link line. The control full link line is used to transmit the control instructions received by the control panel to the inspection robot.

[0099] S302: In response to the control instruction, an image of an internal area of ​​the target blade cavity is collected in the target blade cavity to obtain an internal image of the target blade cavity.

[0100] In some embodiments, in response to a control instruction, image capture is performed on the internal area of ​​the target blade cavity within the target blade cavity. The control system of the inspection robot may output a drive signal corresponding to the control instruction to the drive system, and the drive system adjusts the speed of the motor according to the drive signal so that the inspection robot moves according to the control instruction and performs image capture.

[0101] Specifically, the start and stop and speed regulation of the motor can be realized based on the driver program and control logic written by the single-chip microcomputer. Under the control of the program, the single-chip microcomputer regularly and continuously outputs a pulse wave (PWM) signal with a variable duty cycle to the drive circuit, thereby changing the motor speed to control the inspection robot to move forward or backward, and control the steering of the inspection robot by controlling the differential speed of the left and right wheels.

[0102] In some embodiments, the image acquisition of the internal area of ​​the target blade cavity in the target blade cavity can be completed by using a camera system provided on the inspection robot to shoot the inside of the target blade cavity. Specifically, the camera system can be used to shoot the inside of the target blade cavity to obtain an inspection video, and the internal image of the target blade cavity can be obtained by acquiring a video frame image in the inspection video.

[0103] S303: Send the target blade cavity internal image to the defect detection platform, so that the defect detection platform can identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0104] In some embodiments, sending the target blade cavity internal image to the defect detection platform may be achieved through a second cable connected between the defect detection platform and the inspection robot.

[0105] The second cable may include a network transmission line, which is used to transmit the internal image of the target blade cavity collected by the inspection robot to the defect detection platform, so that the defect detection platform can perform defect detection based on the internal image of the target blade cavity.

[0106] In the embodiment of the present application, the inspection robot sends the collected target blade cavity internal image to the defect detection platform, so that the defect detection platform recognizes the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiment of the present application can improve the efficiency and accuracy of blade cavity internal defect detection.

[0107] In some embodiments, the camera system 402 may include a camera, a rotating base 403 and a rotating motor; the rotating base is provided with a rotating motor, and the camera is arranged on the rotating base. S302 may include:

[0108] The rotation of the rotating base is controlled by a rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity within different angle ranges.

[0109] By setting a rotating motor on the rotating base and setting a camera on the rotating base, the rotating motor can be used to control the rotation of the rotating base to drive the camera to rotate, thereby realizing the shooting of the internal images of the target blade cavity in different angle ranges, ensuring the comprehensiveness of the internal images of the target blade cavity, and further improving the accuracy of identifying internal defects of the target blade cavity.

[0110] As an example, the rotation of the rotating base is controlled by a rotating motor to drive the camera to rotate. The rotation of the camera can be driven by converting the high speed and low torque output by the rotating motor into the low speed and high torque required by the rotating base.

[0111] As an example, the camera system 402 may include one camera or two cameras. The two cameras are respectively arranged at the front and rear of the inspection robot car, including a 4-megapixel front camera and a 2-megapixel rear camera. The 4-megapixel front camera is used to shoot the inspection video in real time; the 2-megapixel rear camera is used to observe the route of the car during its backward movement.

[0112] In some embodiments, in order to obtain a clearer and more complete defect image, the camera may be a smart camera; the rotating base is controlled to rotate by a rotating motor to drive the camera to rotate, so as to capture the internal image of the target blade cavity in different angle ranges, which may include:

[0113] When the intelligent camera determines that there is a cavity defect in the internal image of the target blade cavity, the rotating base is controlled to rotate by the rotating motor to drive the camera to rotate, so as to shoot the cavity defect at different angles.

[0114] Among them, the smart camera can have basic image recognition functions, and can determine whether there are cavity defects in the internal image of the target blade cavity.

[0115] When the intelligent camera determines that there is a cavity defect in the internal image of the target blade cavity, the rotating base is controlled by a rotating motor to drive the camera to rotate, and the cavity defect is photographed at different angles. This can obtain a clearer and more complete cavity defect image, and defect recognition is performed based on the defect image, which can further improve the accuracy of defect recognition.

[0116] In some embodiments, the camera has a positioning function; the rotating base is controlled to rotate by a rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity in different angle ranges, including:

[0117] Determine the position information of the inspection robot in the target blade cavity through the camera;

[0118] When the position information indicates that the inspection robot has reached a preset position, the rotating base is controlled to rotate by a rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity within different angle ranges.

[0119] The preset position may be set manually. As an example, the preset position may be a position within the cavity that is 10 meters away from the blade tail end in a straight line.

[0120] When the position information indicates that the inspection robot has reached the preset position, the rotating motor is used to control the rotation of the rotating base to drive the camera to rotate, so as to capture the internal images of the target blade cavity in different angle ranges. This enables the interior of the cavity at a specified position in the internal area of ​​the target blade cavity to be captured at different angles, thereby obtaining a more comprehensive and clearer image of the interior of the cavity.

[0121] Figure 5 Schematic diagram of the structure of a blade defect detection device 500 provided in an embodiment of the present application. Figure 5 As shown, the device is applied to a defect detection platform and may include a receiving module 510 , a sending module 520 , and an identification module 530 .

[0122] The receiving module 510 is used to generate a control instruction in response to a first input of a user on the control panel;

[0123] The sending module 520 is used to send a control instruction to the inspection robot located in the target blade cavity, so that the inspection robot can collect images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity;

[0124] A receiving module 510 is used to receive the internal image of the target blade cavity sent by the inspection robot;

[0125] The identification module 530 is used to identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0126] In the embodiment of the present application, the target blade cavity internal image sent by the inspection robot is received through the defect detection platform, and the blade cavity defect feature information in the target blade cavity internal image is identified through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiment of the present application can improve the efficiency and accuracy of blade cavity internal defect detection.

[0127] In some embodiments, the blade defect detection device 500 further includes an acquisition module, a processing module, and a training module;

[0128] An acquisition module, used for acquiring initial blade cavity internal image samples corresponding to a plurality of defect categories;

[0129] A processing module, used for performing sample expansion on the initial blade cavity internal image samples corresponding to each defect category to obtain expanded blade cavity internal image samples;

[0130] The processing module is further used to obtain the blade cavity internal image sample based on the initial blade cavity internal image sample and the expanded blade cavity internal image sample;

[0131] The processing module is also used to label the blade cavity internal image samples to obtain cavity defect label samples corresponding to the blade cavity internal image samples;

[0132] The training module is used to train a preset target detection model based on the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples to obtain a blade cavity defect recognition model.

[0133] In some embodiments, the processing module is also used to randomly crop the background image in the initial blade cavity internal image sample and rotate, translate, perspective transform, change light and dark, and mirror the foreground image in the initial blade cavity internal image sample to obtain an expanded blade cavity internal image sample.

[0134] In some embodiments, the blade defect detection device 500 further includes a detection module;

[0135] The training module is also used to respectively train a plurality of preset target detection models using the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples until a training stop condition is met, thereby obtaining a plurality of trained target detection models;

[0136] A detection module is used to detect multiple trained target detection models respectively using the internal image test samples of the blade cavity to obtain the detection results of the multiple trained target detection models; the detection results at least include the defect detection rate;

[0137] The detection module is also used to determine the target detection model with the highest defect detection rate among multiple trained target detection models to obtain a blade cavity defect recognition model.

[0138] In some embodiments, the blade defect detection device 500 further includes a generating module;

[0139] The generation module is used to generate a blade cavity defect report of the target blade according to the basic information of the target blade, the blade cavity defect recognition result corresponding to the target blade, the video frame image and the meter mark information corresponding to the video frame image.

[0140] In some embodiments, the generation module is also used to identify blade cavity defect feature information in the internal image of the target blade cavity through a blade cavity defect recognition model, and after obtaining the target blade cavity defect recognition result, if the target blade cavity defect recognition result indicates that there is a defect in the target blade cavity, output an alarm message.

[0141] Figure 5 The various modules in the device shown can be implemented Figure 1 Each step in the above steps can achieve the corresponding technical effect, which will not be repeated here for the sake of brevity.

[0142] like Figure 6 As shown, the device is applied to an inspection robot and may include a receiving module 610 , a sending module 620 , and a collection module 630 .

[0143] A receiving module 610 is used to receive a control instruction sent by a control panel on the defect detection platform;

[0144] The acquisition module 630 is used to acquire an image of the internal area of ​​the target blade cavity in response to the control instruction, so as to obtain an internal image of the target blade cavity;

[0145] The sending module 620 is used to send the target blade cavity internal image to the defect detection platform, so that the defect detection platform can identify the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

[0146] In the embodiment of the present application, the inspection robot sends the collected target blade cavity internal image to the defect detection platform, so that the defect detection platform recognizes the blade cavity defect feature information in the target blade cavity internal image through the blade cavity defect recognition model to obtain the target blade cavity defect recognition result. Since the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples, the blade cavity defect recognition model can quickly and accurately identify the defects inside the blade cavity based on the blade cavity internal image. Therefore, compared with the manual defect detection method, the embodiment of the present application can improve the efficiency and accuracy of blade cavity internal defect detection.

[0147] In some embodiments, the blade defect detection device 600 further includes a control module;

[0148] The control module is used to control the rotation of the rotating base through the rotating motor to drive the camera to rotate so as to capture the internal images of the target blade cavity within different angle ranges.

[0149] In some embodiments, the control module is also used to control the rotation of the rotating base through the rotating motor to drive the camera to rotate when the smart camera determines that there is a cavity defect in the internal image of the target blade cavity, so as to photograph the cavity defect at different angles.

[0150] In some embodiments, the camera has a positioning function; the rotating motor controls the rotating base to rotate, so as to drive the camera to rotate, so as to capture images of the interior of the target blade cavity at different angles, including:

[0151] Determine the position information of the inspection robot in the target blade cavity by using the camera;

[0152] When the position information indicates that the inspection robot has reached a preset position, the rotating base is controlled to rotate by a rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity within different angle ranges.

[0153] Figure 6 The various modules in the device shown can be implemented Figure 3 Each step in the above steps can achieve the corresponding technical effect, which will not be repeated here for the sake of brevity.

[0154] Figure 7 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0155] The electronic device may include a processor 701 and a memory 702 storing computer program instructions.

[0156] Specifically, the processor 701 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0157] The memory 702 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 702 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 702 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 702 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 702 is a non-volatile solid-state memory.

[0158] In certain embodiments, the memory may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, typically, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of the present application.

[0159] The processor 701 implements any one of the blade defect detection methods in the above embodiments by reading and executing computer program instructions stored in the memory 702 .

[0160] In one example, the electronic device may further include a communication interface 703 and a bus 710. Figure 7As shown, the processor 701, the memory 702, and the communication interface 703 are connected via a bus 710 and communicate with each other.

[0161] The communication interface 703 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0162] Bus 710 includes hardware, software or both, and the parts of electronic equipment are coupled to each other.For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 710 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0163] In addition, in combination with the blade defect detection method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any blade defect detection method in the above embodiment is implemented.

[0164] An embodiment of the present application also provides a computer program product, including a computer program, which implements any one of the blade defect detection methods in the above embodiments when the computer program is processed and executed.

[0165] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0166] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0167] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0168] The above reference is to the flowchart and / or block diagram of the method, device, equipment, storage medium and computer program product according to the embodiment of the present application. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. This processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of the boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0169] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A blade defect detection method, applied to a defect detection platform, characterized in that: The defect detection platform includes a control panel, and the defect detection platform is communicatively connected with the inspection robot; generating a control instruction in response to a first input of a user on the control panel; Sending the control instruction to the inspection robot located in the target blade cavity, so that the inspection robot collects images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity; Receiving the internal image of the target blade cavity sent by the inspection robot; The blade cavity defect recognition model is used to identify the blade cavity defect feature information in the target blade cavity internal image to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on the blade cavity internal image samples and the corresponding cavity defect label samples.

2. The method according to claim 1, characterized in that Before the blade cavity defect feature information in the target blade cavity internal image is identified by the blade cavity defect recognition model to obtain the target blade cavity defect recognition result, the method further includes: Acquire initial blade cavity internal image samples corresponding to a plurality of defect categories respectively; Perform sample expansion on the initial blade cavity internal image samples corresponding to each defect category to obtain expanded blade cavity internal image samples; Based on the initial blade cavity internal image sample and the expanded blade cavity internal image sample, obtaining a blade cavity internal image sample; Annotating the blade cavity internal image samples to obtain cavity defect label samples corresponding to the blade cavity internal image samples; A preset target detection model is trained based on the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples to obtain a blade cavity defect recognition model.

3. The method according to claim 2, characterized in that The step of performing sample expansion on the initial blade cavity internal image samples corresponding to each defect category to obtain expanded blade cavity internal image samples includes: The background image in the initial blade cavity internal image sample is randomly cropped, and the foreground image in the initial blade cavity internal image sample is rotated, translated, perspective transformed, light and dark changed, and mirrored to obtain an expanded blade cavity internal image sample.

4. The method according to claim 2, characterized in that: The method of training a preset target detection model based on the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples to obtain a blade cavity defect recognition model includes: Using the blade cavity internal image samples and the cavity defect label samples corresponding to the blade cavity internal image samples, respectively training a plurality of preset target detection models until a training stop condition is met, thereby obtaining a plurality of trained target detection models; Using the blade cavity internal image test samples to respectively detect the multiple trained target detection models to obtain the detection results of the multiple trained target detection models; the detection results at least include the defect detection rate; The target detection model with the highest defect detection rate is determined among the multiple trained target detection models to obtain a blade cavity defect recognition model.

5. The method according to any one of claims 1 to 4, characterized in that: After identifying the blade cavity defect feature information in the target blade cavity internal image by the blade cavity defect recognition model and obtaining the target blade cavity defect recognition result, the method further includes: A blade cavity defect report of the target blade is generated according to the basic information of the target blade, the blade cavity defect recognition result corresponding to the target blade, the internal image of the target blade cavity, and the meter mark information corresponding to the internal image of the target blade cavity.

6. The method according to any one of claims 1 to 4, characterized in that: After identifying the blade cavity defect feature information in the target blade cavity internal image by the blade cavity defect recognition model and obtaining the target blade cavity defect recognition result, the method further includes: When the target blade cavity defect identification result indicates that the target blade cavity has a defect, an alarm message is output.

7. A blade defect detection method, applied to an inspection robot, characterized in that: include: Receive control instructions sent by the control panel on the defect detection platform; In response to the control instruction, an image of an internal area of ​​the target blade cavity is collected in the target blade cavity to obtain an internal image of the target blade cavity; The target blade cavity internal image is sent to the defect detection platform so that the defect detection platform identifies the blade cavity defect feature information in the target blade cavity internal image through a blade cavity defect recognition model to obtain a target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on blade cavity internal image samples and corresponding cavity defect label samples.

8. The method according to claim 7, characterized in that The inspection robot includes a camera system, which includes a camera head, a rotating base and a rotating motor; the rotating base is provided with a rotating motor, and the camera head is arranged on the rotating base; The step of collecting an image of an internal area of ​​the target blade cavity in the target blade cavity to obtain an internal image of the target blade cavity includes: The rotating base is controlled to rotate by the rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity within different angle ranges.

9. The method according to claim 8, characterized in that The camera is a smart camera; the rotating motor controls the rotating base to rotate, so as to drive the camera to rotate, so as to capture the internal images of the target blade cavity in different angle ranges, including: When the smart camera determines that there is a cavity defect in the internal image of the target blade cavity, the rotating motor controls the rotating base to rotate, thereby driving the camera to rotate, so as to photograph the cavity defect at different angles.

10. The method according to claim 8, characterized in that The camera has a positioning function; the rotating motor controls the rotating base to rotate, so as to drive the camera to rotate, so as to capture the internal image of the target blade cavity in different angle ranges, including: Determining the position information of the inspection robot in the target blade cavity by using the camera; When the position information indicates that the inspection robot has reached a preset position, the rotating base is controlled to rotate by the rotating motor to drive the camera to rotate, so as to capture images of the interior of the target blade cavity within different angle ranges.

11. A blade defect detection device, applied to a defect detection platform, characterized in that: The defect detection platform includes a control panel, and the defect detection platform is communicatively connected with the inspection robot; the device includes: A receiving module, configured to generate a control instruction in response to a first input of a user on the control panel; A sending module, used for sending the control instruction to the inspection robot located in the target blade cavity, so that the inspection robot collects images of the internal area of ​​the target blade cavity to obtain an internal image of the target blade cavity; A receiving module, used for receiving the internal image of the target blade cavity sent by the inspection robot; The recognition module is used to identify the blade cavity defect feature information in the target blade cavity internal image through a blade cavity defect recognition model to obtain the target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on blade cavity internal image samples and corresponding cavity defect label samples.

12. A blade defect detection device, applied to an inspection robot, characterized in that: The device comprises: A receiving module, used for receiving control instructions sent by a control panel on the defect detection platform; A collection module, configured to collect images of an internal area of ​​the target blade cavity in response to the control instruction, and obtain an internal image of the target blade cavity; A sending module is used to send the target blade cavity internal image to the defect detection platform, so that the defect detection platform can identify the blade cavity defect feature information in the target blade cavity internal image through a blade cavity defect recognition model to obtain a target blade cavity defect recognition result; the blade cavity defect recognition model is trained based on blade cavity internal image samples and corresponding cavity defect label samples.

13. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the blade defect detection method according to any one of claims 1 to 10 is implemented.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the blade defect detection method according to any one of claims 1 to 10 is implemented.

15. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the blade defect detection method according to any one of claims 1 to 10.

Images