A Method and System for Obtaining the Angular Velocity of Offshore Wind Turbine Blades Based on Key Points

Abstract

The present application discloses a method and system for obtaining the angular velocity of an offshore wind turbine blade based on key points. The method includes: collecting offshore wind turbine blade images in real time and performing preprocessing; using a key point detection algorithm to respectively identify key points in the offshore wind turbine blade images at the first moment and the second moment collected; taking the nacelle center point identified in the offshore wind turbine blade images at the first moment and the second moment of any selected blade as the origin and the blade tip as the end point to generate a first vector and a second vector; calculating the angle difference between the first vector and the second vector, and calculating the offshore wind turbine blade angular velocity sequence by dividing the sum of the angle difference and the angle difference compensation value by the time interval between the first moment and the second moment; comparing the offshore wind turbine blade angular velocity obtained by the sensor with the offshore wind turbine blade angular velocity sequence to determine the finally output offshore wind turbine blade angular velocity. The present application can obtain the angular velocity of the offshore wind turbine blade with relatively high accuracy and strong real-time performance.

Description

Technical Field

[0001] This application relates to the field of new energy and wind power technology, and specifically relates to a method and system for obtaining the angular velocity of an offshore wind turbine blade based on key points. Background Technique

[0002] With the increasing emphasis on renewable energy, offshore wind power generation has become an important part of clean energy. During the operation and maintenance of wind power equipment, accurately identifying the angular velocity of the fan blades is crucial, which not only affects the operation efficiency of the fan but also the safety and service life of the wind power equipment.

[0003] Traditional methods for identifying the angular velocity of wind turbine blades are mainly based on physical sensors or image processing technology. The installation and maintenance costs of physical sensors are high, and they are easily affected by harsh marine environments, resulting in measurement accuracy errors. The technology based on image processing relies on high-quality image acquisition and complex image processing algorithms, and has high requirements for image acquisition equipment and processing platforms. Therefore, the existing technology has certain limitations and deficiencies.

[0004] Currently, the methods for obtaining the angular velocity of offshore wind turbine blades have problems such as low accuracy and weak real-time performance, and it is difficult to meet the requirements of the efficient and safe operation of modern wind power equipment. Therefore, it has become an urgent task to develop a method for obtaining the angular velocity of offshore wind turbine blades with higher accuracy and stronger real-time performance. Summary of the Invention

[0005] To solve the above problems, this application provides a method and system for obtaining the angular velocity of an offshore wind turbine blade based on key points.

[0006] In the first aspect, this application provides a method for obtaining the angular velocity of an offshore wind turbine blade based on key points, including:

[0007] Real-time collect images of offshore wind turbine blades and perform preprocessing;

[0008] Use a key point detection algorithm to respectively identify key points in the offshore wind turbine blade image at the first moment and the offshore wind turbine blade image at the second moment; the key points include: the center point of the nacelle, the tip of the blade, and the connection point between the nacelle and the blade; the time interval between the first moment and the second moment is less than the preset time interval; the yolov8-pose model is selected as the basic model for the key point detection algorithm;

[0009] Taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin and the blade tip of any blade identified in the offshore wind turbine blade image at the first moment as the end point, a first vector is generated; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin and the blade tip of the same blade identified in the offshore wind turbine blade image at the second moment as the end point, a second vector is generated; the same blade is determined by matching according to the key point matching algorithm.

[0010] Calculate the angular difference between the first vector and the second vector, and calculate the angular velocity sequence of the offshore wind turbine blade by dividing the sum of the angular difference and the angular difference compensation value by the time interval between the first moment and the second moment; the angular compensation value is n times 360 degrees, and n takes 0 and integers greater than 0; according to the sensor, obtain the angular velocity of the offshore wind turbine blade at the first moment or the second moment and compare it with the angular velocity sequence of the offshore wind turbine blade, and select the angular velocity of the offshore wind turbine blade closest to the angular velocity of the offshore wind turbine blade collected by the sensor from the angular velocity sequence of the offshore wind turbine blade as the finally output angular velocity of the offshore wind turbine blade.

[0011] By adopting the above scheme, the yolov8-pose model is used to more accurately detect the key points of the offshore wind turbine blade; based on the detected key points, the key points of each blade are matched, the angular velocity sequence is calculated according to the positional relationship of the identified key points, and combined with the angular velocity collected by the single technology of the sensor, a more accurate angular velocity is selected from the angular velocity sequence.

[0012] Preferably, it further includes:

[0013] Use the sensor to collect environmental data in real time; the environmental data includes: wind direction, wind speed and wind force.

[0014] Input the environmental data collected in the period from the first moment to the second moment into the first neural network model respectively, and output the result of whether the offshore wind turbine blade deforms in the period from the first moment to the second moment; the input of the first neural network model is environmental data, and the output is whether the offshore wind turbine blade deforms, which is trained and generated by historical environmental data and the result of whether the offshore wind blade deforms obtained at the corresponding moment.

[0015] If the offshore wind turbine blade deforms at any moment from the first moment to the second moment, the similarity threshold of the key points in the key point matching algorithm is correspondingly reduced, and the same blade at the first moment and the second moment is matched based on the adjusted key point matching algorithm.

[0016] By adopting the above solution, considering that the blades will undergo a certain deformation under specific environmental conditions, it is determined whether deformation occurs during the current period based on the neural network model. For the case of deformation, the similarity threshold for key point matching is adjusted correspondingly to ensure as many matching key points as possible and prevent the inability to obtain the key point matching result due to deformation problems.

[0017] Preferably, it further includes:

[0018] Taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin, and taking the blade tip of each remaining unselected blade identified in the offshore wind turbine blade image at the first moment as the end point, a corresponding first vector of the blade is generated; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin, and taking the blade tip of the same blade as each remaining unselected blade identified in the offshore wind turbine blade image at the second moment as the end point, a corresponding second vector of the blade is generated;

[0019] Compare the angular differences calculated from the first vector and the second vector of the corresponding blade obtained for each blade. If there is any angular difference between any two calculated values greater than the preset angular difference, increase the key point type and quantity to optimize the training of the key point detection algorithm. Re-identify the key points according to the optimized key point detection algorithm, and re-determine the same blade based on the re-identified key points.

[0020] By adopting the above solution, considering that incorrect key point matching will lead to incorrect angular difference calculation, calculate the angular difference corresponding to the angular velocity of the remaining blades to verify whether there is an error in the angular difference obtained by calculating the corresponding selected blades, and then re-match the key points according to the verification result to ensure an accurate angular velocity.

[0021] Preferably, it further includes:

[0022] When the angular velocity of the offshore wind turbine blade cannot be obtained in real time according to the sensor, the environmental data collected at the first moment or the second moment and the historical blade speed at a certain time interval from the first moment or the second moment are input into the second neural network model, and the predicted offshore wind turbine blade speed at the first moment or the second moment is output; the input of the second neural network model is the environmental data at the current moment and the historical blade speed at a certain time interval from the current moment, and the output is the offshore wind turbine blade speed at the current moment, which is trained and generated by the historical offshore wind turbine environmental data, the blade speed at a certain time interval from the historical offshore wind turbine environmental data collection moment, and the speed at the historical offshore wind turbine environmental data collection moment.

[0023] Calculate the angular velocity of the offshore wind turbine blade based on the predicted offshore wind turbine blade speed, and compare the calculated angular velocity of the offshore wind turbine blade with the angular velocity sequence of the offshore wind turbine blade.

[0024] By adopting the above scheme, considering that when the angular velocity cannot be obtained according to the sensor, the rotational speed is predicted and generated based on historical environmental data and blade rotational speed, and then the angular velocity of the offshore wind turbine blade is obtained to assist in determining the angular velocity of the offshore wind turbine blade in the angular velocity sequence of the offshore wind turbine blade.

[0025] Preferably, it further includes:

[0026] Judge whether the finally output angular velocity of the offshore wind turbine blade is greater than the preset angular velocity threshold of the offshore wind turbine blade; if it is greater, use the angular velocity of the offshore wind turbine blade at the first moment or the second moment collected by the backup sensor to replace the angular velocity of the offshore wind turbine blade collected by the original sensor to compare the angular velocity sequence of the offshore wind turbine blade.

[0027] By adopting the above scheme, considering that when the offshore wind force and wind speed reach a certain value, the offshore wind turbine blade control system will control the offshore wind turbine blade to operate at a certain angular velocity to avoid blade damage, so there is a corresponding angular velocity threshold. Once the angular velocity obtained by the sensor is greater than the angular velocity threshold, it indicates that the current sensor is damaged, and the value collected by the backup sensor is correspondingly used for auxiliary judgment.

[0028] Preferably, it further includes:

[0029] Obtain the finally output angular velocity sequence of the offshore wind turbine blade, monitor the moment when the absolute value of the angular velocity change value in the angular velocity sequence of the offshore wind turbine blade is greater than the preset angular velocity change threshold, record it as the distortion moment, and correspondingly obtain the environmental data at the distortion moment and analyze it;

[0030] If the number of types of environmental data that meet the first condition is greater than the preset number, use the angular velocity of the offshore wind turbine blade at the distortion moment collected by the backup sensor to replace the angular velocity of the offshore wind turbine blade collected by the original sensor to compare the angular velocity sequence of the offshore wind turbine blade; the first condition is: the absolute value of the change value of a certain type of environmental data at the distortion moment compared with the corresponding type of environmental data at the previous moment of the distortion moment is greater than the corresponding preset environmental data change threshold.

[0031] By adopting the above scheme, considering that the angular velocity change of the offshore wind turbine blade mainly depends on the changes in wind speed and wind force in the environmental data. Once it is calculated that the moment with a large angular velocity change corresponds to a huge change in environmental data, it is verified that there is a calculation error in the current angular velocity, indicating that the current sensor is damaged, and the value collected by the backup sensor is correspondingly used for auxiliary judgment.

[0032] Preferably, add Pose as CompositionalTokens to the selected yolov8-pose model network architecture and use DCNv3 to replace the convolutional layer.

[0033] By adopting the above solution, preferably using Pose as Compositional Tokens and DCNv3 to improve the accuracy of the model. Due to the introduction of these technologies, the model has higher performance and stability in key point detection, further improving the accuracy of rotational speed recognition.

[0034] In a second aspect, the present application provides a key point-based offshore wind turbine blade angular velocity system, including:

[0035] An image acquisition module for real-time acquisition of offshore wind turbine blade images and preprocessing;

[0036] A key point detection module for using a key point detection algorithm to respectively identify key points in the offshore wind turbine blade image at the first moment and the offshore wind turbine blade image at the second moment collected; the key points include: the center point of the nacelle, the tip of the blade, and the connection point between the nacelle and the blade; the key point detection algorithm selects the yolov8-pose model;

[0037] A vector acquisition module for generating a first vector with the center point of the nacelle identified in the offshore wind turbine blade image at the first moment as the origin and the tip of any blade identified in the offshore wind turbine blade image at the first moment as the end point; generating a second vector with the center point of the nacelle identified in the offshore wind turbine blade image at the second moment as the origin and the tip of the same blade identified in the offshore wind turbine blade image at the second moment as the end point; the same blade is determined by matching according to the key point matching algorithm;

[0038] An angular velocity calculation module for calculating the angle difference between the first vector and the second vector, and calculating the offshore wind turbine blade angular velocity sequence by dividing the sum of the angle difference and the angle difference compensation value by the time interval between the first moment and the second moment; the angle compensation value is n times 360 degrees, where n takes an integer of 0 and greater than 0; comparing the offshore wind turbine blade angular velocity obtained from the sensor with the offshore wind turbine blade angular velocity sequence, and selecting the offshore wind turbine blade angular velocity closest to the offshore wind turbine blade angular velocity collected by the sensor from the offshore wind turbine blade angular velocity sequence.

[0039] By adopting the above solution, the key points of the offshore wind turbine blade are detected by the key point detection algorithm, the angle between the blades within a certain time interval is calculated based on these key points, and then the angular velocity of the blade is calculated based on the change of the angle, improving the accuracy and efficiency of angular velocity recognition.

[0040] In a third aspect, the present application provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method as described above.

[0041] Fourthly, the present application provides a computer device, which includes a memory, a processor, and a program stored on the memory and operable. When the program is executed by the processor, the steps of the above method are implemented.

[0042] In summary, the present application has the following beneficial effects:

[0043] 1. By integrating image detection technology and sensor detection technology, the yolov8-pose model is selected to more accurately detect the key points of the offshore wind turbine blade; according to the positional relationship of the identified and matched key points, an angular velocity sequence is calculated, and combined with the angular velocity collected by the single sensor technology, a more accurate angular velocity is selected from the angular velocity sequence to improve the accuracy of angular velocity recognition; 2. Combining the collected environmental data, the judgment and optimization of key point matching are carried out to further improve the accuracy of the finally calculated angular velocity;

[0044] 3. The calculated angle differences of each blade are used to verify each other, and according to the verification results, the key points are further matched and optimized to improve the accuracy of the finally calculated angular velocity. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flowchart of the method for obtaining the angular velocity of an offshore wind turbine blade based on key points in a specific embodiment;

[0046] Figure 2 is an illustration of the identification of key points of an offshore wind turbine blade using the method for obtaining the angular velocity of an offshore wind turbine blade based on key points in a specific embodiment;

[0047] Figure 3 is an illustration of the positional relationship of key points of an offshore wind turbine blade using the method for obtaining the angular velocity of an offshore wind turbine blade based on key points in a specific embodiment;

[0048] Figure 4 is a schematic structural diagram of the system for obtaining the angular velocity of an offshore wind turbine blade based on key points in a specific embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] As Figure 1 shown, an embodiment of the present application discloses a method for obtaining the angular velocity of an offshore wind turbine blade based on key points, which specifically includes:

[0051] S1. Real-time collect the images of the offshore wind turbine blade and perform preprocessing.

[0052] Specifically, a drone is used to take real-time pictures of the offshore wind turbine blades within a preset shooting distance. In this embodiment, the preset shooting distance is set within 80 meters from the center of the diameter.

[0053] Preprocess the real-time offshore wind turbine blade images collected by shooting, including adjusting the brightness and contrast of the images, and performing enhancement processing on the images.

[0054] S2. Use a key point detection algorithm to identify key points in the preprocessed image data.

[0055] Specifically, the key point detection algorithm uses the yolov8-pose model as the basic model or selects other neural network models for identifying key points; the key points include: the center point of the nacelle, the tip of the blade, the maximum bending point of the blade, and the connection points between the nacelle and the blade, etc. Among them, the connection points between each nacelle and the blade are symmetrically arranged. For example, the connection points between the three blades and the nacelle are symmetrically arranged at intervals of 120° clockwise. As Figure 2 shown, the tip of the blade and the connection point between the nacelle and the blade are associated and set at a distance equal to the length of the blade.

[0056] Use the yolov8-pose model trained and constructed by using historical offshore wind turbine blade images with key point information (position, type, association relationship, etc.) manually marked at different angles to set the loss function, adjust the parameters of the yolov8-pose model, and perform iterative training to generate a yolov8-pose model with a loss function value greater than the preset loss threshold; in addition, add Pose as Compositional Tokens to the selected yolov8-pose model network architecture and use DCNv3 to replace the convolutional layer; among them, encode the key points in historical offshore wind images with different angles and forms to further improve the accuracy of key detection.

[0057] Input the preprocessed image data of each frame into the trained yolov8-pose model, and select the key point recognition results of the offshore wind turbine blade image at the first moment and the key point recognition results of the offshore wind turbine blade image at the second moment. The second moment is the next moment after the first moment, and the time interval between the first moment and the second moment is less than the preset time interval. This preset time interval is relatively small. In this embodiment, it is taken as 1s. As Figure 3 shown, use the key point detection algorithm to obtain the key point information of a frame of image.

[0058] S3. Obtain the angle difference of the blades in different frame image data according to the identified key points.

[0059] Specifically, taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin, and taking the blade tip of any blade identified in the offshore wind turbine blade image at the first moment as the end point, a first vector is generated; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin, and taking the blade tip of the same blade identified in the offshore wind turbine blade image at the second moment as the end point, a second vector is generated; and the angular difference between the first vector and the second vector is calculated.

[0060] In this embodiment, the offshore wind turbine blade has three blades; according to the key point recognition result in the offshore wind turbine blade image at the first moment, a first vector corresponding to a randomly selected blade (such as the first blade) is selected, and a second vector corresponding to the first blade is determined in the key point recognition result of the offshore wind turbine blade at the second moment, and the included angle between the same blade at two moments is calculated. The specific calculation formula is:

[0061]

[0062] In the formula, θ is the included angle between the same blade at two moments, that is, the angular difference between the same blade at two moments, is the first vector, is the second vector.

[0063] Among them, the same blade can be selected and determined by matching according to the key point matching algorithm; the key point matching algorithm is to match the feature information of the key points identified in different frame image data. The feature information of the key points includes: position, texture, color, mark, shape, etc.; in the matching process, in addition to comparing the similarity of the feature information between the key points, the key point matching is also based on the type of each key point and the distance relationship between the key points.

[0064] In addition, considering that the feature information between the three blades in the offshore wind turbine blade is relatively similar, when training the key point detection algorithm, the type of training data can be increased, that is, the number and type of key points. For example, a key point is determined at a certain distance interval for each blade; or the key point matching recognition is combined with the vibration feature signal. Vibration sensors are installed at each key point to obtain the feature information of the vibration signals corresponding to different key points, and the vibration feature information is combined with other feature information of the key points for comprehensive matching.

[0065] In this embodiment, it is selected in the scenario where the offshore wind turbine blade rotates at a low speed, that is, in the scenario where the rotational speed of the offshore wind turbine blade is lower than the preset rotational speed according to the sensors installed on the offshore wind turbine blade. Among them, at the preset rotational speed, the rotation angle of the blade within the time interval between the first moment and the second moment is less than the angle between two blades, that is, the rotation angle of the blade within 1 s is less than 120 degrees. In this scenario, the blade closest to the blade tip position selected in the offshore wind turbine blade image at the first moment among the blade tip positions of each blade recognized in the offshore wind turbine blade image at the second moment can be directly used as the same blade as in the offshore wind turbine blade image at the first moment.

[0066] S4. Calculate the angular velocity of the offshore wind turbine blade according to the calculated angle difference between the first vector and the second vector.

[0067] Specifically, calculate the angular velocity sequence of the offshore wind turbine blade by dividing the sum of the angle difference and the angle difference compensation value by the time interval between the first moment and the second moment; the specific formula is:

[0068] {θ} = θ + θ0

[0069] θ0 = n * 360°

[0070] In the formula, {θ} is the angular velocity sequence of the offshore wind turbine blade, θ0 is the angle difference compensation value, and n takes integers of 0 and greater than 0.

[0071] Since the time interval from the first moment to the second moment is short, it is defaulted that the angular velocity values at the two moments are not very different. Therefore, compare the angular velocity of the offshore wind turbine blade at the first moment or the second moment obtained by the sensor with the angular velocity sequence of the offshore wind turbine blade to assist in determining the final angular velocity of the offshore wind turbine blade; that is, select the angular velocity of the offshore wind turbine blade closest to the angular velocity of the offshore wind turbine blade collected by the sensor from the angular velocity sequence of the offshore wind turbine blade as the final output angular velocity of the offshore wind turbine blade; among them, the angular velocity of the offshore wind turbine blade at a certain moment obtained by the sensor can be the average value of the angular velocities obtained by the sensors installed on multiple blades.

[0072] Or directly compare the average value of the angular velocities of the offshore wind turbine blade at the first moment and the second moment with the angular velocity sequence of the offshore wind turbine blade to assist in determining the final angular velocity of the offshore wind turbine blade; that is, select the angular velocity of the offshore wind turbine blade closest to the average value of the angular velocities of the offshore wind turbine blade collected by the sensor from the angular velocity sequence of the offshore wind turbine blade as the final output angular velocity of the offshore wind turbine blade.

[0073] In this embodiment, it is selected in the scenario where the offshore wind turbine blade rotates at a low speed. Since it is impossible for the same blade to rotate one or more full circles, n takes 0, and directly calculate the angular velocity of the offshore wind turbine blade by dividing the angle difference between the first vector and the second vector by the time interval between the first moment and the second moment.

[0074] By adopting the solution described in this embodiment, the correct rate of obtaining the angular velocity of the target offshore wind turbine blade at the corresponding moment is relatively high, and the angular velocity can be obtained in real time.

[0075] In a specific embodiment, considering that specific environmental conditions, such as strong winds, etc., may cause the shape of the offshore wind turbine blade to deform, to avoid reducing the accuracy of angular velocity acquisition due to the influence of deformation on key point matching, the method further includes: using sensors to collect environmental data in real time; the environmental data includes: wind direction, wind speed, wind force, etc.

[0076] Input the environmental data collected in the period from the first moment to the second moment into the first neural network model respectively, and output the result of whether the offshore wind turbine blade deforms in the period from the first moment to the second moment; the input of the first neural network model is environmental data, and the output is whether the offshore wind turbine blade deforms; the first neural network is trained and generated through historical environmental data and the result of whether the offshore wind turbine blade deforms obtained at the corresponding moment.

[0077] If the offshore wind turbine blade deforms at any moment from the first moment to the second moment, the similarity threshold of the key points in the key point matching algorithm is correspondingly reduced, and the same blade is matched between the first moment and the second moment based on the adjusted key point matching algorithm; in this embodiment, the reduced value of the similarity threshold of the key points is 5%, for example: from 95% to 90%.

[0078] In addition, during the key point matching process, if there is more than one blade in the offshore wind turbine blade image at the second moment that matches the first blade identified in the offshore wind turbine blade image at the first moment, such as 3 blades, then determine multiple (3 kinds) blade matching results for the association relationship between each matching result and the key points, that is, the first blade, the second blade, and the third blade identified in the offshore wind turbine blade image at the first moment, retain multiple matching results, calculate the angle differences corresponding to different matching results, calculate the angular velocities of multiple offshore wind turbine blades subsequently, and compare with the angular velocity of the offshore wind turbine blade collected by the sensor to finally determine the angular velocity of the offshore wind turbine blade.

[0079] In a specific embodiment, the accuracy of the angular velocity acquisition result is verified by calculating multiple angle differences, and then the key point detection algorithm is optimized to improve the accuracy of key point detection; the method includes:

[0080] Taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin, and taking the blade tip of each remaining unselected blade identified in the offshore wind turbine blade image at the first moment as the end point, a first vector corresponding to the blade is generated; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin, and taking the blade tip of the same blade as each remaining unselected blade identified in the offshore wind turbine blade image at the second moment as the end point, a second vector corresponding to the blade is generated;

[0081] For example, according to the key points identified in the offshore wind turbine blade images of the second blade at the first moment and the second moment, a first vector and a second vector of the second blade are generated; according to the key points identified in the offshore wind turbine blade images of the third blade at the first moment and the second moment, a first vector and a second vector of the third blade are generated.

[0082] Compare the angular differences calculated from the first vector and the second vector of the corresponding blade obtained for each blade; for example, the angular difference calculated for the first blade is 30 degrees, the angular difference calculated for the second blade is 28 degrees, and the angular difference calculated for the third blade is 30 degrees;

[0083] Ignoring the accuracy of the matching algorithm, if there is any two calculated angular differences greater than the preset angular difference, increase the type and quantity of key points, that is, optimize the training key point detection algorithm by increasing the training data; in this embodiment, the preset angular difference is 1 degree, then the angular difference calculated for the first blade corresponding to the second blade is 2 degrees greater than 1 degree, and the key point detection algorithm needs to be further optimized for training.

[0084] Re - identify the key points according to the optimized key point detection algorithm, and re - perform key point matching according to the re - identified key points to determine the same blade.

[0085] In a specific embodiment, considering the scenario where the sensor is damaged and the angular velocity of the offshore wind turbine blade cannot be obtained by the sensor to assist in determining the final angular velocity, a method of predicting the rotational speed and calculating the angular velocity is adopted to assist in generating the angular velocity; the method includes:

[0086] Considering that the offshore wind turbine blade control system will adjust the rotational speed of the offshore wind turbine blade according to environmental data, when the angular velocity of the offshore wind turbine blade cannot be obtained in real time by the sensor, the environmental data collected at the first moment or the second moment and the historical blade rotational speed at a time interval from the first moment or the second moment are input into the second neural network model, and the predicted rotational speed of the offshore wind turbine blade at the first moment or the second moment is output.

[0087] The input of the second neural network model is the environmental data at any moment and the historical blade rotation speed at an interval of a certain period of time from any moment, and the output is the rotation speed of the offshore wind turbine blade at the current moment; the second neural network is trained and generated by the historical offshore wind turbine blade environmental data, the blade rotation speed at an interval of a certain period of time from the collection moment of the historical offshore wind turbine blade environmental data, and the rotation speed at the collection moment of the historical offshore wind turbine blade environmental data; among them, the preset period of time is set to 2 hours.

[0088] The angular velocity of the offshore wind turbine blade is calculated based on the predicted rotation speed of the offshore wind turbine blade, and the calculated angular velocity of the offshore wind turbine blade is compared with the angular velocity sequence of the offshore wind turbine blade. Among them, the calculation formula between the rotation speed of the offshore wind turbine blade and the angular velocity of the offshore wind turbine blade is w = 2πn / 60.

[0089] In a specific embodiment, considering the scenario where the sensor is damaged, based on the obtained angular velocity, the result detected by the sensor is verified in reverse to avoid reducing the accuracy rate of the finally obtained angular velocity; the method includes:

[0090] Considering that the offshore wind turbine blade control system will set a maximum offshore wind turbine blade rotation speed during the process of adjusting the rotation speed of the offshore wind turbine blade according to the environmental data, a corresponding angular velocity threshold of the offshore wind turbine blade is set. Thus, it is judged whether the finally output angular velocity of the offshore wind turbine blade is greater than the preset angular velocity threshold of the offshore wind turbine blade; if it is greater, the angular velocity of the offshore wind turbine blade at the first moment or the second moment collected by the backup sensor is used to replace the angular velocity of the offshore wind turbine blade collected by the original sensor to compare with the angular velocity sequence of the offshore wind turbine blade.

[0091] Considering that the offshore wind turbine blade control system will complete the adjustment at a certain rotation speed change rate during the process of adjusting the rotation speed of the offshore wind turbine blade according to the environmental data to avoid damaging the offshore wind turbine blade due to rapid adjustment. Thus, the angular velocity sequence of the finally output offshore wind turbine blade is obtained, and the moment when the absolute value of the angular velocity change value in the angular velocity sequence of the offshore wind turbine blade is greater than the preset angular velocity change threshold is monitored and recorded as the distortion moment; for example: the angular velocity change value at the first moment is the angular velocity at the first moment minus the initial angular velocity; the angular velocity change value at the second moment is the angular velocity at the second moment minus the angular velocity at the first moment; if the absolute value of the angular velocity change value at the second moment is greater than the preset angular change threshold, the second moment is recorded as the distortion moment.

[0092] Collect the environmental data corresponding to the distortion moment and analyze it; if the number of types of environmental data that meet the first condition is greater than the preset number, use the angular velocity of the offshore wind turbine blade at the distortion moment collected by the backup sensor to replace the angular velocity of the offshore wind turbine blade collected by the original sensor and compare the angular velocity sequence of the offshore wind turbine blade; the first condition is that the absolute value of the change value of a certain type of environmental data at the distortion moment compared to the corresponding type of environmental data at the previous moment of the distortion moment is greater than the corresponding preset environmental data change threshold.

[0093] For example, the wind force data corresponding to the second moment is X1, the wind force data corresponding to the first moment is X2, the change value of the wind force data at the first moment compared to the wind force data at the second moment is X2 - X1, and the absolute value of this change value is greater than the corresponding wind force data change threshold.

[0094] As Figure 4 shown, an embodiment of the present application discloses a system for obtaining the angular velocity of an offshore wind turbine blade based on key points, including:

[0095] An image acquisition module 101 for real-time collecting the images of the offshore wind turbine blade and performing preprocessing;

[0096] A key point detection module 102 for respectively identifying key points in the image of the offshore wind turbine blade at the first moment and the image of the offshore wind turbine blade at the second moment collected by using a key point detection algorithm; the key points include: the center point of the nacelle, the tip of the blade, and the connection point between the nacelle and the blade; the key point detection algorithm selects the yolov8-pose model;

[0097] A vector acquisition module 103 for generating a first vector with the center point of the nacelle identified in the image of the offshore wind turbine blade at the first moment as the origin and the tip of any blade identified in the image of the offshore wind turbine blade at the first moment as the end point; generating a second vector with the center point of the nacelle identified in the image of the offshore wind turbine blade at the second moment as the origin and the tip of the same blade identified in the image of the offshore wind turbine blade at the second moment as the end point; the same blade is determined by matching according to the key point matching algorithm; an angular velocity calculation module 104 for calculating the angle difference between the first vector and the second vector, and calculating the angular velocity sequence of the offshore wind turbine blade by dividing the sum of the angle difference and the angle difference compensation value by the time interval between the first moment and the second moment; the angle compensation value is n times 360 degrees, and n takes an integer of 0 and greater than 0; compare the angular velocity sequence of the offshore wind turbine blade according to the angular velocity of the offshore wind turbine blade obtained by the sensor, and select the angular velocity of the offshore wind turbine blade closest to the angular velocity of the offshore wind turbine blade collected by the sensor from the angular velocity sequence of the offshore wind turbine blade.

[0098] The system further includes:

[0099] The key point detection optimization module 105 is used to collect environmental data in real time by using sensors; the environmental data includes: wind direction, wind speed, and wind force; the environmental data collected in the period from the first moment to the second moment is respectively input into the first neural network model, and the result of whether the offshore wind turbine blade deforms in the period from the first moment to the second moment is output; the input of the first neural network model is environmental data, and the output is whether the offshore wind turbine blade deforms, through the historical environmental data and the result of whether the offshore blade deforms obtained at the corresponding moment; if the offshore wind turbine blade deforms at any moment from the first moment to the second moment, the similarity threshold of the key points in the key point matching algorithm is correspondingly reduced, and the same blade at the first moment and the second moment is matched based on the adjusted key point matching algorithm.

[0100] The key point detection optimization module 105 is further used to generate a first vector corresponding to each blade with the center point of the nacelle identified in the offshore wind turbine blade image at the first moment as the origin and the blade tip of each remaining unselected blade identified in the offshore wind turbine blade image at the first moment as the end point; generate a second vector corresponding to each blade with the center point of the nacelle identified in the offshore wind turbine blade image at the second moment as the origin and the blade tip of the same blade as each remaining unselected blade identified in the offshore wind turbine blade image at the second moment as the end point; compare the angular differences calculated from the first vector and the second vector of the corresponding blade obtained for each blade, and if there is any angular difference between any two calculated values greater than the preset angular difference, increase the key point type and the number to optimize the training of the key point detection algorithm, re-identify the key points according to the optimized key point detection algorithm, and re-determine the same blade according to the re-identified key points.

[0101] The angular velocity calculation module 104 is further used to, when the angular velocity of the offshore wind turbine blade cannot be obtained in real time by the sensor, input the environmental data collected at the first moment or the second moment and the historical blade rotation speed at a certain time interval from the first moment or the second moment into the second neural network model, and output the predicted rotation speed of the offshore wind turbine blade at the first moment or the second moment; the input of the second neural network model is the environmental data at the current moment and the historical blade rotation speed at a certain time interval from the current moment, and the output is the rotation speed of the offshore wind turbine blade at the current moment, which is generated by training with the historical offshore wind turbine blade environmental data, the blade rotation speed at a certain time interval from the historical offshore wind turbine blade environmental data collection moment, and the rotation speed at the historical offshore wind turbine blade environmental data collection moment; calculate the angular velocity of the offshore wind turbine blade based on the predicted rotation speed of the offshore wind turbine blade, and compare the calculated angular velocity of the offshore wind turbine blade with the angular velocity sequence of the offshore wind turbine blade.

[0102] The angular velocity calculation module 104 is further configured to determine whether the finally output angular velocity of the offshore wind turbine blade is greater than a preset angular velocity threshold of the offshore wind turbine blade; if it is greater, the angular velocity of the offshore wind turbine blade at the first moment or the second moment collected by the backup sensor is used to replace the angular velocity of the offshore wind turbine blade collected by the original sensor to compare the angular velocity sequence of the offshore wind turbine blade.

[0103] The angular velocity calculation module 104 obtains the finally output angular velocity sequence of the offshore wind turbine blade, monitors the moment when the absolute value of the angular velocity change value in the angular velocity sequence of the offshore wind turbine blade is greater than a preset angular velocity change threshold, and records it as a distortion moment, and correspondingly obtains the environmental data at the distortion moment and analyzes it; if the number of types of environmental data that meet the first condition is greater than a preset number, the angular velocity of the offshore wind turbine blade at the distortion moment collected by the backup sensor is used to replace the angular velocity of the offshore wind turbine blade collected by the original sensor to compare the angular velocity sequence of the offshore wind turbine blade; the first condition is that the absolute value of the change value of a certain type of environmental data at the distortion moment compared with the corresponding type of environmental data at the previous moment of the distortion moment is greater than the corresponding preset environmental data change threshold.

[0104] An embodiment of the present application also discloses a computer-readable storage medium.

[0105] Specifically, the computer-readable storage medium stores a computer program that can be loaded and executed by a processor such as the above method for obtaining the angular velocity of an offshore wind turbine blade based on key points. The computer-readable storage medium includes, for example: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0106] An embodiment of the present application also discloses a computer device.

[0107] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded and executed by the processor such as the above method for obtaining the angular velocity of an offshore wind turbine blade based on key points.

[0108] The above are all preferred embodiments of the present application. The protection scope of the present application is not limited by this. Any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example in a series of equivalent or similar features.

Claims

1. A method for obtaining the angular velocity of an offshore wind turbine blade based on key points, characterized in that Including: Collecting real-time images of the offshore wind turbine blade and performing preprocessing; Using a key-point detection algorithm to respectively identify key points in the offshore wind turbine blade image at the first moment and the offshore wind turbine blade image at the second moment; the key points include: the nacelle center point, the blade tip, and the connection point between the nacelle and the blade; the time interval between the first moment and the second moment is less than the preset time interval; the yolov8-pose model is selected as the key-point detection algorithm; Taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin, and taking the blade tip of any blade identified in the offshore wind turbine blade image at the first moment as the end point, to generate a first vector; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin, and taking the blade tip of the same blade identified in the offshore wind turbine blade image at the second moment as the end point, to generate a second vector; the same blade is determined by matching according to the key-point matching algorithm; Calculating the angular difference between the first vector and the second vector, and using the sum of the angular difference and the angular difference compensation value Dividing by the time interval between the first moment and the second moment to calculate and obtain the angular velocity sequence of the offshore wind turbine blade; the angular difference compensation value is n times of 360 degrees, and n takes 0 and integers greater than 0; comparing the angular velocity of the offshore wind turbine blade at the first moment or the second moment obtained by the sensor with the angular velocity sequence of the offshore wind turbine blade, and selecting the angular velocity of the offshore wind turbine blade closest to the angular velocity of the offshore wind turbine blade collected by the sensor from the angular velocity sequence of the offshore wind turbine blade as the finally output angular velocity of the offshore wind turbine blade; Also including: Using the sensor to collect environmental data in real time; the environmental data includes: wind direction, wind speed, and wind force; Inputting the environmental data collected in the period from the first moment to the second moment into the first neural network model respectively, and outputting the result of whether the offshore wind turbine blade deforms in the period from the first moment to the second moment; the input of the first neural network model is environmental data, and the output is whether the offshore wind turbine blade deforms, which is trained and generated by historical environmental data and the result of whether the offshore blade deforms obtained at the corresponding moment; If the offshore wind turbine blade deforms at any moment from the first moment to the second moment, then correspondingly reduce the similarity threshold of the key points in the key-point matching algorithm, and perform matching of the same blade at the first moment and the second moment based on the adjusted key-point matching algorithm; Also including: Taking the nacelle center point identified in the offshore wind turbine blade image at the first moment as the origin, and taking the blade tip of each remaining unselected blade identified in the offshore wind turbine blade image at the first moment as the end point, to generate the corresponding blade first vector; taking the nacelle center point identified in the offshore wind turbine blade image at the second moment as the origin, and taking the blade tip of the same blade as each remaining unselected blade identified in the offshore wind turbine blade image at the second moment as the end point, to generate the corresponding blade second vector; Compare the angular difference calculated from the first vector and the second vector of the corresponding blade obtained for each blade. If there is any angular difference between any two calculated values greater than the preset angular difference, increase the key point type and quantity to optimize the training of the key point detection algorithm. Re-identify the key points according to the optimized key point detection algorithm, and re-determine the same blade based on the re-identified key points.

2. The method for obtaining the angular velocity of an offshore wind turbine blade based on key points according to claim 1, wherein, It further includes: When the angular velocity of the offshore wind turbine blade cannot be obtained in real time by the sensor, input the environmental data collected at the first moment or the second moment and the historical blade rotation speed at a certain time interval from the first moment or the second moment into the second neural network model, and output the predicted rotation speed of the offshore wind turbine blade at the first moment or the second moment; the input of the second neural network model is the environmental data at the current moment and the historical blade rotation speed at a certain time interval from the current moment, and the output is the rotation speed of the offshore wind turbine blade at the current moment, which is trained and generated by the historical environmental data of the offshore wind turbine blade, the blade rotation speed at a certain time interval from the historical environmental data collection moment of the offshore wind turbine blade, and the rotation speed at the historical environmental data collection moment of the offshore wind turbine blade. Calculate the angular velocity of the offshore wind turbine blade based on the predicted rotation speed of the offshore wind turbine blade, and compare the calculated angular velocity of the offshore wind turbine blade with the angular velocity sequence of the offshore wind turbine blade.

3. The method for obtaining the angular velocity of an offshore wind turbine blade based on key points according to claim 1, wherein It further includes: Judge whether the finally output angular velocity of the offshore wind turbine blade is greater than the preset angular velocity threshold of the offshore wind turbine blade; If it is greater, replace the angular velocity of the offshore wind turbine blade collected by the original sensor with the angular velocity of the offshore wind turbine blade collected by the backup sensor at the first moment or the second moment to compare with the angular velocity sequence of the offshore wind turbine blade.

4. The method for obtaining the angular velocity of an offshore wind turbine blade based on key points according to claim 1, wherein It further includes: Obtain the finally output angular velocity sequence of the offshore wind turbine blade, monitor the moment when the absolute value of the angular velocity change value in the angular velocity sequence of the offshore wind turbine blade is greater than the preset angular velocity change threshold, record it as the distortion moment, and correspondingly obtain and analyze the environmental data at the distortion moment; If the number of types of environmental data that meet the first condition is greater than the preset number, replace the angular velocity of the offshore wind turbine blade collected by the original sensor with the angular velocity of the offshore wind turbine blade collected by the backup sensor at the distortion moment to compare with the angular velocity sequence of the offshore wind turbine blade; the first condition is that the absolute value of the change value of a certain type of environmental data at the distortion moment compared with the corresponding type of environmental data at the previous moment of the distortion moment is greater than the corresponding preset environmental data change threshold.

5. The method for obtaining the angular velocity of an offshore wind turbine blade based on key points according to claim 1, wherein Add Pose as Compositional Tokens to the selected yolov8-pose model network architecture, and replace the convolutional layer with DCNv3.

6. A key-point-based angular velocity system for an offshore wind turbine blade, characterized in that, Adopt the method for obtaining the angular velocity of the offshore wind turbine blade based on key points as described in any one of claims 1-5, including: An image acquisition module for real-time acquisition and preprocessing of the offshore wind turbine blade image; A key point detection module for using a key point detection algorithm to respectively identify key points in the offshore wind turbine blade image at the first moment and the offshore wind turbine blade image at the second moment; the key points include: the center point of the nacelle, the tip of the blade, and the connection point between the nacelle and the blade; the key point detection algorithm selects the yolov8-pose model; A vector acquisition module is configured to generate a first vector with the center point of the nacelle identified in the offshore wind turbine blade image at the first moment as the origin and the tip of any blade identified in the offshore wind turbine blade image at the first moment as the end point; generate a second vector with the center point of the nacelle identified in the offshore wind turbine blade image at the second moment as the origin and the tip of the same blade identified in the offshore wind turbine blade image at the second moment as the end point; the same blade is determined by matching according to the key point matching algorithm. An angular velocity calculation module is configured to calculate the angular difference between the first vector and the second vector, and calculate the offshore wind turbine blade angular velocity sequence by dividing the sum of the angular difference and the angular difference compensation value by the time interval between the first moment and the second moment; the angular difference compensation value is n times of 360 degrees, and n takes an integer of 0 and greater than 0; compare the offshore wind turbine blade angular velocity sequence obtained according to the sensor with the offshore wind turbine blade angular velocity, and select the offshore wind turbine blade angular velocity closest to the offshore wind turbine blade angular velocity collected by the sensor from the offshore wind turbine blade angular velocity sequence.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 5.

8. A computer device, characterized in that, The computer device includes a memory, a processor, and a program stored and executable on the memory. When the program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 5.

Images