Fan blade clearance monitoring method and system, electronic equipment and storage medium

Abstract

The invention relates to the technical field of wind turbine generators, particularly provides a fan blade clearance monitoring method and system, electronic equipment and a storage medium, and aims to solve the problems of low detection precision and high false alarm rate in the existing clearance monitoring technology. Therefore, the draught fan blade clearance monitoring method comprises the steps that monitoring data of a draught fan blade at the current moment are obtained, and the monitoring data comprise clearance data, pitch angle data and blade rotating speed data; the monitoring data are input into a pre-constructed clearance prediction model, operation data of the fan blade at the next moment are predicted, and the clearance prediction model is constructed based on actual operation data and simulation operation data; and controlling to output alarm information based on the operation data of the fan blade at the next moment. According to the method, the clearance prediction model is constructed by using the actual operation data and the simulation operation data, and the accuracy of judging the dangerous tower sweeping working condition of the fan blade is improved, so that the operation safety of a wind turbine generator is effectively guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of wind power generation, and in particular to a method, system, electronic device and storage medium for monitoring the clearance of wind turbine blades. Background Art

[0002] In the field of wind power generation, with the large-scale wind turbine blades and the rapid development of wind farms, it is particularly important to ensure the safe operation of wind turbines. The clearance distance between the blades and the tower is one of the key factors affecting the safety of wind turbines. The introduction of machine learning and machine vision technology provides a new solution for the clearance monitoring of wind turbines, making the monitoring and prediction of blade clearance more intuitive and accurate. There are related video clearance systems to monitor and predict blade clearance to ensure the safety of the unit. However, the current clearance prediction system has problems such as low accuracy and high false alarm rate, which leads to the problem of "warning but not prevention" in the clearance warning and anti-sweeping tower system. In addition, as wind farms develop towards more complex and harsh environments, such as deserts, Gobi, and desert areas, higher requirements are placed on the environmental adaptability and accuracy of clearance prediction technology.

[0003] Accordingly, the art needs a new fan blade clearance monitoring solution to solve the above problems. Summary of the invention

[0004] In order to overcome the above-mentioned defects, the present application is proposed to solve or at least partially solve the technical problems of low accuracy and high false alarm rate in the clearance prediction system.

[0005] In a first aspect, a method for monitoring wind turbine blade clearance is provided, the method comprising: obtaining monitoring data of the wind turbine blade at a current moment, wherein the monitoring data comprises clearance data, pitch angle data and blade speed data; inputting the monitoring data into a pre-constructed clearance prediction model to predict operating data of the wind turbine blade at a next moment, the clearance prediction model being constructed based on actual operating data and simulated operating data; and controlling output of alarm information based on the operating data of the wind turbine blade at the next moment.

[0006] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the clearance prediction model is constructed by the following steps: obtaining actual operation data; obtaining simulation operation data; determining a training data set based on the actual operation data and the simulation operation data; and constructing a clearance prediction model based on the training data set.

[0007] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the obtaining of actual operation data includes: obtaining clearance data, corresponding pitch angle data and blade speed data of wind turbines of different models under normal operating conditions as actual operation data.

[0008] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the acquisition of simulation operation data includes: using a computational fluid dynamics method to solve the Navier-Stokes equations to determine the flow field information of the wind turbine; using a wind turbine actuation line method to perform performance analysis on the wind turbine rotor blades based on the flow field information of the wind turbine, the aerodynamic characteristics of the two-dimensional airfoil and momentum theory; using momentum theory to consider the influence of the wind rotor on the incoming flow to determine the aerodynamic performance-related parameters of the wind rotor; based on the aerodynamic performance-related parameters of the wind rotor, obtaining the operation data of the wind turbine under different loads; based on preset screening conditions, screening the operation data under different loads to obtain simulation clearance data under dangerous conditions, corresponding simulation pitch angle data and simulation blade speed data as simulation operation data, wherein the dangerous conditions include extreme wind speed conditions, extreme wind direction change conditions, fault conditions, special environmental conditions and operating conditions.

[0009] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the training data set is determined based on the actual operation data and the simulated operation data, including: performing data preprocessing on the actual operation data and the simulated operation data respectively to obtain the training data set, and the data preprocessing includes at least one of missing value processing, outlier processing, noise reduction processing, and normalization processing.

[0010] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the clearance prediction model is constructed based on the training data set, including: constructing a long short-term memory neural network model, using the training data set to train the long short-term memory neural network model, taking the clearance data, pitch angle data and blade speed data at the current moment as input, the long short-term memory neural network outputs the clearance data, pitch angle data and blade speed data at the next moment, using the mean square error as the model loss function, and selecting Adam as the model optimizer, using the mean absolute error to evaluate the prediction performance of the long short-term memory neural network model, and obtaining the trained clearance prediction model when the prediction performance reaches the preset standard.

[0011] In a technical solution of the above-mentioned wind turbine blade clearance monitoring method, the control output of alarm information is based on the operating data of the wind turbine blade at the next moment, including: judging whether the wind turbine blade has a tower sweep risk based on the operating data of the wind turbine blade at the next moment, and if so, generating an alarm information; if not, judging it to be in a safe state.

[0012] In a second aspect, a wind turbine blade clearance monitoring system is provided, the system comprising: an acquisition module for acquiring monitoring data of the wind turbine blade at a current moment; a prediction module for using a pre-built clearance prediction model to predict the operating data of the wind turbine blade at the next moment, the clearance prediction model being constructed based on actual operating data and simulated operating data; and a control module for controlling output of alarm information based on the operating data of the wind turbine blade at the next moment.

[0013] In a third aspect, an electronic device is provided, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any one of the technical solutions of the above-mentioned wind blade clearance monitoring method is implemented.

[0014] In a fourth aspect, a computer-readable storage medium is provided, wherein a plurality of program codes are stored therein, wherein the program codes are suitable for being loaded and run by a processor to execute the method described in any one of the technical solutions of the above-mentioned wind turbine blade clearance monitoring method.

[0015] The above one or more technical solutions of the present application have at least one or more of the following beneficial effects:

[0016] The wind turbine blade clearance monitoring method provided in the present application includes: obtaining monitoring data of the wind turbine blade at the current moment, wherein the monitoring data includes clearance data, pitch angle data and blade speed data; inputting the monitoring data into a pre-constructed clearance prediction model to predict the operation data of the wind turbine blade at the next moment, wherein the clearance prediction model is constructed based on actual operation data and simulated operation data; and controlling the output of alarm information based on the operation data of the wind turbine blade at the next moment. The present application uses a numerical simulation method to obtain simulation operation data, and the simulation operation data can simulate and supplement dangerous working conditions that may be difficult to capture in actual operation. By combining actual operation data and simulated operation data to construct a clearance prediction model, the prediction ability of the model under dangerous working conditions can be significantly improved, thereby issuing alarm information according to the prediction results, effectively ensuring the operation safety of the wind turbine set, and reducing losses caused by unexpected shutdowns. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The disclosure of the present application will become easier to understand with reference to the accompanying drawings. It is easy for those skilled in the art to understand that these drawings are only for illustrative purposes and are not intended to limit the scope of protection of the present application. Among them:

[0018] Figure 1 It is a schematic flow chart of the main steps of a method for monitoring the clearance of wind turbine blades according to an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of detailed steps of a fan blade clearance monitoring method according to an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of the main structure of a wind turbine blade clearance monitoring system according to an embodiment of the present application;

[0021] Figure 4 It is a schematic diagram of the main structure of an electronic device according to an embodiment of the present application.

[0022] Reference numerals:

[0023] 11: memory; 12: processor; 31: acquisition module; 32: prediction module; 33: control module. DETAILED DESCRIPTION

[0024] Some embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application and are not intended to limit the protection scope of the present application.

[0025] In the description of the present application, "module" and "processor" may include hardware, software or a combination of the two. A module may include hardware circuits, various suitable sensors, communication ports, memory, and may also include software parts, such as program code, or a combination of software and hardware. The processor may be a central processing unit, a microprocessor, an image processor, a digital signal processor or any other suitable processor. The processor has data and / or signal processing functions. The processor may be implemented in software, hardware or a combination of the two. Computer-readable storage media include any suitable medium that can store program code, such as a disk, a hard disk, an optical disk, a flash memory, a read-only memory, a random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as only A, only B or A and B. The term "at least one A or B" or "at least one of A and B" has a similar meaning to "A and / or B" and may include only A, only B or A and B. The singular terms "one" and "the" may also include plural forms.

[0026] The current clearance prediction system has problems such as low accuracy and high false alarm rate, which leads to the problem that the clearance warning and anti-sweeping tower system can "warn but cannot prevent". In addition, as wind farms develop into more complex and harsh environments, such as deserts, Gobi and wasteland areas, higher requirements are placed on the environmental adaptability and accuracy of clearance prediction technology.

[0027] To this end, the wind turbine blade clearance monitoring method provided by the present application includes: obtaining monitoring data of the wind turbine blade at the current moment, wherein the monitoring data includes clearance data, pitch angle data and blade speed data; inputting the monitoring data into a pre-constructed clearance prediction model to predict the operation data of the wind turbine blade at the next moment, wherein the clearance prediction model is constructed based on actual operation data and simulated operation data; and controlling the output of alarm information based on the operation data of the wind turbine blade at the next moment. The present application constructs a clearance prediction model by combining the actual operation data collected and the simulated operation data supplemented by numerical simulation, which can accurately predict the operation data at the next moment and provide buffer time for potential dangerous clearance. It can be widely used in wind turbines of various types and sizes, and provide strong support for the safe operation of the wind power industry.

[0028] See attached Figure 1 , Figure 1 FIG. 1 is a flow chart of the main steps of a method for monitoring the clearance of wind turbine blades according to an embodiment of the present application. Figure 1 As shown, the wind turbine blade clearance monitoring method in the embodiment of the present application mainly includes the following steps S101 to S103.

[0029] Step S101: acquiring monitoring data of a wind turbine blade at a current moment, wherein the monitoring data includes clearance data, pitch angle data and blade speed data;

[0030] In this embodiment, the clearance data of the fan blade refers to the minimum gap between the blade and the surrounding environment when the fan rotates. The pitch angle (PitchAngle) is also called the pitch angle, which refers to the angle between the chord line of the airfoil at the top of the fan blade and the rotating plane.

[0031] Step S102: inputting the monitoring data into a pre-built clearance prediction model to predict the operation data of the wind turbine blade at the next moment, wherein the clearance prediction model is built based on actual operation data and simulation operation data.

[0032] In this embodiment, the monitoring data at the current moment is input into the clearance prediction model. After receiving the input data, the clearance prediction model will perform calculation and analysis to predict the clearance data, pitch angle data and blade speed data of the wind turbine blades at the next moment.

[0033] Step S103: Based on the operation data of the wind turbine blade at the next moment, control the output of alarm information.

[0034] Based on the method described in steps S101 to S103 above, the present application uses the monitoring data obtained in real time from the wind turbine blades as the input of the clearance prediction model to predict the operating data of the wind turbine at the next moment, and further analyzes and judges the predicted data, thereby controlling the output of alarm information to provide buffer time for potential dangerous clearance. The application can be widely used in wind turbines of various types and sizes, and provide strong support for the safe operation of the wind power industry.

[0035] The above steps S101 to S103 are further explained below.

[0036] See attached Figure 2 , Figure 2 It is a detailed step flow chart of a method for monitoring wind turbine blade clearance according to an embodiment of the present application.

[0037] like Figure 2 As shown, the fan blade clearance monitoring method of this embodiment includes the following steps:

[0038] In one embodiment, the clearance prediction model is constructed by the following steps: obtaining actual operation data; obtaining simulation operation data; determining a training data set based on the actual operation data and the simulation operation data; and constructing a clearance prediction model based on the training data set.

[0039] Specifically, the clearance prediction algorithm currently using machine learning is designed based on short-term clearance data and operating condition data. The proportion of dangerous operating condition data in the data set is small, and it cannot cover all dangerous operating conditions, so the prediction ability of the model built by machine learning is limited. The method of the present application first collects the actual operating data of the wind turbine set, and simulates the dangerous operating conditions of the wind turbine set through a simulation method to obtain simulated operating data. The simulated operating data includes the operating data of the wind turbine set under dangerous operating conditions obtained by numerical simulation. The simulated operating data can effectively supplement the actual operating data, and then determine the training data set based on the actual operating data and the simulated operating data, and build a clearance prediction model based on the training data set, so that the clearance prediction model built with the actual operating data and the simulated operating data as training data has the ability to judge dangerous conditions. Improved, thereby improving the early warning capability and effectively ensuring the safe operation of the unit.

[0040] In one embodiment, the obtaining of actual operation data includes: obtaining clearance data, corresponding pitch angle data, and blade speed data of wind turbines of different models under normal operating conditions as actual operation data.

[0041] Specifically, the clearance data of wind turbines of different models under normal operating conditions and the corresponding pitch angle data and blade speed data are collected.

[0042] In one embodiment, the acquisition of simulation operation data includes: using a computational fluid dynamics method to solve the Navier-Stokes equations to determine the flow field information of the wind turbine; using a wind turbine actuation line method to perform performance analysis on the wind turbine blades based on the flow field information of the wind turbine, the aerodynamic characteristics of the two-dimensional airfoil, and momentum theory; using momentum theory to consider the impact of the wind rotor on the incoming flow to determine the aerodynamic performance-related parameters of the wind rotor; based on the aerodynamic performance-related parameters of the wind rotor, obtaining the operation data of the wind turbine under different loads; based on preset screening conditions, screening the operation data under different loads to obtain simulated clearance data, corresponding simulated pitch angle data, and simulated blade speed data under dangerous conditions as simulation operation data, wherein the dangerous conditions include extreme wind speed conditions, extreme wind direction change conditions, fault conditions, special environmental conditions, and operating conditions.

[0043] Specifically, first, a three-dimensional geometric model of the wind turbine is constructed, including the tower, nacelle, wind rotor and other components, to ensure that the model can truly reflect the actual structure and size of the wind turbine. Then, the three-dimensional meshing technology is used to divide the calculation domain around the wind turbine into discrete grid units. For key areas such as the wind rotor, encryption processing is performed to improve the calculation accuracy. And according to the actual atmospheric environmental conditions, the wind speed, wind direction, turbulence intensity and other parameters of the inlet boundary are set; the outlet boundary is set to the fully developed condition; the wall boundary is set to the no-slip boundary condition according to the surface characteristics of the wind turbine components. And, select the simple solver in Fluent to solve the Navier-Stokes equations, and set the control parameters of the solution, such as time step, number of iterations, etc., to obtain the flow field information of the wind turbine.

[0044] Then, the wind rotor blades are divided into multiple micro-element segments (blade elements) at equal intervals along the span direction or according to specific rules, and each blade element has independent aerodynamic characteristics. Through experimental measurement or numerical simulation and other methods, the aerodynamic characteristic data of the two-dimensional airfoil corresponding to each blade element at different angles of attack, such as lift coefficient, drag coefficient, etc., are obtained. Based on momentum theory, the momentum exchange of each blade element to the incoming flow is considered, combined with the aerodynamic characteristics of the two-dimensional airfoil, the force condition of each blade element is calculated, and then the performance analysis of the wind turbine rotor blades is carried out.

[0045] Also, considering the disturbance of the wind rotor on the incoming flow, the momentum conservation equation before and after the wind rotor is established through momentum theory, and the changes in the incoming flow parameters such as velocity and pressure are analyzed. According to the force of the wind rotor and the changes in the incoming flow parameters, the aerodynamic performance related parameters of the wind rotor, such as power, axial force, torque and other parameters, are calculated. And by calculating the aerodynamic performance related parameters of the wind rotor under different working conditions, the operating data of the wind turbine under different loads are obtained.

[0046] Finally, the calculated operating data of wind turbines under different loads are sorted and stored, and classified according to different working conditions (such as wind speed, wind direction, turbulence intensity, etc.). According to the design standards and actual operating requirements of wind turbines, the judgment criteria of extreme load conditions are determined as preset screening conditions, and then the operating data under different loads are screened according to the preset screening conditions, and the operating data under the corresponding extreme load conditions are screened out, and the screened operating data under the extreme load conditions are used as simulation operating data under dangerous conditions.

[0047] Dangerous operating conditions specifically include extreme wind speed conditions, extreme wind direction change conditions, fault conditions, special environmental conditions and operating conditions, among which extreme wind speed conditions refer to extreme working gusts, extreme wind speed models or extreme wind speed shear; extreme working gusts, such as extreme gusts, may cause power outages in the wind turbine at the beginning, the lowest wind speed, the highest wind speed and when the gust acceleration is maximum, within the range from the cut-in wind speed to the cut-out wind speed, resulting in huge blade flapping bending moment and tower bottom overturning bending moment, causing serious impact on the unit structure.

[0048] Extreme wind speed model: The simulation takes into account extreme wind conditions that exceed the design wind speed of the unit to test the structural strength and stability of the unit under extreme wind speeds. For example, short-term extreme wind speeds may cause components such as blades and towers to bear loads exceeding the limit, resulting in damage.

[0049] Extreme wind speed shear: When the wind speed changes too fast in the vertical direction, different parts of the wind rotor will be subjected to uneven forces, increasing the fatigue load on the blades and the transmission system, and may even cause distortion of the blades and damage to the transmission components.

[0050] Extreme wind direction change conditions refer to extreme wind direction changes. For example, a drastic change in wind direction in a short period of time may make the yaw system unable to respond in time, causing the wind turbine to bear loads in non-design directions, generating additional torque and bending moment on the tower, nacelle and blades, and increasing the risk of structural damage.

[0051] Fault conditions refer to grid faults, mechanical faults, and control system faults; grid faults include grid voltage sag, frequency anomalies, three-phase imbalance, and other faults that affect the grid-connected operation and power output of wind turbines, and may cause generator overload, inverter damage, or even cause the unit to go offline, causing harm to the grid and the unit itself.

[0052] Mechanical failures such as blade breakage, gearbox failure, bearing damage and other mechanical component failures, analyze the impact of the failure on the overall performance and structure of the unit. For example, blade breakage will destroy the balance of the wind wheel, cause violent vibrations, and have a huge impact on the tower and foundation.

[0053] Control system failures such as pitch system failures make it impossible to adjust the blade pitch angle normally, resulting in the inability of the unit to effectively control power and load; yaw system failures make it impossible for the wind turbine to align with the wind direction in time, reducing power generation efficiency and may also cause the unit to bear abnormal loads.

[0054] Special environmental conditions refer to special conditions such as freezing conditions and earthquake conditions. Frozen conditions are when ice forms on the surface of blades, towers and other components in a low-temperature environment, increasing the weight of the components and aerodynamic resistance, changing the aerodynamic performance of the blades, causing the power of the unit to drop, and may even cause ice to fall off, posing a safety hazard. Earthquake conditions are when earthquakes are considered in earthquake-prone areas, and the effects of earthquake acceleration and frequency on the foundation, tower and superstructure of wind turbines are analyzed to evaluate the seismic performance of the unit in earthquakes and prevent foundation settlement and tower tilt or collapse.

[0055] Operating conditions refer to emergency shutdown conditions or startup and grid-connected conditions; Emergency shutdown conditions: If an emergency shutdown command is suddenly triggered during operation, the unit quickly feathers and disconnects the inverter, which may cause the blade root and the bottom of the tower to bear huge load impacts. It is necessary to analyze the load changes during the shutdown process through numerical simulation and optimize the shutdown strategy. Startup and grid-connected conditions: The speed increase process and grid-connected moment when the wind turbine is started will produce current shocks and mechanical stress changes. These processes need to be paid attention to in numerical simulation to ensure that the electrical and mechanical components of the unit can withstand the corresponding stress.

[0056] In one embodiment, determining the training data set based on the actual operation data and the simulation operation data includes: performing data preprocessing on the actual operation data and the simulation operation data, respectively, to obtain the training data set, wherein the data preprocessing includes at least one of missing value processing, outlier processing, noise reduction processing, and normalization processing.

[0057] Specifically, the collected actual operation data and simulation operation data are preprocessed, such as checking whether there are missing values ​​in the data. For missing values, the samples containing missing values ​​can be deleted, or the mean, median and other methods can be used for processing; checking whether there are outliers in the data. If there are outliers, strategies such as deletion and replacement can be used for processing; denoising processing refers to the use of commonly used denoising methods, such as smoothing filtering, wavelet transform, Kalman filtering or particle filtering, to reduce the noise in the data and improve the signal-to-noise ratio of the data, thereby improving the training effect of the model; normalization processing refers to the use of commonly used normalization methods to scale data with different features to the same scale, thereby eliminating the dimensional differences between the features and improving the training speed and performance of the model. After completing the above data preprocessing steps, the processed actual operation data and simulation operation data are merged into a training data set.

[0058] In one embodiment, the construction of the clearance prediction model based on the training data set includes: constructing a long short-term memory neural network model, using the training data set to train the long short-term memory neural network model, taking the clearance data, pitch angle data and blade speed data at the current moment as input, the long short-term memory neural network outputs the clearance data, pitch angle data and blade speed data at the next moment, using the mean square error as the model loss function, and selecting Adam as the model optimizer, using the mean absolute error to evaluate the prediction performance of the long short-term memory neural network model, and obtaining the trained clearance prediction model when the prediction performance reaches the preset standard.

[0059] Specifically, the long short-term memory neural network model (LSTM) is a special recurrent neural network (RNN) that is good at processing time series data. In this embodiment, a multi-layer LSTM network is constructed as a long short-term memory neural network model, and the specific structure is as follows:

[0060] Input layer: The input layer receives the training data set as input, specifically using clearance data, pitch angle data, and blade speed data as input.

[0061] Network layer: Set up a three-layer LSTM. The number of hidden units in the first LSTM layer is 128, and the output sequence is passed to the next layer. The number of units in the second LSTM layer is 84, and the number of units in the third LSTM layer can be adjusted according to actual conditions.

[0062] Dropout layer: A Dropout layer is added after each LSTM layer, and the dropout rate is set to 0.2 to prevent overfitting and randomly drop some neurons.

[0063] The output layer is a fully connected layer (Dense layer) with an output dimension of 3, corresponding to the predicted clearance data, pitch angle data, and blade speed data at the next moment.

[0064] The construction of the long short-term memory neural network model specifically uses the Sequential model of Keras as the basic framework of the model. Add an LSTM layer to process time series data or data with sequence dependencies. Add a Dropout layer to randomly discard the output of some neurons during training to reduce overfitting. Add an output layer (fully connected layer, Dense layer) to output the final prediction results.

[0065] Then, the preprocessed training data set is used to train the blade clearance prediction model. The clearance prediction model continuously optimizes the prediction model by continuously learning the actual operation data and combining the simulation operation data obtained by numerical simulation. The specific training steps of the blade clearance prediction model include:

[0066] Select the Adam optimizer and set the learning rate to 0.001. The Adam optimizer is an adaptive learning rate method based on first-order and second-order moment estimates.

[0067] The mean square error (MSE) is used as the loss function to measure the difference between the model's predicted value and the actual value. The mean absolute error (MAE) is selected as the evaluation indicator to monitor the performance of the model during the training process.

[0068] Use the fit method, input the input data X and target data Y, set epochs to 100, and batch_size to 32 for training. epochs indicates the number of training rounds. Setting epochs to 100 means that the model will traverse the entire training data set 100 times. In each round of training, the model will update its internal weights and parameters to minimize the loss function. batch_size indicates the number of samples used each time the model weights are updated. Setting batch_size to 32 means that the model will use 32 samples of data to calculate a gradient update.

[0069] In each epoch, the training data will be divided into multiple batches. For batch_size of 32, if the training data has 1000 samples, each epoch will contain 1000 / 32 = 31.25 batches, which are usually rounded up to 32 batches. For each batch, the model will predict the output based on the input data and the current weights, and then calculate the difference between the predicted result and the target data Y, which is measured by the loss function (such as MSE).

[0070] The model then uses the Adam optimizer to update the weights based on the gradient of the loss function to minimize the loss. The optimizer adjusts the weights based on the gradient to make the model's predictions closer to the true target data. This process is repeated in each batch until the entire training dataset has been used once, completing an epoch.

[0071] For step S101, the monitoring data of the wind turbine blades at the current moment is obtained, wherein the monitoring data includes clearance data, pitch angle data and blade speed data;

[0072] Specifically, the distance between the blade and the surrounding environment (such as a tower, other blades or obstacles) can be monitored in real time by installing sensors or laser rangefinders and other equipment. By using image recognition technology, the clearance distance between the blade and the surrounding environment can be calculated by analyzing the video or image captured by the camera to obtain clearance data. The pitch angle is directly measured by an angle sensor installed on the blade to obtain pitch angle data. The blade speed is directly measured by a speed sensor installed on the generator shaft or blade. In some specific embodiments, the blade speed can also be determined by analyzing parameters such as the generator output power or current.

[0073] With respect to step S102, in one embodiment, the monitoring data is input into a pre-built clearance prediction model to predict the operation data of the wind turbine blade at the next moment, and the clearance prediction model is built based on actual operation data and simulated operation data.

[0074] Specifically, the monitoring data of wind turbine blades is collected. Before the monitoring data is input into the clearance prediction model, the monitoring data can be preprocessed, such as removing noise, outliers, and normalizing. The preprocessed monitoring data is input into the pre-built clearance prediction model. The model will predict the operating data of the wind turbine blades at the next moment based on the learned knowledge and rules.

[0075] For step S103, in one embodiment, the control of outputting alarm information based on the operating data of the wind turbine blade at the next moment includes: judging whether the wind turbine blade has a tower sweep risk based on the operating data of the wind turbine blade at the next moment, and if so, generating an alarm message; if not, judging it to be in a safe state.

[0076] Specifically, based on the predicted operation data at the next moment, it is determined whether the wind turbine blades have a tower sweep risk. Tower sweep risk refers to the risk that the wind turbine blades may collide with the wind turbine tower during rotation. If it is determined that the wind turbine blades have a tower sweep risk, an alarm message is immediately generated. The alarm message can be a display alarm, a speaker alarm, or an indicator light alarm, and can also include information such as the specific type of risk, possible consequences, and recommended countermeasures. In addition, corresponding control functions can be set so that emergency measures can be automatically taken when a tower sweep risk is detected, such as stopping the wind turbine operation, adjusting the blade angle, etc., to prevent accidents and reduce losses.

[0077] If it is determined that there is no risk of tower sweep, the wind turbine blades are considered to be in a safe state and their operating status can continue to be monitored.

[0078] It should be pointed out that although the various steps in the above embodiments are described in a specific order, those skilled in the art can understand that in order to achieve the effect of the present application, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders. These adjusted schemes are equivalent to the technical schemes described in this application, and therefore will also fall within the scope of protection of this application.

[0079] It is understood by those skilled in the art that all or part of the processes in the method for implementing the above-mentioned embodiment of the present application can also be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device, medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium, etc. that can carry the computer program code.

[0080] The present application also provides a fan blade clearance monitoring system.

[0081] See attached Figure 3 , Figure 3 FIG. 1 is a schematic diagram of the main structure of a fan blade clearance monitoring system according to an embodiment of the present application. Figure 3 As shown, a wind turbine blade clearance monitoring system of the present application includes: an acquisition module 31, a prediction module 32 and a control module 33. In some embodiments, one or more of the acquisition module 31, the prediction module 32 and the control module 33 can be combined into one module.

[0082] In some embodiments, the acquisition module 31 can be configured to acquire the monitoring data of the fan blade at the current moment; the prediction module 32 can be configured to use a pre-built clearance prediction model to predict the operation data of the fan blade at the next moment, and the clearance prediction model is built based on the actual operation data and the simulation operation data; the control module 33 can be configured to control the output of the alarm information based on the operation data of the fan blade at the next moment. In one embodiment, the description of the specific implementation function can be referred to the description of steps S101 to S103.

[0083] The above-mentioned wind turbine blade clearance monitoring system is used to perform Figure 1The embodiment of the wind blade clearance monitoring method shown in the figure has similar technical principles, technical problems solved and technical effects produced. Technicians in this technical field can clearly understand that for the convenience and conciseness of description, the specific working process and related instructions of the wind blade clearance monitoring system can refer to the contents described in the embodiment of the wind blade clearance monitoring method, which will not be repeated here.

[0084] Another aspect of the present application provides an electronic device.

[0085] In an embodiment of an electronic device according to the present application, the electronic device may include at least one processor; and a memory connected to the at least one processor in communication; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any of the above embodiments is implemented. The electronic device described in the present application may include a driving device, a smart car, a robot, and the like. Figure 4 , Figure 4 FIG. 4 exemplarily shows that the memory 11 and the processor 12 are communicatively connected via a bus.

[0086] In some embodiments of the present application, the electronic device may further include at least one sensor, which is used to sense information and is communicatively connected to any type of processor mentioned in the present application.

[0087] Another aspect of the present application also provides a computer-readable storage medium.

[0088] In an embodiment of a computer-readable storage medium according to the present application, the computer-readable storage medium can be configured to store a program for executing the wind blade clearance monitoring method of the above-mentioned method embodiment, and the program can be loaded and run by the processor to implement the above-mentioned wind blade clearance monitoring method. For ease of explanation, only the parts related to the embodiment of the present application are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present application. The computer-readable storage medium can be a storage device formed by various electronic devices. Optionally, the computer-readable storage medium in the embodiment of the present application is a non-temporary computer-readable storage medium.

[0089] So far, the technical solution of the present application has been described in conjunction with an embodiment shown in the accompanying drawings, but it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments. Without departing from the principles of the present application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present application.

Claims

1. A fan blade clearance monitoring method, characterized in that: The method comprises: Acquire monitoring data of the wind turbine blades at the current moment, wherein the monitoring data includes clearance data, pitch angle data and blade speed data; Inputting the monitoring data into a pre-built clearance prediction model to predict the operation data of the wind turbine blade at the next moment, wherein the clearance prediction model is built based on actual operation data and simulated operation data; Based on the operation data of the fan blades at the next moment, the output of alarm information is controlled.

2. The fan blade clearance monitoring method according to claim 1, characterized in that: The headroom prediction model is constructed using the following steps: Obtain actual operation data; Get simulation running data; Determining a training data set based on the actual operation data and the simulation operation data; Based on the training data set, a clearance prediction model is constructed.

3. The fan blade clearance monitoring method according to claim 2, characterized in that: The obtaining of actual operation data includes: The clearance data, corresponding pitch angle data and blade speed data of wind turbines of different models under normal operating conditions are obtained as actual operating data.

4. The fan blade clearance monitoring method according to claim 2, characterized in that: The obtaining of simulation operation data includes: Computational fluid dynamics methods are used to solve the Navier-Stokes equations and determine the flow field information of the wind turbine; Using a wind turbine actuation line method, the performance of the wind turbine rotor blades is analyzed based on the flow field information of the wind turbine, the aerodynamic characteristics of the two-dimensional airfoil, and momentum theory; Using momentum theory and considering the effect of the wind rotor on the incoming flow, determine the relevant parameters of the wind rotor's aerodynamic performance; Based on the aerodynamic performance related parameters of the wind rotor, the operation data of the wind turbine under different loads are obtained; Based on preset screening conditions, the operating data under the different loads are screened to obtain simulated clearance data, corresponding simulated pitch angle data and simulated blade speed data under dangerous conditions as simulated operating data, wherein the dangerous conditions include extreme wind speed conditions, extreme wind direction change conditions, fault conditions, special environmental conditions and operating conditions.

5. The fan blade clearance monitoring method according to claim 2, characterized in that: The determining of a training data set based on the actual operation data and the simulation operation data includes: The actual operation data and the simulation operation data are respectively preprocessed to obtain a training data set, wherein the data preprocessing includes at least one of missing value processing, outlier processing, noise reduction processing, and normalization processing.

6. The wind turbine blade clearance monitoring method according to claim 2, characterized in that: The step of constructing a clearance prediction model based on the training data set includes: A long short-term memory neural network model is constructed, and the long short-term memory neural network model is trained using a training data set. The clearance data, pitch angle data and blade speed data at the current moment are taken as input, and the long short-term memory neural network outputs the clearance data, pitch angle data and blade speed data at the next moment. The mean square error is used as the model loss function, and Adam is selected as the model optimizer. The mean absolute error is used to evaluate the prediction performance of the long short-term memory neural network model. When the prediction performance reaches the preset standard, the trained clearance prediction model is obtained.

7. The wind turbine blade clearance monitoring method according to claim 1, characterized in that: The controlling and outputting of alarm information based on the operation data of the fan blade at the next moment includes: Determine whether the fan blade has a tower sweep risk based on the operation data of the fan blade at the next moment, and if so, generate an alarm message; If not, it is judged to be a safe state.

8. A fan blade clearance monitoring system, characterized in that: The system comprises: An acquisition module is used to acquire the monitoring data of the fan blades at the current moment; A prediction module, used to predict the operation data of the wind turbine blade at the next moment by using a pre-built clearance prediction model, wherein the clearance prediction model is built based on actual operation data and simulated operation data; The control module is used to control the output of alarm information based on the operation data of the fan blades at the next moment.

9. An electronic device comprising at least one processor and at least one memory, wherein the memory is suitable for storing a plurality of program codes, wherein: The program code is suitable for being loaded and run by the processor to execute the wind turbine blade clearance monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by a processor to execute the wind turbine blade clearance monitoring method according to any one of claims 1 to 7.

Images