Fault diagnosis and control method, device and electronic equipment of wind turbine generator system

Abstract

This application provides a method, apparatus, and electronic device for fault diagnosis and control of wind turbine generators. The method includes: acquiring the state parameters and environmental parameters of the wind turbine generator under test; acquiring the fault type and fault level of the wind turbine generator under test based on the state parameters and environmental parameters; selecting a target control scheme from candidate control schemes based on the fault type and fault level; and determining the target control parameters and target control strategy of the wind turbine generator under test based on the target control scheme. This application improves the accuracy and efficiency of fault diagnosis of wind turbine generators, and can select the corresponding target control scheme according to different fault types and fault levels, dynamically adjust the control parameters and strategies of the wind turbine generator to determine the target control parameters and target control strategy of the wind turbine generator under test, optimize the operating efficiency and reliability of the wind turbine generator, reduce the wear and loss of the wind turbine generator, and extend the service life of the wind turbine generator.

Description

Technical Field

[0001] This application relates to the field of wind power generation technology, and in particular to a method, apparatus and electronic equipment for fault diagnosis and control of wind turbine generators. Background Technology

[0002] Currently, wind turbine fault diagnosis solutions often rely on quasi-static detection of state parameters, such as vibration signals, oil parameters, temperature parameters, and generator output current and voltage. However, this approach fails to meet the requirements for wind turbine fault diagnosis. Furthermore, these solutions often analyze only a single or limited number of signal sources, neglecting the correlation and complementarity between various state parameters and environmental information. They cannot comprehensively reflect the operating status and fault characteristics of the wind turbine. When relying on experience or models, these solutions do not consider the uncertainties and complexities of wind turbine operation, and cannot adapt to the changing operating conditions and fault types. These fault diagnosis methods reduce the accuracy and practicality of wind turbine fault diagnosis, while also impacting operating efficiency and reliability. Therefore, improving the accuracy of fault diagnosis and enhancing the operating efficiency of wind turbines through fault diagnosis and control has become an urgent problem to be solved. Summary of the Invention

[0003] The purpose of this application is to at least partially solve one of the technical problems in the aforementioned technologies.

[0004] The first aspect of this application provides a method for fault diagnosis and control of wind turbine generators, comprising: acquiring state parameters and environmental parameters of the wind turbine generator to be tested; acquiring the fault type and fault level of the wind turbine generator to be tested based on the state parameters and the environmental parameters; selecting a target control scheme from candidate control schemes based on the fault type and the fault level; and determining the target control parameters and target control strategy of the wind turbine generator to be tested based on the target control scheme.

[0005] A second aspect of this application provides a fault diagnosis and control device for wind turbine generators, comprising: a first acquisition module for acquiring state parameters and environmental parameters of the wind turbine generator under test; a second acquisition module for acquiring the fault type and fault level of the wind turbine generator under test based on the state parameters and the environmental parameters; a selection module for selecting a target control scheme from candidate control schemes based on the fault type and fault level; and a determination module for determining the target control parameters and target control strategy of the wind turbine generator under test based on the target control scheme.

[0006] A third aspect of this application provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the fault diagnosis and control method for wind turbine generators provided in the first aspect of this application.

[0007] A fourth aspect of this application provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the fault diagnosis and control method for wind turbine generators provided in the first aspect of this application.

[0008] A fifth aspect of this application provides a computer program product that, when executed by an instruction processor, performs the fault diagnosis and control method for wind turbine generators provided in the first aspect of this application.

[0009] The fault diagnosis and control method and apparatus for wind turbines provided in this application acquires the state parameters and environmental parameters of the wind turbine under test, obtains the fault type and fault level of the wind turbine under test based on the state parameters and environmental parameters, selects a target control scheme from candidate control schemes based on the fault type and fault level, and determines the target control parameters and target control strategy of the wind turbine under test based on the target control scheme. This application improves the accuracy and efficiency of fault diagnosis of wind turbines, and can select the corresponding target control scheme according to different fault types and fault levels, dynamically adjust the control parameters and strategies of the wind turbine to determine the target control parameters and target control strategy of the wind turbine under test, optimize the operating efficiency and reliability of the wind turbine, reduce the wear and loss of the wind turbine, and extend the service life of the wind turbine.

[0010] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0011] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0012] Figure 1 This is a flowchart illustrating a method for fault diagnosis and control of a wind turbine generator according to an embodiment of this application.

[0013] Figure 2 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0014] Figure 3This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0015] Figure 4 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0016] Figure 5 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0017] Figure 6 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0018] Figure 7 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application.

[0019] Figure 8 This is a schematic diagram of the structure of a fault diagnosis and control device for a wind turbine according to an embodiment of this application;

[0020] Figure 9 This is a block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0021] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0022] The following description, with reference to the accompanying drawings, outlines a method, apparatus, electronic device, and medium for fault diagnosis and control of wind turbine generators according to embodiments of this application.

[0023] Figure 1 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine according to an embodiment of this application, as shown below. Figure 1 As shown, the method includes:

[0024] S101, obtain the status parameters and environmental parameters of the wind turbine to be tested.

[0025] It should be noted that this application does not limit the specific method for obtaining the state parameters and environmental parameters of the wind turbine under test, and the appropriate method can be selected according to the actual situation.

[0026] Alternatively, the internal sensors of the wind turbine can be used to obtain the status parameters of the wind turbine under test.

[0027] For example, the temperature, pressure, power, and speed of the wind turbine can be obtained through internal sensors such as temperature sensors, pressure sensors, power sensors, and speed sensors.

[0028] Alternatively, external sensors of the wind turbine can be used to obtain environmental parameters of the wind turbine under test.

[0029] For example, environmental parameters such as wind speed, wind direction, temperature, and air pressure of the wind turbine can be obtained through external sensors such as wind speed sensors, wind direction sensors, temperature sensors, and air pressure sensors of the wind turbine.

[0030] It should be noted that the status parameters and environmental parameters of the wind turbine under test can reflect the operating status and external conditions of the wind turbine under test, and can provide a data basis for fault diagnosis and adaptive control of the wind turbine under test.

[0031] S102, based on the status parameters and environmental parameters, obtain the fault type and fault level of the wind turbine unit under test.

[0032] In this embodiment of the application, after obtaining the status parameters and environmental parameters, the fault type and fault level of the wind turbine to be tested can be obtained based on the status parameters and environmental parameters.

[0033] It should be noted that this application does not limit the specific method for obtaining the fault type and fault level of the wind turbine under test based on state parameters and environmental parameters, and the method can be selected according to the actual situation.

[0034] Alternatively, deep learning technology can be used to input state parameters and environmental parameters into a target fault diagnosis model based on a convolutional neural network (CNN), and the target fault diagnosis model can output the fault type and fault level of the wind turbine to be detected.

[0035] Optionally, the fault type can be blade crack, gearbox wear, bearing damage, generator short circuit, etc.

[0036] Optionally, the fault level can be set according to the actual situation, such as: minor, moderate, severe, etc.

[0037] S103. Select the target control scheme from the candidate control schemes based on the fault type and fault level.

[0038] In the application embodiment, after obtaining the fault type and fault level, the target control scheme can be selected from the candidate control schemes according to the fault type and fault level.

[0039] Optionally, candidate control schemes include control schemes based on fuzzy logic, control schemes based on genetic algorithms, and control schemes based on reinforcement learning.

[0040] Optionally, a mapping database between fault types and fault levels and control schemes can be pre-established. The mapping database includes control schemes based on fuzzy logic, control schemes based on genetic algorithms, and control schemes based on reinforcement learning. The target control scheme is determined by querying the mapping database according to the fault type and fault level.

[0041] It should be noted that by establishing a mapping database between fault types and fault levels and control schemes, the mapping database can match the optimal control scheme according to different fault types and fault levels, so as to give full play to the advantages of each control scheme.

[0042] For example, if the fault type of the wind turbine to be tested is A and the fault degree is a, the target control scheme can be determined as A by querying the mapping relationship database.

[0043] S104. Based on the target control scheme, determine the target control parameters and target control strategy of the wind turbine to be tested.

[0044] Among them, control parameters refer to various parameters that affect the operating efficiency and lifespan of the wind turbine unit under test.

[0045] For example, control parameters can be mechanical parameters such as blade angle, generator speed, transmission gear ratio, pitch blade length, and counterweight position; and electrical parameters such as generator axial clearance, winding current, excitation current, and grid connection parameters.

[0046] Among them, the control strategy refers to the reasonable control strategy selected based on the real-time status parameters of the wind turbine to be tested.

[0047] For example, the control strategy could be to select the blade overspeed protection strategy when the wind speed is too high, and reduce the energy taken up by the wind turbine by adjusting the blade angle; and select the power limiting output strategy when the grid demand is low, and control the power generation by adjusting the generator terminal electrical parameters.

[0048] For example, when the target control scheme is a fuzzy logic-based control scheme, fuzzy logic can be used to perform fuzzy reasoning on the operating state and environmental information of the wind turbine under test to obtain the target control parameters and target control strategy; when the target control scheme is a genetic algorithm-based control scheme, the genetic algorithm can be used to optimize and search for the control parameters and control strategy of the wind turbine under test to obtain the target control parameters and target control strategy; when the target control scheme is a reinforcement learning-based control scheme, reinforcement learning can be used to learn and adjust the control parameters and control strategy of the wind turbine under test online to ensure that the wind turbine under test can obtain the target control parameters and target control strategy of the wind turbine under test according to different state parameters and environmental parameters.

[0049] The fault diagnosis and control method for wind turbines proposed in this application obtains the state parameters and environmental parameters of the wind turbine under test. Based on these parameters, it determines the fault type and fault level of the wind turbine. Then, based on the fault type and fault level, it selects a target control scheme from candidate control schemes. Finally, based on the target control scheme, it determines the target control parameters and target control strategy for the wind turbine under test. This application improves the accuracy and efficiency of fault diagnosis for wind turbines. Furthermore, it allows for the selection of corresponding target control schemes based on different fault types and fault levels, dynamically adjusting the control parameters and strategies of the wind turbine to determine the target control parameters and strategy, thereby optimizing the operating efficiency and reliability of the wind turbine, reducing wear and tear, and extending its lifespan.

[0050] In the above embodiments, the specific process of obtaining the fault type and fault level of the wind turbine under test based on state parameters and environmental parameters can be described in conjunction with... Figure 2 To understand further, Figure 2 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 2 As shown, the method includes:

[0051] S201, Obtain the trained target fault diagnosis model.

[0052] It should be noted that the target fault diagnosis model based on convolutional neural network (CNN) proposed in this application is built on the basis of traditional convolutional neural network LeNet-5, adding convolutional layers, pooling layers and fully connected layers, and using the rectified linear unit (ReLU) function as the activation function.

[0053] The training process of the target fault diagnosis model will be explained below.

[0054] In this embodiment, normal and fault signals of the wind turbine under different operating conditions can be obtained. The normal and fault signals are randomly divided into training and test sets. The signals in the training set are input into the fault diagnosis model to be trained, and the cross-entropy loss function and stochastic gradient descent algorithm are used to iteratively train the fault diagnosis model to obtain a trained fault diagnosis model. The signals in the test set are input into the trained fault diagnosis model. In response to the trained fault diagnosis model meeting the training termination condition, the trained fault diagnosis model is determined as the target fault diagnosis model.

[0055] Optionally, metrics such as accuracy and confusion matrix can be used to evaluate the performance of the target fault diagnosis model.

[0056] S202, input the state parameters and environmental parameters into the target fault diagnosis model, and output the fault type and fault level of the wind turbine to be tested.

[0057] It should be noted that the target fault diagnosis model has the following structure: The first layer is the input layer, receiving a one-dimensional signal (state parameters and environmental parameters) of length 2048 as input. The second layer is a convolutional layer, using 16 one-dimensional convolutional kernels of size 64, stride 16, and padding 24 to convolve the input signal, resulting in 16 feature maps of length 128. The third layer is a batch normalization layer, which normalizes the feature maps output from the second layer to improve the model's stability and convergence speed. The fourth layer is an activation layer, using the ReLU function to perform a non-linear transformation on the feature maps output from the third layer to enhance the model's expressive power. The fifth layer is a pooling layer, using max pooling with a size of 2 and a stride of 2. The fourth layer's output feature map is downsampled using pooling, resulting in 16 feature maps of length 64. Layers six through ten consist of convolutional layers, batch normalization layers, activation layers, pooling layers, and another convolutional layer, respectively. The sixth layer uses 32 one-dimensional convolutional kernels of size 3, stride 1, and padding 1 to convolve the fifth layer's output feature map, resulting in 32 feature maps of length 64. These feature maps are then subjected to batch normalization, ReLU activation, and max pooling, resulting in 32 feature maps of length 32. Finally, 64 one-dimensional convolutional kernels of size 3, stride 1, and padding 1 are used to further refine these feature maps. The feature maps are convolved to obtain 64 feature maps of length 32. Layers 11 to 15 are batch normalization layers, activation layers, pooling layers, convolutional layers, and batch normalization layers, respectively. Layer 11 performs batch normalization, ReLU activation, and max pooling on the feature maps output from layer 10, resulting in 64 feature maps of length 16. These feature maps are then convolved with 64 one-dimensional convolutional kernels of size 3 and stride 1, resulting in 64 feature maps of length 14. Finally, batch normalization is performed on these feature maps. Layers 16 to 18 are activation layers, pooling layers, and fully connected layers, respectively. The sixteenth layer performs ReLU activation and max pooling operations on the feature map output from the fifteenth layer, resulting in 64 feature maps of length 7. These feature maps are flattened into a vector of length 192 and input into a fully connected layer with 100 neurons, resulting in a vector of length 100. The nineteenth layer is a fully connected layer that inputs the vector output from the eighteenth layer into a fully connected layer with 10 neurons, resulting in a vector of length 10. This gives the fault type and fault level of the wind turbine.

[0058] It should be noted that, in order to improve the accuracy of determining the fault type and fault level of the wind turbine under test, this application performs cluster analysis on the wind turbine set to determine the final fault type and fault level of the wind turbine under test.

[0059] In the above embodiments, the specific process for determining the final fault type and fault level of the wind turbine under test can be combined with... Figure 3 To understand further, Figure 3 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 3 As shown, the method includes:

[0060] S301 performs cluster analysis on the wind turbine set, dividing the wind turbine set into multiple subsets based on the state parameters and environmental parameters of each wind turbine set.

[0061] The wind turbine set includes multiple wind turbine units.

[0062] Alternatively, clustering algorithms, such as the K-means algorithm, can be used to perform cluster analysis on multiple wind turbine units, dividing the set of wind turbine units into multiple subsets based on the state parameters and environmental parameters of each wind turbine unit.

[0063] For example, based on wind speed, the set of wind turbines can be divided into low-wind-speed subsets, medium-wind-speed subsets, and high-wind-speed subsets according to the wind speed level and range of variation; based on temperature, the set of wind turbines can be divided into low-temperature subsets and high-temperature subsets; based on air pressure, the set of wind turbines can be divided into high-pressure subsets and flat-pressure subsets.

[0064] It should be noted that dividing the wind turbine set into multiple subsets based on the state and environmental parameters of each wind turbine can make the states and environments of wind turbines located in the same subset similar, laying the foundation for subsequent training of the corresponding fault diagnosis model.

[0065] S302, for each subset of wind turbines, train the corresponding target fault diagnosis model.

[0066] It should be noted that after dividing the wind turbine set into multiple subsets, a corresponding target fault diagnosis model can be trained for the wind turbines in each subset.

[0067] It should be noted that the specific process for training the target fault diagnosis model can be found in the above embodiments, and will not be repeated here.

[0068] For example, for the low-temperature subset, the corresponding target fault diagnosis model is trained as M1; for the high-temperature subset, the corresponding target fault diagnosis model is trained as M2; for the high-atomic set, the corresponding target fault diagnosis model is trained as M3; and for the plain subset, the corresponding target fault diagnosis model is trained as M4.

[0069] S303: Input the state parameters and environmental parameters of the wind turbine to be tested into all the trained target fault diagnosis models to obtain the fault type and fault level output by each target fault diagnosis model.

[0070] It should be noted that after training the corresponding target fault diagnosis model for each subset of wind turbines, all trained target fault diagnosis models can be obtained. The state parameters and environmental parameters of the wind turbine to be tested can be input into all trained target fault diagnosis models to obtain the fault type and fault level output by each target fault diagnosis model.

[0071] S304 uses an ensemble learning method to process the fault type and fault level output by each target fault diagnosis model to obtain the final fault type and fault level of the wind turbine to be tested.

[0072] Optionally, an ensemble learning method, such as a voting method, can be used to comprehensively judge the fault type and fault level output by each target fault diagnosis model to obtain the final fault type and fault level of the wind turbine to be tested.

[0073] The fault diagnosis and control method for wind turbines proposed in this application trains a target fault diagnosis model using deep learning technology. State and environmental parameters are input into the target fault diagnosis model, which outputs the fault type and fault level of the wind turbine under test. Simultaneously, through an ensemble learning method, the fault types and fault levels output by all target fault diagnosis models are comprehensively judged to obtain the final fault type and fault level of the wind turbine under test. This improves the accuracy and efficiency of fault diagnosis for wind turbines and lays the foundation for subsequently determining the target control parameters and target control strategies for the wind turbine under test.

[0074] In this embodiment of the application, after obtaining the fault type and fault level of the wind turbine to be tested, a target control scheme can be adaptively selected according to the fault type and fault level, and the target control parameters and target control strategy of the wind turbine to be tested can be determined based on the target control scheme.

[0075] The following section explains the specific process by which the target control parameters and target control strategy of the wind turbine to be tested are determined based on the target control scheme proposed in this application.

[0076] In the above embodiments, if the target control scheme is a fuzzy logic-based control scheme, the specific process of determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme can be combined with... Figure 4 To understand further, Figure 4 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 4 As shown, the method includes:

[0077] S401, determine the input and output variables of the wind turbine in the fuzzy inference process, and set the corresponding fuzzy set and membership function for each variable.

[0078] It should be noted that the input variables can be wind speed, rotational speed, power, etc., and the output variable is the blade angle.

[0079] It should be noted that a corresponding fuzzy set and membership function can be set for each variable, such as high, medium, and low.

[0080] For example, regarding wind speed, three fuzzy sets can be defined: 0-3 m / s for low wind speed, 2-5 m / s for medium wind speed, and 4-8 m / s for high wind speed.

[0081] It should be noted that "high", "medium" and "low" in the membership function represent the membership relationship of fuzzy sets, that is, the degree to which a specific value belongs to different fuzzy sets.

[0082] For example, regarding wind speed, when the wind speed is 2.5 m / s, the membership degree of belonging to the "low wind speed set" may be 0.2, the membership degree of belonging to the "medium wind speed set" may be 0.8, and the membership degree of belonging to the "high wind speed set" may be 0.

[0083] S402, input the real-time input variable values ​​of the wind turbine to be tested into the fuzzy rule base, perform fuzzy inference, and obtain the corresponding output variable values.

[0084] It should be noted that fuzzy rules for wind turbines can be preset, such as "if the wind speed is high and the rotational speed is low, then increase the blade angle", and the preset fuzzy rules can be stored in the fuzzy rule library.

[0085] In this embodiment of the application, by inputting the real-time input variable values ​​of the wind turbine to be tested into the fuzzy rule base, fuzzy reasoning can be performed to obtain the corresponding output variable values.

[0086] For example, when the current wind speed is 3 m / s and the rotation speed is 10 rpm, according to the pre-set fuzzy rule of the wind turbine, "if the wind speed is low and the rotation speed is low, then slightly increase the blade angle", through fuzzy reasoning, it can be obtained that the increment of the blade angle is 2 degrees.

[0087] S403 defuzzifies the output variable values ​​to determine the target control parameters and target control strategy of the wind turbine to be tested.

[0088] Optionally, the fuzzy inference results can be summarized to obtain the membership function of the output variable value, and the centroid of the membership function can be calculated as the final output value to obtain the target control parameters and target control strategy of the wind turbine to be tested.

[0089] In the above embodiments, if the target control scheme is a control scheme based on a genetic algorithm, the specific process of determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme can be combined with... Figure 5 To understand further, Figure 5 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 5 As shown, the method includes:

[0090] S501 encodes the control parameters and control strategies of the wind turbine and obtains a binary string, where the binary string is treated as an individual.

[0091] It should be noted that the control parameters and control strategies of wind turbine units can be encoded into a binary string, as an individual.

[0092] S502 generates N individuals randomly to construct an initial population and calculates the fitness value of each individual in the initial population, where N is a positive integer.

[0093] S503, based on fitness values, selects each individual and forms a new population through crossover and mutation until a preset termination condition is reached, thereby determining the target control parameters and target control strategy of the wind turbine to be tested.

[0094] It should be noted that after obtaining the fitness value, selection can be made based on the fitness value, retaining excellent individuals and eliminating inferior individuals, and forming a new population through crossover and mutation. The above steps are repeated until the preset termination condition is reached, and the optimal individual, i.e. the target control parameters and target control strategy, is output.

[0095] It should be noted that this application does not limit the setting of preset termination conditions, which can be set according to actual conditions. Optionally, the termination condition can be the maximum number of iterations, the accuracy of the optimal solution, etc.

[0096] In the above embodiments, if the target control scheme is a reinforcement learning-based control scheme, the specific process of determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme can be combined with... Figure 6 To understand further, Figure 6 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 6 As shown, the method includes:

[0097] S601, pre-sets the state parameters of the wind turbine's state space, the action parameters of its action space, and the reward function.

[0098] It should be noted that the wind turbine can be set in a state space and an action space. The state parameters in the state space can include wind speed, rotational speed, power, etc., while the action parameters in the action space can include blade angle, braking force, etc.

[0099] S602 obtains the reward value corresponding to each state parameter-action parameter through the reward function.

[0100] It should be noted that, through the pre-set reward function of the wind turbine, the reward value corresponding to each state parameter-action parameter can be given according to the wind turbine's operating efficiency and reliability indicators.

[0101] S603 selects target action parameters based on the current state parameters and reward value through a pre-set strategy function for the wind turbine.

[0102] It should be noted that by pre-setting the strategy function of the wind turbine, the optimal or random action can be selected based on the current state parameters and reward value in order to explore or utilize environmental information.

[0103] S604, based on the current reward value and the future reward value, obtains the value corresponding to each state parameter-action parameter through a pre-set value function of the wind turbine.

[0104] S605, using a reinforcement learning algorithm, the policy function and the value function are iteratively updated to obtain the target policy function and the target value function.

[0105] For example, reinforcement learning algorithms such as Q-learning and SARSA can be used to learn and update the policy function and value function online, thereby obtaining the target policy function and target value function.

[0106] S606, based on the target strategy function and the target value function, determine the target control parameters and target control strategy of the wind turbine to be tested.

[0107] It should be noted that after obtaining the target strategy function and the target value function, the wind turbine can make the best control decision based on different state parameters and environmental parameters, so as to determine the target control parameters and the target control strategy.

[0108] The fault diagnosis and control method for wind turbines proposed in this application, after determining the fault type and fault level of the wind turbine under test, can determine a target control scheme that matches the fault type and fault level, dynamically adjust the control parameters and control strategies of the wind turbine under test to obtain the target control parameters and target control strategies of the wind turbine under test, and control the wind turbine through the target control parameters and target control strategies, thereby improving the operating efficiency and reliability of the wind turbine, reducing the wear and loss of the wind turbine, and extending the service life of the wind turbine.

[0109] In this embodiment of the application, after obtaining the fault diagnosis data (fault type and fault level), target control scheme and target control strategy of the wind turbine, the above data can be stored and shared.

[0110] In the above embodiments, the specific process after determining the target control parameters and target control strategy of the wind turbine to be tested can be described in conjunction with... Figure 7 To understand further, Figure 7 This is a flowchart illustrating a fault diagnosis and control method for a wind turbine generator according to another embodiment of this application, as shown below. Figure 7 As shown, the method includes:

[0111] S701 encrypts, hashes, and signs the fault detection data and experience data of the wind turbine to be tested, forming a data block.

[0112] It should be noted that fault diagnosis data and experience data include both quantitative operating parameters and signal data, as well as qualitative experience knowledge, fault diagnosis models, and maintenance plans.

[0113] For example, fault detection data may include the operating data of the wind turbine under test (wind speed, power generation, speed and temperature, etc.), fault data (alarm information, fault codes and maintenance records, etc.), wind farm environmental data (temperature, air pressure, humidity, etc.), and experience data may include fault modes, fault cause analysis, maintenance suggestions, fault diagnosis models (model parameters of fault diagnosis models trained based on deep learning), maintenance plans and maintenance results (handling procedures and maintenance effect evaluation for various faults).

[0114] It should be noted that the fault detection data and experience data of the wind turbine under test are encrypted, hashed, and signed to form a data block.

[0115] S702 sends the data block to the blockchain network and verifies the data block.

[0116] Optionally, data blocks can be sent to a blockchain network where they can be verified and consensus-based among multiple nodes to determine their validity and order.

[0117] Blockchain technology is a distributed ledger technology that uses cryptographic principles to encrypt, verify, store, and share data. It has the following characteristics: decentralization—it does not rely on any centralized institution or platform; data is maintained and updated jointly by multiple nodes; immutability—once data is written into the blockchain, it cannot be modified or deleted, ensuring data integrity and authenticity; traceability—data is recorded and linked in chronological order in the blockchain, forming an indivisible data chain that allows tracing the source and changes of the data; and smart contracts—a blockchain-based automated execution protocol that can complete transactions or tasks without human intervention based on preset conditions and rules.

[0118] S703 stores verified data blocks in the blockchain, and uses smart contracts on the blockchain to trade fault detection data and experience data, and to remotely monitor and maintain wind turbine units.

[0119] It should be noted that storing verified data blocks in the blockchain can form an immutable data chain, enabling distributed storage and sharing of data. Through smart contracts, wind turbine fault diagnosis data and experience data can be valued, traded, and incentivized, realizing the circulation and co-creation of knowledge value. Furthermore, through smart contracts, wind turbines can be remotely monitored and maintained, enabling functions such as fault early warning, fault handling, and fault feedback.

[0120] It should be noted that this application utilizes blockchain technology to treat wind turbine fault diagnosis data and experience data as a resource, enabling encrypted, distributed storage and sharing within the blockchain network. This achieves decentralized management and circulation of resources. Simultaneously, smart contracts can be used to implement an incentive mechanism for wind turbine fault diagnosis knowledge. Relevant personnel in the wind turbine industry can jointly create, contribute, verify, evaluate, use, and improve wind turbine fault diagnosis resources, forming an open, collaborative, and innovative knowledge ecosystem. Furthermore, remote monitoring and maintenance of wind turbines can be achieved, with smart contracts automatically triggering corresponding control commands or maintenance tasks, thereby improving the operating efficiency and reliability of wind turbines.

[0121] The wind turbine fault diagnosis and control method proposed in this application, after obtaining the wind turbine fault diagnosis data, target control scheme and target control strategy, can use blockchain technology to realize the circulation and co-creation of wind turbine fault diagnosis resources, and realize remote monitoring and maintenance of wind turbines, thereby improving the quality and quantity of wind turbine fault diagnosis resources and improving the operating efficiency and reliability of wind turbines.

[0122] The specific process of the fault diagnosis and control method for wind turbines proposed in this application will be explained below.

[0123] This application uses a 1.5MW double-fed induction generator (DFIG) wind turbine as the wind turbine to be tested, which is operating in a mountain wind farm, as an example for explanation and illustration.

[0124] It should be noted that the wind turbine under test is equipped with internal temperature sensors, pressure sensors, power sensors, speed sensors, etc., and external wind speed sensors, wind direction sensors, air temperature sensors, air pressure sensors, etc. It can collect the status parameters and environmental parameters of the wind turbine under test in real time, and send the status parameters and environmental parameters to the cloud server through the wireless communication module.

[0125] It should be noted that when the wind turbine under test experiences gearbox wear failure, the temperature and pressure inside the gearbox increase, and the power and speed fluctuate, affecting the operating efficiency and reliability of the wind turbine under test. This failure is reflected in the collected status parameters and environmental parameters and sent to the cloud server.

[0126] It should be noted that the collected state parameters and environmental parameters are input into the target fault diagnosis model to obtain the fault type and fault level of the wind turbine under test. The target fault diagnosis model is trained with a large amount of normal and fault signal data, and can automatically learn the fault characteristics of the wind turbine without the need for manual setting of feature extraction and classification rules. The target fault diagnosis model identifies that the wind turbine under test has a gearbox wear fault and gives the corresponding fault level as medium.

[0127] Furthermore, based on the fault type and fault level, the selected adaptive control method is a target control scheme based on a genetic algorithm. The genetic algorithm is used to optimize and search for the control parameters and control strategy of the wind turbine, thereby obtaining the optimal control parameters and control strategy. The specific steps are as follows:

[0128] The control parameters and control strategies of the wind turbine are encoded into a binary string as an individual. The control parameters can be blade angle and braking force, and the control strategy can be starting braking and stopping braking. For example, an individual can be represented as "00101101", where "00" represents a blade angle of 0°, "10" represents a braking force of 50%, "11" represents starting braking, and "01" represents stopping braking. A certain number of individuals are randomly generated to form an initial population. Optionally, the population size can be set to 100, that is, 100 individuals are generated, and the fitness value of each individual is calculated.

[0129] Alternatively, the fitness value of an individual can be calculated using the following formula:

[0130] f = α × P - β × T - γ × B

[0131] Where f is the fitness value, P is the output power of the wind turbine, T is the gearbox temperature, B is the braking force, and α, β and γ are weighting coefficients used to adjust the weight of each performance index. They can be set to α = 0.8, β = 0.1 and γ = 0.1.

[0132] Furthermore, a roulette wheel selection method can be used, where the probability of each individual being selected is proportional to its fitness value. Superior individuals are selected and inferior ones are eliminated. In this example, a roulette wheel method is used for selection, and crossover and mutation probabilities are applied to generate new individuals and form a new population. The crossover probability can be set to 0.8, and the mutation probability to 0.01. Crossover refers to two individuals exchanging some genes to produce two new individuals. Mutation refers to an individual randomly changing a single gene position to produce a new individual. For example, for two individuals "00101101" and "11010010", a single-point crossover will produce two new individuals "00100010" and "11011101". A single-point mutation of "00100010" will produce a new individual "00100011".

[0133] Repeat the above steps until the preset termination condition is reached, such as the maximum number of iterations, the accuracy of the optimal solution, etc. Optionally, the preset termination condition can be that the maximum number of iterations is 100, that is, repeat the above steps 100 times, and output the optimal individual, that is, the target control parameters and the target control strategy. For example, the optimal individual is "01011011", that is, the target control parameters are the blade angle is 15° and the braking force is 75%, and the target control strategy is to start braking.

[0134] Furthermore, fault diagnosis data and experience data from wind turbines can be encrypted, stored in a distributed manner, and shared. Blockchain technology can be used to facilitate the circulation and co-creation of wind turbine fault diagnosis resources, as well as remote monitoring and maintenance of wind turbines. The specific steps are as follows:

[0135] Alternatively, the Ethereum platform can be used as the infrastructure of the blockchain network to issue and trade wind turbine fault diagnosis data and experience data as a non-fungible token (NFT). NFT is a digital asset based on blockchain technology, which has the characteristics of uniqueness, non-fungibility, and verifiability. It can be used to represent anything unique. Issuing and trading wind turbine fault diagnosis data and experience data as NFT can guarantee its intellectual property rights, knowledge traceability, and knowledge anti-counterfeiting.

[0136] By leveraging smart contracts, an incentive mechanism for wind turbine fault diagnosis knowledge can be implemented, encouraging relevant personnel to collaboratively create, contribute, verify, evaluate, use, and improve such knowledge. This fosters an open, collaborative, and innovative knowledge ecosystem. The Ethereum platform can be used to establish incentive mechanisms. For example, wind turbine maintenance personnel can receive token rewards for timely reporting of fault diagnosis and experience data, which is verified by other participants. Wind turbine experts and researchers can receive token rewards for providing effective fault diagnosis solutions and suggestions, which are evaluated by other participants. Other participants can receive token rewards for participating in the verification and evaluation of wind turbine fault diagnosis knowledge and providing valuable feedback and improvement suggestions. These tokens can be used to buy or sell other wind turbine fault diagnosis knowledge on the blockchain network or exchanged for fiat currency.

[0137] Through the above incentive mechanism, the value circulation and co-creation of wind turbine fault diagnosis knowledge can be promoted, the quality and quantity of wind turbine fault diagnosis knowledge can be improved, an open, collaborative and innovative knowledge ecosystem can be formed, and remote monitoring and maintenance of wind turbines can be realized by using blockchain technology. Through smart contracts, corresponding control commands or maintenance tasks can be automatically triggered, thereby improving the operating efficiency and reliability of wind turbines.

[0138] Optionally, smart contracts on the Ethereum platform can be used to set up the following monitoring and maintenance mechanisms. For example, for wind turbine maintenance personnel, if they receive control commands or maintenance tasks from a cloud server or blockchain network, they can confirm receipt through a smart contract and perform operations according to the commands or tasks. After the operation is completed, they can report the results through a smart contract and receive corresponding token rewards. For the monitoring system in the cloud server or blockchain network, if it detects that a wind turbine has malfunctioned or that control parameters and control strategies need to be adjusted, it can send control commands or maintenance tasks to the wind turbine maintenance personnel through a smart contract. For the wind turbine itself, if it receives control commands or maintenance tasks from a cloud server or blockchain network, it will automatically execute them through a smart contract and report the results. If it detects that a wind turbine has malfunctioned or that control parameters and control strategies need to be adjusted, it can automatically send a request to the monitoring system in the cloud server or blockchain network through a smart contract and make adjustments based on the feedback. Through these monitoring and maintenance mechanisms, remote monitoring and maintenance of wind turbines can be achieved, improving the operating efficiency and reliability of wind turbines.

[0139] In summary, the wind turbine fault diagnosis and control method proposed in this application, through deep learning technology, trains a target fault diagnosis model that can automatically learn the fault characteristics of wind turbines without requiring manual setting of feature extraction and classification rules. This improves the accuracy and efficiency of wind turbine fault diagnosis. It can dynamically adjust the control parameters and strategies of wind turbines according to different fault types and levels, optimizing the operating efficiency and reliability of wind turbines, reducing wear and tear, and extending their lifespan. Furthermore, based on blockchain technology, it enables the circulation and co-creation of wind turbine fault diagnosis resources and remote monitoring and maintenance of wind turbines, improving the quality and quantity of wind turbine fault diagnosis resources and ultimately enhancing the operating efficiency and reliability of wind turbines.

[0140] Figure 8 This is a schematic diagram of the structure of a fault diagnosis and control device for a wind turbine according to an embodiment of this application, as shown below. Figure 8 As shown, the wind turbine fault diagnosis and control device 800 includes a first acquisition module 81, a second acquisition module 82, a selection module 83, and a determination module 84, wherein:

[0141] The first acquisition module 81 is used to acquire the status parameters and environmental parameters of the wind turbine generator to be tested.

[0142] The second acquisition module 82 is used to acquire the fault type and fault level of the wind turbine to be tested based on the state parameters and the environmental parameters.

[0143] The selection module 83 is used to select a target control scheme from the candidate control schemes based on the fault type and fault level.

[0144] The determination module 84 is used to determine the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme.

[0145] The fault diagnosis and control device for wind turbine generators provided in the second aspect of this application also has the following technical features, including:

[0146] According to one embodiment of this application, the second acquisition module 82 is used to: acquire the trained target fault diagnosis model; input the state parameters and the environmental parameters into the target fault diagnosis model, and output the fault type and fault level of the wind turbine to be tested.

[0147] According to one embodiment of this application, the second acquisition module 82 is configured to: acquire normal signals and fault signals of the wind turbine under different operating conditions; randomly divide the normal signals and the fault signals into a training set and a test set; input the signals in the training set into a fault diagnosis model to be trained, and iteratively train the fault diagnosis model using a cross-entropy loss function and a stochastic gradient descent algorithm to obtain a trained fault diagnosis model; input the signals in the test set into the trained fault diagnosis model; and, in response to the trained fault diagnosis model satisfying the training termination condition, determine the trained fault diagnosis model as the target fault diagnosis model.

[0148] According to one embodiment of this application, the second acquisition module 82 is configured to: perform cluster analysis on the wind turbine set, and divide the wind turbine set into multiple subsets according to the state parameters and environmental parameters of each wind turbine set; train a corresponding target fault diagnosis model for the wind turbine sets in each subset; input the state parameters and environmental parameters of the wind turbine set to be tested into all the trained target fault diagnosis models to obtain the fault type and fault level output by each target fault diagnosis model; and process the fault type and fault level output by each target fault diagnosis model through an ensemble learning method to obtain the final fault type and fault level of the wind turbine set to be tested.

[0149] According to one embodiment of this application, module 83 is selected for: pre-establishing a mapping relationship database between fault types and fault levels and control schemes, wherein the mapping relationship database includes control schemes based on fuzzy logic, control schemes based on genetic algorithms, and control schemes based on reinforcement learning; and querying the mapping relationship database according to the fault type and fault level to determine the target control scheme.

[0150] According to one embodiment of this application, the determining module 84 is used to: determine the input variables and output variables of the wind turbine during the fuzzy inference process, and set a corresponding fuzzy set and membership function for each variable; input the real-time input variable values ​​of the wind turbine to be detected into the fuzzy rule base, perform fuzzy inference, and obtain the corresponding output variable values; defuzzify the output variable values, and determine the target control parameters and target control strategy of the wind turbine to be detected.

[0151] According to one embodiment of this application, the determining module 84 is configured to: encode the control parameters and control strategies of the wind turbine to obtain a binary string, wherein the binary string is considered as an individual; generate N individuals randomly to construct an initial population, calculate the fitness value of each individual in the initial population, where N is a positive integer; screen each individual based on the fitness value, and form a new population through crossover and mutation until a preset termination condition is reached, thereby determining the target control parameters and target control strategy of the wind turbine to be tested.

[0152] According to one embodiment of this application, a determining module 84 is configured to: pre-set state parameters of the state space of the wind turbine, action parameters of the action space, and a reward function; obtain a reward value corresponding to each state parameter-action parameter through the reward function; select a target action parameter based on the current state parameter and reward value, using a pre-set strategy function of the wind turbine; obtain a value corresponding to each state parameter-action parameter based on the current reward value and future reward value, using a pre-set value function of the wind turbine; iteratively update the strategy function and the value function using a reinforcement learning algorithm to obtain a target strategy function and a target value function; and determine the target control parameters and target control strategy of the wind turbine to be detected based on the target strategy function and the target value function.

[0153] According to one embodiment of this application, the apparatus 800 is used to: encrypt, hash, and sign fault detection data and experience data of the wind turbine to be tested to form a data block; send the data block to a blockchain network and verify the data block; store the verified data block in the blockchain; and conduct transactions on the fault detection data and experience data through a smart contract on the blockchain, and remotely monitor and maintain the wind turbine.

[0154] The fault diagnosis and control device for wind turbines proposed in this application acquires the state parameters and environmental parameters of the wind turbine under test. Based on these parameters, it determines the fault type and fault level of the wind turbine. Then, based on the fault type and fault level, it selects a target control scheme from candidate control schemes. Finally, based on the target control scheme, it determines the target control parameters and target control strategy for the wind turbine under test. This application improves the accuracy and efficiency of fault diagnosis for wind turbines. Furthermore, it allows for the selection of corresponding target control schemes based on different fault types and fault levels, dynamically adjusting the control parameters and strategies of the wind turbine to determine the target control parameters and strategy, thereby optimizing the operating efficiency and reliability of the wind turbine, reducing wear and tear, and extending its lifespan.

[0155] To achieve the above embodiments, this application also provides an electronic device, a computer-readable storage medium, and a computer program product.

[0156] Figure 9 This is a block diagram of an electronic device according to an embodiment of this application, such as... Figure 9 As shown, device 1000 includes memory 101, processor 102, and a computer program stored in memory 101 and executable on processor 102. When processor 102 executes program instructions, it implements the execution of... Figures 1 to 7 An example of a fault diagnosis and control method for wind turbine units.

[0157] To implement the above embodiments, this application also provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute... Figures 1 to 7 The embodiment of the method for fault diagnosis and control of wind turbine units.

[0158] To implement the above embodiments, this application also provides a computer program product that, when the instruction processor in the computer program product is executed, performs... Figures 1 to 7 The embodiment of the method for fault diagnosis and control of wind turbine units.

[0159] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0160] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0161] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0162] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0163] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0164] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0165] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0166] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for fault diagnosis and control of wind turbine generators, characterized in that, The method includes: Obtain the status and environmental parameters of the wind turbine generator under test; Based on the state parameters and the environmental parameters, the fault type and fault level of the wind turbine under test are obtained; Based on the fault type and the fault level, a target control scheme is selected from the candidate control schemes; Based on the target control scheme, the target control parameters and target control strategy of the wind turbine to be tested are determined; The step of obtaining the fault type and fault level of the wind turbine under test based on the state parameters and the environmental parameters includes: Obtain the trained target fault diagnosis model; The state parameters and environmental parameters are input into the target fault diagnosis model, and the fault type and fault level of the wind turbine under test are output. Cluster analysis is performed on the set of wind turbine units to divide the set of wind turbine units into multiple subsets based on the state parameters and environmental parameters of each wind turbine unit; For each subset of wind turbines, a corresponding target fault diagnosis model is trained; The state parameters and environmental parameters of the wind turbine to be tested are input into all the trained target fault diagnosis models to obtain the fault type and fault level output by each target fault diagnosis model. By using an ensemble learning method, the fault type and fault level output by each target fault diagnosis model are processed to obtain the final fault type and fault level of the wind turbine under test. The step of selecting a target control scheme from candidate control schemes based on the fault type and fault level includes: A database of mapping relationships between fault types and fault levels and control schemes is pre-established, wherein the mapping relationship database includes control schemes based on fuzzy logic, control schemes based on genetic algorithms, and control schemes based on reinforcement learning; Based on the fault type and fault level, the mapping relationship database is queried to determine the target control scheme.

2. The method according to claim 1, characterized in that, The process of acquiring the trained target fault diagnosis model includes: Acquire normal and fault signals of wind turbine under different operating conditions, and randomly divide the normal and fault signals into training and test sets. The signals in the training set are input into the fault diagnosis model to be trained, and the cross-entropy loss function and stochastic gradient descent algorithm are used to iteratively train the fault diagnosis model to obtain the trained fault diagnosis model. The signals in the test set are input into the trained fault diagnosis model. In response to the trained fault diagnosis model meeting the training termination condition, the trained fault diagnosis model is determined as the target fault diagnosis model.

3. The method according to claim 2, characterized in that, If the target control scheme is a fuzzy logic-based control scheme, determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme includes: Determine the input and output variables of the wind turbine generator in the fuzzy inference process, and set the corresponding fuzzy set and membership function for each variable; The real-time input variable values ​​of the wind turbine to be tested are input into the fuzzy rule base, fuzzy reasoning is performed, and the corresponding output variable values ​​are obtained. The output variable values ​​are defuzzified to determine the target control parameters and target control strategy of the wind turbine to be tested.

4. The method according to claim 2, characterized in that, If the target control scheme is a control scheme based on a genetic algorithm, the step of determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme includes: The control parameters and control strategies of the wind turbine are encoded to obtain binary strings, wherein the binary strings are treated as individual entities; N individuals are generated randomly to construct an initial population. The fitness value of each individual in the initial population is calculated, where N is a positive integer. Based on the fitness value, each individual is screened, and a new population is formed through crossover and mutation until a preset termination condition is reached, thereby determining the target control parameters and target control strategy of the wind turbine to be tested.

5. The method according to claim 2, characterized in that, If the target control scheme is a reinforcement learning-based control scheme, determining the target control parameters and target control strategy of the wind turbine to be tested according to the target control scheme includes: Pre-set the state parameters of the wind turbine's state space, the action parameters of its action space, and the reward function. The reward function is used to obtain the reward value corresponding to each state parameter-action parameter. Based on the current state parameters and reward value, the target action parameters are selected through the pre-set strategy function of the wind turbine. Based on the current reward value and the future reward value, the value corresponding to each state parameter-action parameter is obtained through a pre-set value function of the wind turbine. By using a reinforcement learning algorithm, the policy function and the value function are iteratively updated to obtain the target policy function and the target value function; Based on the target strategy function and the target value function, the target control parameters and target control strategy of the wind turbine to be tested are determined.

6. The method according to claim 1, characterized in that, The method further includes: The fault detection data and experience data of the wind turbine under test are encrypted, hashed and signed to form a data block; The data block is sent to the blockchain network and verified. The verified data blocks are stored in the blockchain, and the fault detection data and experience data are traded through smart contracts on the blockchain, enabling remote monitoring and maintenance of the wind turbine units.

7. A fault diagnosis and control device for wind turbine generators, characterized in that, The apparatus implements the method as described in claim 1, the apparatus comprising: The first acquisition module is used to acquire the status parameters and environmental parameters of the wind turbine generator to be tested. The second acquisition module is used to acquire the fault type and fault level of the wind turbine to be tested based on the state parameters and the environmental parameters. The selection module is used to select a target control scheme from the candidate control schemes based on the fault type and fault level. The determination module is used to determine the target control parameters and target control strategy of the wind turbine to be tested based on the target control scheme.

Images