Wind turbine generator operation and maintenance management method and system, equipment and medium

Abstract

The invention relates to the technical field of wind turbine generator data processing, in particular to a wind turbine generator operation and maintenance management method and system, equipment and a medium, and the method mainly comprises the steps that a result of whether a current fan is abnormal or not is output through an optimal wind turbine generator fault prediction model and to-be-detected fan operation data; if abnormity occurs, a data evaluation model is established, the data evaluation model is used for evaluating the floating degree of the data, evaluation values of the multiple groups of types of operation data are calculated through the data evaluation model, and the multiple groups of types of operation data are sorted according to the evaluation values. According to the method, when the wind turbine generator data is abnormal, the subsequent evaluation result is provided, which group of types of data fluctuate more currently is judged according to the floating degree of the data, and the problem or equipment corresponding to the data with the high floating degree is processed preferentially according to the sorting result, so that the working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine data processing, and in particular, to a method, system, device, and medium for operation and maintenance management of wind turbines. Background Art

[0002] With the continuous growth of global energy demand and the increasing awareness of environmental protection, the new energy industry has become a strategic industry promoting the economic development of our country. As an important part of wind power generation, the construction scale and efficiency of wind farms directly affect the output of wind power generation. The operation and maintenance (O&M) problems have thus become prominent. The O&M of wind turbines is an important link to ensure the efficient and stable operation of wind power generation systems, which covers multiple aspects such as daily monitoring, fault diagnosis, regular maintenance, and fault repair.

[0003] However, the current O&M systems generally only perform fault diagnosis on wind turbines and then output the diagnosis results, without analyzing the diagnosis results. When abnormalities or faults occur, how to handle them to achieve better effects. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, system, device, and medium for operation and maintenance management of wind turbines to solve the above problems in the prior art.

[0005] The embodiments of the present invention are realized through the following technical solutions:

[0006] A method for operation and maintenance management of wind turbines includes:

[0007] Set a collection period, obtain initial real-time wind turbine operation data within a collection period, preprocess the initial real-time wind turbine operation data, and output the wind turbine operation data to be detected;

[0008] Obtain the historical operation data of the wind turbine, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through the historical operation data of the wind turbine, and output the optimal wind turbine fault prediction model;

[0009] Output the result of whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the wind turbine operation data to be detected;

[0010] If no abnormality occurs, output the result of normal detection. If an abnormality occurs, obtain the wind turbine operation data to be detected, group it according to the data type, and obtain several groups of operation data of different types;

[0011] Establish a data evaluation model, which is used to evaluate the fluctuation degree of data. Normalize several groups of operation data of different types, calculate the evaluation values of several groups of operation data of different types respectively through the data evaluation model, and sort several groups of operation data of different types according to the size of the evaluation values, and output the sorting result.

[0012] Preferably, the preprocessing of the initial real-time fan operation data includes:

[0013] Identify whether there are missing values in the initial real-time fan operation data. If there are missing values, establish an interpolation function model and fill in the missing values through the interpolation function model;

[0014] If there are no missing values, directly output it as the fan operation data to be detected.

[0015] Preferably, it also includes obtaining a first value and a second value, including:

[0016] Obtain two values adjacent to the missing value, and judge whether the two adjacent values are normal values. If they are normal values, record the two values as the first value and the second value respectively;

[0017] If there is a non-normal value, obtain the next value adjacent to the non-normal value and judge whether the next adjacent value is a normal value until a normal value is obtained, and save the normal value as the first value or the second value;

[0018] Whether it is a normal value includes calculating the average value of the same type of data, setting a difference threshold, and judging whether the difference between two adjacent values and the average value is within the difference threshold. If so, it is judged as a normal value, and if not, it is judged as an abnormal value.

[0019] Preferably, the interpolation function model includes:

[0020]

[0021] In the formula, X z Replacement interpolation of the missing value, a1 is the first value, and a2 is the second value.

[0022] Preferably, establishing the wind turbine fault prediction model includes setting an input layer, a convolutional layer, an activation function layer, a pooling layer, a full average pooling layer, a fully connected layer, and an output layer;

[0023] The input layer is used to transfer sample data and extract features. The convolutional layer is used to perform convolutional operations on the data. The activation function layer is used to simulate the transmission process of the nervous system. The pooling layer is used to reduce the dimension of the output of the convolutional layer. The fully connected layer is used to classify the data. The output layer is used to output the diagnosis result.

[0024] Preferably, the convolutional layer includes:

[0025]

[0026] Where x i+1 is the output of the i-th convolutional layer, and x i is the input of the i-th convolutional layer, W i is the convolutional kernel, b i is the bias term, is the convolutional operation;

[0027] The activation function layer includes;

[0028]

[0029] Where f(x) is the activation function and x is the input value of the activation function;

[0030] The pooling layer includes;

[0031]

[0032] Where p l(i,i) is the output from the i-th pooling layer to the j-th fully connected layer, and x l(i,t) is the input from the t-th activation function layer to the i-th pooling layer, W is the regularization coefficient, and j and t are natural numbers;

[0033] The fully connected layer includes:

[0034]

[0035] Where Z l+1(j) is the output of the j-th neuron in the (l + 1)-th layer, and x l(i) is the output of the i-th neuron in the previous layer, is the weight of the connection between the j-th neuron in the (l + 1)-th layer and the i-th neuron in the previous layer, and O j is the probability corresponding to the output of the j-th neuron in the (l + 1)-th layer, where l and k are natural numbers, is the bias term of the j-th neuron in the (l + 1)-th layer.

[0036] Preferably, the establishment of the data evaluation model includes:

[0037]

[0038] Where N is the total number of data of this group of types, f i is the i-th data of this group of types of data, and R f is the evaluation value.

[0039] Second aspect, the present invention provides a wind turbine operation and maintenance management system, including:

[0040] A data preprocessing module, configured to set a collection period, obtain initial real-time wind turbine operation data within a collection period, preprocess the initial real-time wind turbine operation data, and output the wind turbine operation data to be detected;

[0041] A prediction model module, configured to obtain historical wind turbine operation data, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through the historical wind turbine operation data, output an optimal wind turbine fault prediction model, and output a result indicating whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the wind turbine operation data to be detected;

[0042] A result output module, configured to, if no abnormality occurs, output a result indicating normal detection, and if an abnormality occurs, obtain the wind turbine operation data to be detected, group it according to data types, and obtain operation data of several types; establish a data evaluation model, where the data evaluation model is used to evaluate the degree of data fluctuation, calculate the evaluation values of the operation data of several types through the data evaluation model respectively, and sort the operation data of several types according to the magnitude of the evaluation values, and output the sorting result.

[0043] A main control module, connected to the data preprocessing module, the prediction model module, and the result output module, and used to execute the above-mentioned wind turbine operation and maintenance management method.

[0044] Third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, it implements the above-mentioned wind turbine operation and maintenance management method.

[0045] A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the above-mentioned wind turbine operation and maintenance management method.

[0046] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0047] The method provided by the present invention mainly includes outputting the result of whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the operation data of the wind turbine to be detected. If an abnormality occurs, a data evaluation model is established. The data evaluation model is used to evaluate the degree of data fluctuation. The evaluation values of several types of operation data are calculated respectively through the data evaluation model, and several types of operation data are sorted according to the size of the evaluation values. Through the above method, when the wind turbine data is abnormal, subsequent evaluation results are provided. By judging the degree of data fluctuation, it is determined which type of current data fluctuates more. Through the sorting results, the problems or equipment corresponding to the data with a large degree of fluctuation are preferentially processed to improve work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown in the accompanying drawings here can be arranged and designed in various different configurations.

[0051] The division of modules in this application is a logical division. In actual implementation, there may be other division methods. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0052] The independently described modules or sub-modules may be physically separated or not: they can be implemented in software or in hardware. And some modules or sub-modules can be implemented through software, and the functions of these modules or sub-modules are called by the processor through the software, while other parts of the templates or sub-modules are implemented through hardware, such as through hardware circuits. In addition, some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this application.

[0053] Please refer to Figure 1 , a wind turbine operation and maintenance management method provided by the present invention includes:

[0054] S101: Set the acquisition period, obtain the initial real-time wind turbine operation data within one acquisition period, preprocess the initial real-time wind turbine operation data, and output the wind turbine operation data to be detected;

[0055] Among them, the setting of the period can be adjusted according to specific circumstances. The wind turbine operation data includes various categories, such as blade speed, temperature, vibration and other parameters. Missing values in the operation data are removed through preprocessing.

[0056] Specifically, various sensors installed on the wind power generation unit are used to monitor the operation status of the equipment in real time, including various parameters such as wind speed, temperature, and vibration. These data are transmitted to the monitoring center in real time through the network, and in-depth analysis is carried out using big data technology to grasp the health status of the equipment in real time.

[0057] S102: Obtain the historical operation data of the wind turbine, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through the historical operation data of the wind turbine, and output the optimal wind turbine fault prediction model;

[0058] Introduce artificial intelligence technology, especially machine learning and deep learning algorithms, to conduct intelligent diagnosis of equipment faults. By analyzing the equipment operation data and fault history, these algorithms can quickly and accurately identify potential problems of the equipment and give corresponding solutions.

[0059] Fault prediction technology analyzes historical operation data and real-time data, combines the fault modes and fault mechanisms of wind turbines, and uses technologies such as data mining and machine learning to build prediction models to achieve the prediction of possible fault types and fault times of wind turbines in the future. This technology can help the operation and maintenance team formulate maintenance plans and preventive measures in advance, reducing or avoiding losses caused by faults.

[0060] S103: Through the optimal wind turbine fault prediction model and the wind turbine operation data to be detected, output the result of whether the current wind turbine is abnormal;

[0061] S104: If no abnormality occurs, output the result of normal detection. If an abnormality occurs, obtain the wind turbine operation data to be detected, group it according to the data type, and obtain several groups of operation data of different types;

[0062] S105: Establish a data evaluation model, which is used to evaluate the degree of data fluctuation, perform normalization processing on several groups of operation data of different types, calculate the evaluation values of several groups of operation data of different types respectively through the data evaluation model, and sort several groups of operation data of different types according to the size of the evaluation values, and output the sorting result.

[0063] Through the intelligent management system, the operating status of wind turbines can be monitored in real time, data can be automatically collected and analyzed, greatly improving the control ability of wind turbine performance. With remote monitoring and diagnosis, operation and maintenance personnel can timely understand the operating conditions of the turbines, reduce the need for on-site inspections, optimize the maintenance plan, and quickly respond to various faults and abnormal conditions, reducing labor costs and time costs.

[0064] The method provided by the present invention mainly includes outputting the result of whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the operating data of the wind turbine to be detected. If an abnormality occurs, a data evaluation model is established. The data evaluation model is used to evaluate the degree of data fluctuation. The evaluation values of several groups of types of operating data are calculated respectively through the data evaluation model, and several groups of types of operating data are sorted according to the size of the evaluation values. Through the above method, when the wind turbine data is abnormal, subsequent evaluation results are provided, and it is judged which group of types of data fluctuates more currently through the degree of data fluctuation. Through the sorting result, the problems or equipment corresponding to the data with a large degree of fluctuation are preferentially processed to improve work efficiency.

[0065] In an exemplary embodiment of the present invention, the preprocessing of the initial real-time wind turbine operating data includes:

[0066] Identifying whether there are missing values in the initial real-time wind turbine operating data. If there are missing values, an interpolation function model is established, and the missing values are filled through the interpolation function model.

[0067] If there are no missing values, it is directly output as the operating data of the wind turbine to be detected.

[0068] Specifically, it further includes obtaining a first value and a second value, including:

[0069] S201: Obtain two values adjacent to the missing value, and judge whether the two adjacent values are normal values. If they are normal values, the two values are respectively recorded as the first value and the second value;

[0070] S202: If a non-normal value appears, obtain the next value adjacent to the non-normal value, and judge whether the next adjacent value is a normal value until a normal value is obtained, and save the normal value as the first value or the second value;

[0071] S203: Whether it is a normal value includes calculating the average value of the same type of data and setting a difference threshold, and judging whether the difference between the two adjacent values and the average value is within the difference threshold. If so, it is judged as a normal value, and if not, it is judged as an abnormal value.

[0072] In this example, the aim is to avoid incorporating missing values or outliers into the interpolation algorithm, which may lead to large errors in the values to be inserted.

[0073] Specifically, the interpolation function model includes:

[0074]

[0075] In the formula, X z For replacing missing values by interpolation, a1 is the first value and a2 is the second value.

[0076] In an exemplary embodiment of the present invention, establishing a fault prediction model for a wind turbine includes setting an input layer, a convolutional layer, an activation function layer, a pooling layer, a global average pooling layer, a fully connected layer, and an output layer;

[0077] The input layer is used to transfer sample data and perform feature extraction. It is the entrance of the network, responsible for receiving and transferring sample data, providing necessary inputs for subsequent feature extraction and classification processes. The design of this layer directly affects the network's ability to understand and represent data. The input layer receives the original information from the data set and converts it into a format that can be processed by the network, enabling the network to better understand and learn the features of the data. The design of the input layer not only needs to consider the original form and features of the data, but also needs to combine the network structure and task requirements, so that the network can effectively learn the abstract features of the data and provide a good foundation for subsequent feature extraction and classification tasks.

[0078] The convolutional layer is used to perform convolutional operations on the data. It consists of convolutional kernels, and the convolutional kernel is a weight vector. The operation performed by the convolutional layer is: the convolutional kernel performs a convolutional operation on the data input to the convolutional layer at a certain stride, and then outputs the result of the convolutional operation to the next layer.

[0079] The activation function layer is used to simulate the transmission process of the nervous system. Through the activation function, the neural network can screen and process information, similar to the way the human brain processes information. This design improves the network's expressive ability and learning efficiency. The activation function applies a non-linear transformation to the feature data output by the convolutional layer, expanding the network's expressive ability, enabling it to better capture complex patterns and features in the data. This non-linear transformation enables the convolutional neural network to handle various complex tasks, and the performance of the neural network is also improved. Without the activation layer and non-linear activation function, the network can only perform linear operations, and its expressive ability is limited to linearity and cannot handle non-linear tasks.

[0080] In this embodiment, the activation functions adopted in this article include the ReLU function and the Softplus function.

[0081] The pooling layer is used to reduce the dimension of the output of the convolutional layer to reduce the length of the data. By reducing the dimension, the pooling layer can not only improve the computational efficiency of the model and reduce the number of model parameters, but also retain important feature information. In this way, by avoiding overfitting, the model successfully improves its generalization ability on new data. In actual operation, the pooling layer selects the maximum or average value of the local area of the feature data, thereby realizing the dimensionality reduction processing of the features.

[0082] The fully connected layer is used for data classification. The operation performed by the fully connected layer is to flatten the data output by the previous layer, multiply it by the weights, add the bias, and substitute it into the softplus function, thereby completing the classification of separable data.

[0083] The output layer is used to output the diagnostic result. The input corresponding to the neuron with the maximum output value in the softplus layer is used as the output classification result, that is, the category corresponding to the maximum probability value of the softplus function is output, thereby completing the diagnosis of the faults of the wind turbine.

[0084] Specifically, the convolutional layer includes:

[0085]

[0086] where x i+1 is the output of the i-th convolutional layer, x i is the input of the i-th convolutional layer, W i is the convolutional kernel, b i is the bias term, is the convolutional operation;

[0087] The activation function layer includes;

[0088]

[0089] where f(x) is the activation function and x is the input value of the activation function;

[0090] The pooling layer includes;

[0091]

[0092] where p l(i,j) is the output from the i-th pooling layer to the j-th fully connected layer, x l(i,t) is the input from the t-th activation function layer to the i-th pooling layer, W is the regularization coefficient, and j and t are natural numbers;

[0093] The fully connected layer includes:

[0094]

[0095] where Z l+1(j)is the output of the j-th neuron in the (l + 1)-th layer, x l(i) is the output of the i-th neuron in the previous layer, is the weight of the connection between the j-th neuron in the (l + 1)-th layer and the i-th neuron in the previous layer, O j is the probability corresponding to the output of the j-th neuron in the (l + 1)-th layer, l and k are natural numbers, is the bias term of the j-th neuron in the (l + 1)-th layer.

[0096] In an exemplary embodiment of the present invention, the establishment of the data evaluation model includes:

[0097]

[0098] In the formula, N is the total number of data of this group of types, f i is the i-th data of this group of type data, R f is the evaluation value.

[0099] After normalizing several groups of operation data of types, the evaluation value of the same type of data is calculated according to the above formula. According to the comparison of the evaluation values, the floating degree is determined. The larger the evaluation value, the greater the floating degree of the data, and the problem corresponding to this type of data should be processed preferentially.

[0100] Among them, the normalization of the data includes maximum-minimum normalization: scaling the data to a specified range, usually [0, 1] or [-1, 1]. Or, by moving the decimal point position of the data, the data is mapped to between [-1, 1].

[0101] In a second aspect, the present invention provides a wind turbine operation and maintenance management system, including:

[0102] A data preprocessing module, configured to set a collection period, obtain initial real-time wind turbine operation data within a collection period, preprocess the initial real-time wind turbine operation data, and output the wind turbine operation data to be detected;

[0103] A prediction model module, configured to obtain the historical operation data of the wind turbine, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through the historical operation data of the wind turbine, output the optimal wind turbine fault prediction model, and output the result of whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the wind turbine operation data to be detected;

[0104] The result output module is configured to output a normal detection result if no exception occurs. If an exception occurs, it acquires the operating data of the fan to be detected, groups the data by data type, and obtains the operating data of several types; establishes a data evaluation model, which is used to evaluate the degree of data fluctuation, calculates the evaluation values of the operating data of several types through the data evaluation model respectively, sorts the operating data of several types according to the magnitudes of the evaluation values, and outputs the sorting result.

[0105] In addition, in each embodiment of the present invention, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0106] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs and other various media that can store program codes.

[0107] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A wind turbine operation and maintenance management method, characterized in that: include: Set a collection cycle, obtain initial real-time fan operation data within a collection cycle, pre-process the initial real-time fan operation data, and output the fan operation data to be detected; Obtain historical operation data of wind turbines, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through historical operation data of wind turbines, and output the optimal wind turbine fault prediction model; Through the optimal wind turbine fault prediction model and the operating data of the wind turbine to be tested, the result of whether the current wind turbine is abnormal is output; If no abnormality occurs, the normal detection result is output; if an abnormality occurs, the operation data of the fan to be detected is obtained, and the data is grouped according to the type to obtain several groups of operation data; A data evaluation model is established, wherein the data evaluation model is used to evaluate the floating degree of data, normalize several groups of operating data, calculate evaluation values ​​of several groups of operating data respectively through the data evaluation model, sort the several groups of operating data according to the size of the evaluation values, and output the sorting results.

2. A wind turbine operation and maintenance management method according to claim 1, characterized in that: The preprocessing of the initial real-time fan operation data comprises: Identify whether the initial real-time fan operation data has missing values. If there are missing values, establish an interpolation function model to fill the missing values ​​through the interpolation function model; If there is no missing value, it will be directly output as the operating data of the fan to be tested.

3. A wind turbine operation and maintenance management method according to claim 2, characterized in that: The method further includes obtaining a first value and a second value, including: Obtain two adjacent values ​​of the missing value, and determine whether the two adjacent values ​​are normal values. If they are normal values, record the two values ​​as the first value and the second value respectively. If a non-normal value appears, the next value adjacent to the abnormal value is obtained, and it is determined whether the next value adjacent to the abnormal value is a normal value, until a normal value is obtained, and the normal value is saved as the first value or the second value; Whether it is a normal value includes calculating the average value of the same type of data and setting a difference threshold to determine whether the difference between two adjacent values ​​and the average value is within the difference threshold. If so, it is determined to be a normal value, otherwise it is determined to be an abnormal value.

4. A wind turbine operation and maintenance management method according to claim 3, characterized in that: The interpolation function model includes: In the formula, X z Replacement interpolation of missing values, a1 is the first value, a2 is the second value.

5. A wind turbine operation and maintenance management method according to claim 1, characterized in that: The establishment of the wind turbine fault prediction model includes setting an input layer, a convolution layer, an activation function layer, a pooling layer, a full mean pooling layer, a fully connected layer and an output layer; The input layer is used to transmit sample data and extract features, the convolution layer is used to perform convolution operations on the data, the activation function layer is used to simulate the transmission process of the nervous system, the pooling layer is used to reduce the dimension of the output of the convolution layer, the fully connected layer is used to classify the data, and the output layer is used to output the diagnosis results.

6. A wind turbine operation and maintenance management method according to claim 5, characterized in that: The convolutional layer includes: In the formula, x i+1 is the output of the convolutional layer at layer i, x i is the input of the i-th convolutional layer, W i is the convolution kernel, b i is the bias term, is the convolution operation; The activation function layer includes: In the formula, f(x) is the activation function, and x is the input value of the activation function; The pooling layer comprises: In the formula, p l(i,j) The output of the i-th pooling layer to the j-th fully connected layer, x l(i,t) is the input from the tth activation function layer to the i-th pooling layer, W is the regularization coefficient, j and t are natural numbers; The fully connected layer includes: In the formula, Z l+1(j) is the output of the jth neuron in the l+1th layer, x l(i) is the output of the ith neuron in the previous layer, is the weight of the connection between the jth neuron in the l+1th layer and the ith neuron in the previous layer, O j is the probability of the output corresponding to the jth neuron in the l+1th layer, l and k are natural numbers, is the bias term of the jth neuron in the l+1th layer.

7. A wind turbine operation and maintenance management method according to claim 5, characterized in that: The establishing of the data evaluation model comprises: In the formula, N is the total number of data types in this group, f i is the i-th data of this group of data, R f For the evaluation value.

8. A wind turbine operation and maintenance management system, characterized in that: include: The data preprocessing module is configured to set a collection cycle, obtain initial real-time fan operation data within a collection cycle, preprocess the initial real-time fan operation data, and output the fan operation data to be detected; The prediction model module is configured to obtain historical operation data of the wind turbine, establish a wind turbine fault prediction model, train the wind turbine fault prediction model through the historical operation data of the wind turbine, output the optimal wind turbine fault prediction model, and output the result of whether the current wind turbine is abnormal through the optimal wind turbine fault prediction model and the operation data of the wind turbine to be detected; The result output module is configured to output a normal detection result if no abnormality occurs, and if an abnormality occurs, obtain the operation data of the fan to be detected, group the data by type, and obtain several groups of operation data; establish a data evaluation model, the data evaluation model is used to evaluate the floating degree of the data, calculate the evaluation values ​​of several groups of operation data respectively through the data evaluation model, and sort the several groups of operation data according to the size of the evaluation value, and output the sorting result; A main control module is connected to the data preprocessing module, the prediction model module and the result output module, and is used to execute a wind turbine operation and maintenance management method as described in any one of claims 1-7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the wind turbine operation and maintenance management method described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a wind turbine operation and maintenance management method according to any one of claims 1 to 7 is implemented.

Images