Wind turbine energy efficiency status monitoring method, device and storage medium

Abstract

The present application discloses a method, device and storage medium for monitoring the energy efficiency status of a wind turbine, and relates to the field of artificial intelligence, specifically to the fields of deep learning, neural networks, natural language processing, wind turbines and new energy. The specific implementation scheme is: obtaining wind turbine monitoring data; preprocessing the wind turbine monitoring data; determining first data based on the preprocessed wind turbine monitoring data; determining a test set based on the first data; inputting the data in the test set into the pre-trained knowledge distillation student model to obtain the wind turbine energy efficiency status prediction result. The present application improves the accuracy of abnormal energy efficiency status monitoring of wind turbines and reduces the monitoring and operating costs of wind turbines.

Description

Technical Field

[0001] The patent of this invention relates to the field of artificial intelligence, specifically to deep learning, neural networks, natural language processing, wind turbines and new energy fields, and in particular to a method, device and storage medium for monitoring the energy efficiency status of wind turbines. Background Art

[0002] Among the related technologies, wind power has become one of the important supports for the country's clean energy. With the large-scale development of wind power, there are many challenges faced by wind power equipment, daily maintenance of wind power equipment, energy efficiency status of units, fault warning, etc. Therefore, how to timely and accurately monitor the abnormal state of energy efficiency of wind turbines, so as to give timely warning of abnormal conditions, has become a problem to be solved urgently. Summary of the invention

[0003] The present application provides a method, device and storage medium for monitoring the energy efficiency status of a wind turbine generator set.

[0004] According to a first aspect of the present application, a method for monitoring energy efficiency status of a wind turbine is provided, comprising:

[0005] Obtain wind turbine monitoring data;

[0006] Preprocessing the wind turbine monitoring data;

[0007] Determining first data based on the preprocessed wind turbine monitoring data;

[0008] Based on the first data, determining a test set;

[0009] The data in the test set is input into the pre-trained knowledge distillation student model to obtain the wind turbine energy efficiency status prediction result.

[0010] According to one embodiment of the present disclosure, the first data includes reference degree data and deviation degree data.

[0011] According to one embodiment of the present disclosure, the knowledge distillation student model is trained by the following method:

[0012] Obtain wind turbine monitoring sample data;

[0013] Preprocessing the wind turbine monitoring sample data;

[0014] Determine first sample data based on the preprocessed wind turbine monitoring sample data;

[0015] Determine the energy efficiency status of the wind turbine generator system based on the first sample data;

[0016] Based on the energy efficiency state, labeling the first sample data to obtain a first sample data label;

[0017] Pre-training the knowledge distillation teacher model based on the first sample data;

[0018] Inputting the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result;

[0019] The knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model.

[0020] According to an embodiment of the present disclosure, pre-training the knowledge distillation teacher model based on the first sample data includes:

[0021] Based on the first sample data, determining a training sample set;

[0022] Inputting the training sample set into the knowledge distillation teacher model to obtain a second prediction result;

[0023] According to a preset loss function, calculate the loss value of the second prediction result and the validation set label;

[0024] The knowledge distillation teacher model is pre-trained based on the loss value.

[0025] According to an embodiment of the present disclosure, the knowledge distillation student model includes a bidirectional long-short time calculation unit and a representation capability improvement component; the first sample data is input into the pre-trained knowledge distillation teacher model to obtain a first prediction result, including:

[0026] inputting the first sample data into the bidirectional long-short time calculation unit to obtain feature information of the first sample data;

[0027] Based on the feature information, the first sample data is input into the characterization capability improvement component to obtain the first sample data after the characterization capability is improved;

[0028] According to a preset normalized exponential function, the first prediction result is calculated based on the first sample data with improved characterization capability.

[0029] According to an embodiment of the present disclosure, the step of training the knowledge distillation student model based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model includes:

[0030] Based on the first sample data, determining a test sample set;

[0031] Inputting the test sample set into the knowledge distillation student model to obtain a second prediction result;

[0032] Calculate a first loss value according to the second prediction result and the first prediction result according to a preset loss function;

[0033] Calculate a second loss value according to the second prediction result and the first sample data label according to a preset loss function;

[0034] The knowledge distillation student model is trained according to the first loss value and the second loss value to obtain the trained knowledge distillation student model.

[0035] According to a second aspect of the present application, a wind turbine energy efficiency status monitoring device is provided, comprising:

[0036] An acquisition module is used to acquire wind turbine monitoring data;

[0037] A preprocessing module, used for preprocessing the wind turbine monitoring data;

[0038] A first determination module, configured to determine first data based on the preprocessed wind turbine monitoring data;

[0039] A second determination module, used to determine a test set based on the first data;

[0040] The prediction module is used to input the data in the test set into the pre-trained knowledge distillation student model to obtain the energy efficiency status prediction result of the wind turbine set.

[0041] According to one embodiment of the present disclosure, the first data includes reference degree data and deviation degree data.

[0042] According to one embodiment of the present disclosure, the device further includes:

[0043] A pre-training module is used to pre-train the knowledge distillation student model; wherein the pre-training module is specifically used to:

[0044] Obtain wind turbine monitoring sample data;

[0045] Preprocessing the wind turbine monitoring sample data;

[0046] Determine first sample data based on the preprocessed wind turbine monitoring sample data;

[0047] Determine the energy efficiency status of the wind turbine generator system based on the first sample data;

[0048] Based on the energy efficiency state, labeling the first sample data to obtain a first sample data label;

[0049] Pre-training the knowledge distillation teacher model based on the first sample data;

[0050] Inputting the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result;

[0051] The knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model.

[0052] According to one embodiment of the present disclosure, the pre-training module is specifically used for:

[0053] Based on the first sample data, determining a training sample set;

[0054] Inputting the training sample set into the knowledge distillation teacher model to obtain a second prediction result;

[0055] According to a preset loss function, calculate the loss value of the second prediction result and the validation set label;

[0056] The knowledge distillation teacher model is pre-trained based on the loss value.

[0057] According to an embodiment of the present disclosure, the knowledge distillation student model includes a bidirectional long- and short-time calculation unit and a representation capability improvement component; wherein the pre-training module is specifically used for:

[0058] inputting the first sample data into the bidirectional long-short time calculation unit to obtain feature information of the first sample data;

[0059] Based on the feature information, the first sample data is input into the characterization capability improvement component to obtain the first sample data after the characterization capability is improved;

[0060] According to a preset normalized exponential function, the first prediction result is calculated based on the first sample data with improved characterization capability.

[0061] According to one embodiment of the present disclosure, the pre-training module is specifically used for:

[0062] Based on the first sample data, determining a test sample set;

[0063] Inputting the test sample set into the knowledge distillation student model to obtain a second prediction result;

[0064] Calculate a first loss value according to the second prediction result and the first prediction result according to a preset loss function;

[0065] Calculate a second loss value according to the second prediction result and the first sample data label according to a preset loss function;

[0066] The knowledge distillation student model is trained according to the first loss value and the second loss value to obtain the trained knowledge distillation student model.

[0067] According to a third aspect of the present application, a storage medium is provided. When instructions in the storage medium are executed by a processor of a computer device, the computer device can execute the wind turbine energy efficiency status monitoring method as described in the first aspect.

[0068] According to a fourth aspect of the present application, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the wind turbine energy efficiency status monitoring method described in the first aspect when executing the computer program.

[0069] According to a fifth aspect of the present application, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the method for monitoring the energy efficiency status of a wind turbine generator system as described in the first aspect are implemented.

[0070] According to the technical solution of the present application, based on the monitoring data of the wind turbine set, the abnormal state of energy efficiency of the wind turbine set is predicted through knowledge distillation, which improves the accuracy of monitoring the abnormal state of energy efficiency of the wind turbine set and reduces the monitoring and operation cost of the wind turbine set.

[0071] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application.

[0073] Figure 1 is a schematic diagram according to the first embodiment of the present application;

[0074] Figure 2 is a schematic diagram according to a second embodiment of the present application;

[0075] Figure 3 is a schematic diagram according to a third embodiment of the present application;

[0076] Figure 4 It is a block diagram of a computer device used to implement the wind turbine energy efficiency status monitoring method of the embodiment of the present application. DETAILED DESCRIPTION

[0077] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0078] Among the related technologies, wind power has become one of the important supports for the country's clean energy. With the large-scale development of wind power, there are many challenges in the daily maintenance of wind power equipment, energy efficiency status of units, fault warning, etc. Therefore, how to timely and accurately monitor the abnormal state of wind turbine energy efficiency and give timely warning of abnormal state has become an urgent problem to be solved.

[0079] Based on the above problems, the present application proposes a method, device and storage medium for monitoring the energy efficiency status of a wind turbine set. According to the monitoring data of the wind turbine set, the abnormal energy efficiency status of the wind turbine set can be predicted through knowledge distillation, thereby improving the accuracy of monitoring the abnormal energy efficiency status of the wind turbine set and reducing the monitoring and operation costs of the wind turbine set.

[0080] Figure 1 It is a schematic diagram according to the first embodiment of the present application. It should be noted that the wind turbine energy efficiency status monitoring method in the embodiment of the present application can be used for the wind turbine energy efficiency status monitoring device in the embodiment of the present application, and the device can be configured in a computer device. Figure 1 As shown, the wind turbine energy efficiency status monitoring method includes the following steps:

[0081] Step 101, obtaining wind turbine monitoring data.

[0082] As a possible example, the wind turbine monitoring data can be obtained by real-time data collection of the wind turbine through a data acquisition and monitoring control system. Optionally, the data acquisition and monitoring control system can be a SCADA (Supervisory Control And Data Acquisition) system. Optionally, the wind turbine monitoring data can be data collected in real time by the SCADA system, or can be historical data collected by the SCADA system.

[0083] Step 102: pre-processing the wind turbine monitoring data.

[0084] It can be understood that since there are abnormal data such as null values ​​or missing values ​​in the wind turbine monitoring data collected by the data acquisition and monitoring control system, it is necessary to remove the abnormal data through preprocessing.

[0085] Step 103: Determine first data based on the pre-processed wind turbine monitoring data.

[0086] It can be understood that the first data is calculated based on the monitoring data of the wind turbine generator set.

[0087] In this embodiment of the present application, the first data includes reference data and deviation data.

[0088] Among them, in an embodiment of the present application, the deviation data may include at least one or more of the following 1)-6) items: 1) the deviation of the generator bearing temperature in the non-driving direction; 2) the deviation of the generator assembly temperature; 3) the deviation of the gearbox bearing temperature; 4) the deviation of the gearbox oil temperature; 5) the deviation of the first temperature of the variable pitch motor of the wind turbine; 6) the deviation of the second temperature of the variable pitch motor of the wind turbine; 7) the deviation of the third temperature of the variable pitch motor of the wind turbine.

[0089] Step 104: determine a test set based on the first data.

[0090] As a possible example, part of the data is selected from the first data as the data of the test set.

[0091] Step 105, input the data in the test set into the pre-trained knowledge distillation student model to obtain the wind turbine energy efficiency status prediction result.

[0092] It should be noted that the above-mentioned knowledge distillation student model is a trained student model. The data in the test set is input into the knowledge distillation student model. The knowledge distillation student model predicts the abnormal state in the energy efficiency state of the wind turbine based on the data in the test set to obtain a prediction result. Optionally, the above-mentioned first sample data label can be a state label such as normal or abnormal, or a score value of an efficient state. It can be a state label such as normal or abnormal, or a score value of an efficient state.

[0093] As a possible example, the prediction result of the energy efficiency status of the wind turbine generator set may be output and displayed, thereby facilitating the staff to check the energy efficiency status of the wind turbine generator set.

[0094] Optionally, the knowledge distillation student model can use a BiLSTM (Bi-directional Long Short-Term Memory) with a dropout component to enhance representation capabilities.

[0095] According to the method for monitoring the energy efficiency status of a wind turbine set in an embodiment of the present application, by acquiring wind turbine set monitoring data, the wind turbine set monitoring data is preprocessed, first data is determined based on the preprocessed wind turbine set monitoring data, a test set is determined based on the first data, and the data in the test set is input into a pre-trained knowledge distillation student model to obtain a prediction result of the energy efficiency status of the wind turbine set. The prediction of the abnormal energy efficiency status of the wind turbine set is achieved through knowledge distillation, the accuracy of monitoring the abnormal energy efficiency status of the wind turbine set is improved, and the monitoring operation cost of the wind turbine set is reduced. First, the data collected by the SCADA system is cleaned, abnormal data is screened out, and model training is carried out in combination with the historical data of the SCADA system to ensure the accuracy of predicting anomalies. Therefore, a small model is selected for data prediction and combined with the large model to output soft-labeled results. This is not only beneficial to the time and space calculation complexity of the small model and solves the problem of insufficient generalization ability of the small model, but also makes full use of the historical data of the SCADA system to improve the quality of the pre-trained large model. Therefore, the knowledge distillation method is used to not only improve the accuracy of wind turbine energy efficiency status anomaly detection, but also to optimize and significantly reduce the daily monitoring and operation costs of wind turbine power generation systems, including manpower, material resources, time efficiency and other cost reductions and efficiency increases.

[0096] In order to ensure that the knowledge distillation student model can more accurately predict the abnormal energy efficiency status data of the wind turbine, optionally, by obtaining wind turbine monitoring sample data, the wind turbine monitoring sample data is preprocessed, and first sample data is determined based on the preprocessed wind turbine monitoring sample data, and the energy efficiency status of the wind turbine is determined based on the first sample data. Based on the energy efficiency status, the first sample data is marked to obtain a first sample data label, and the knowledge distillation teacher model is pre-trained based on the first sample data. The first sample data is input into the pre-trained knowledge distillation teacher model to obtain a first prediction result, and the knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain a trained knowledge distillation student model. Figure 2 It is a schematic diagram according to the second embodiment of the present application. It should be noted that the wind turbine energy efficiency status monitoring method of the embodiment of the present application can be executed by the wind turbine energy efficiency status monitoring device in the embodiment of the present application. In some embodiments of the present application, such as Figure 2 As shown in the figure, the knowledge distillation student model proposed in the wind turbine energy efficiency status monitoring method is trained by the following method:

[0097] Step 210: Obtain wind turbine monitoring sample data.

[0098] As a possible example, the wind turbine monitoring sample data may be obtained by real-time data acquisition of the wind turbine by a data acquisition and monitoring control system. Optionally, the data acquisition and monitoring control system may be a SCADA (Supervisory Control And Data Acquisition) system.

[0099] Step 220: pre-process the wind turbine monitoring sample data.

[0100] It can be understood that since abnormal data such as null values ​​or missing values ​​exist in the wind turbine monitoring sample data collected by the data acquisition and monitoring control system, it is necessary to remove the abnormal data through preprocessing.

[0101] Step 230: Determine first sample data based on the pre-processed wind turbine monitoring sample data.

[0102] It can be understood that the first sample data is calculated based on the wind turbine monitoring sample data.

[0103] Step 240: determine the energy efficiency status of the wind turbine generator system based on the first sample data.

[0104] As a possible example, the above energy efficiency status can be obtained by judging the quality of the active power of the wind turbine generator set.

[0105] Step 250: label the first sample data based on the energy efficiency status to obtain a first sample data label.

[0106] As a possible example, the first sample data is labeled according to the energy efficiency state to obtain a first sample data label, and the first sample data label is used as a hard label for training the knowledge distillation student sample. Optionally, the first sample data label can be a state label such as normal or abnormal, or a score value of an efficient state.

[0107] Step 260, pre-training the knowledge distillation teacher model based on the first sample data;

[0108] The pre-training of the knowledge distillation teacher model based on the first sample data includes the following steps:

[0109] Step 261: determine a training sample set based on the first sample data.

[0110] As a possible example, part of the data is selected from the first sample data as sample data of the training sample set.

[0111] Step 262, input the training sample set into the knowledge distillation teacher model to obtain a second prediction result.

[0112] As a possible example, the data in the training sample set is input into the trained knowledge distillation teacher model. The knowledge distillation teacher model predicts the abnormal state of energy efficiency of the wind turbine according to the data in the training sample set to obtain a second prediction result.

[0113] Step 263, according to a preset loss function, calculate the loss value of the second prediction result and the validation set label.

[0114] Optionally, the above loss function may be a mean square error loss function.

[0115] Step 264, pre-train the knowledge distillation teacher model based on the loss value.

[0116] Optionally, the Adam algorithm can be used to pre-train the teacher model.

[0117] Step 270, input the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result.

[0118] It can be understood that the first sample data is input into the pre-trained knowledge distillation teacher model, and the pre-trained knowledge distillation teacher model predicts the abnormal state of the wind turbine energy efficiency based on the first sample data to obtain a first prediction result, and the first prediction result is used as a soft label for training the knowledge distillation student model.

[0119] The knowledge distillation student model includes a bidirectional long- and short-time calculation unit and a representation capability improvement component; inputting the first sample data into the pre-trained knowledge distillation teacher model to obtain the first prediction result includes the following steps:

[0120] Step 271: input the first sample data into a bidirectional long-short time calculation unit to obtain feature information of the first sample data.

[0121] As a possible example, text sequence information of the first sample data is extracted through BiLSTM (Bi-directional Long Short-Term Memory) to obtain text feature information of the first sample data.

[0122] Step 272: Based on the feature information, the first sample data is input into the dropout component for improving the representation capability to obtain the first sample data after the representation capability is improved.

[0123] As a possible example, based on the feature information, the first sample data is input into the representation capability enhancement component dropout, which performs non-optimization on the text feature information, thereby improving the overfitting capability of the knowledge distillation student model, filtering out redundant information in the text feature information, and optimizing the loss function.

[0124] Step 273, according to the preset softmax classification function, a first prediction result is obtained based on the first sample data with improved representation capability.

[0125] As a possible example, a first prediction result is calculated based on the first sample data with improved representation capability through a preset softmax classification function.

[0126] Optionally, an activation function ReLU function is used as the activation function of the distillation teacher model and the distillation student model.

[0127] Step 280: Train the knowledge distillation student model based on the first prediction result and the first sample data label to obtain a trained knowledge distillation student model.

[0128] Optionally, the Adam algorithm can be used to pre-train the teacher model.

[0129] The step of training the knowledge distillation student model based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model includes the following steps:

[0130] Step 281: Determine a test sample set based on the first sample data.

[0131] As a possible example, part of the data is selected from the first sample data as sample data of the test sample set.

[0132] Step 282, input the test sample set into the knowledge distillation student model to obtain a second prediction result.

[0133] As a possible example, the test sample set is input into the knowledge distillation student model to be trained. The knowledge distillation student model predicts the abnormal state of energy efficiency of the wind turbine according to the data in the test sample set to obtain a second prediction result.

[0134] Step 283, according to a preset loss function, calculate the first loss value based on the second prediction result and the first prediction result.

[0135] Step 284, calculate the second loss value according to the second prediction result and the first sample data label according to the preset loss function.

[0136] Optionally, the above loss function may be a mean square error loss function.

[0137] Step 285, training the knowledge distillation student model according to the first loss value and the second loss value to obtain a trained knowledge distillation student model.

[0138] As a possible example, the first loss value and the second loss value are added to obtain a total loss value, and the knowledge distillation student model is trained according to the total loss value, so that the knowledge distillation student model learns to predict the accurate abnormal state of wind turbine energy efficiency.

[0139] According to the wind turbine energy efficiency status monitoring method of the embodiment of the present application, by acquiring wind turbine monitoring sample data, the wind turbine monitoring sample data is preprocessed, and first sample data is determined based on the preprocessed wind turbine monitoring sample data, and the energy efficiency status of the wind turbine is determined based on the first sample data. Based on the energy efficiency status, the first sample data is marked to obtain a first sample data label, and a knowledge distillation teacher model is pre-trained based on the first sample data. The first sample data is input into the pre-trained knowledge distillation teacher model to obtain a first prediction result, and the knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain a trained knowledge distillation student model, so that the knowledge distillation student model can more accurately predict the abnormal energy efficiency status data of the wind turbine, thereby improving the accuracy of abnormal energy efficiency status detection of the wind turbine.

[0140] In order to implement the above embodiment, the present application proposes a wind turbine energy efficiency status monitoring device.

[0141] Figure 3 is a schematic diagram according to the third embodiment of the present application. Figure 3 As shown, the device comprises:

[0142] The acquisition module 301 is used to acquire wind turbine monitoring data.

[0143] The preprocessing module 302 is used to preprocess the wind turbine monitoring data.

[0144] The first determination module 303 is used to determine first data based on the pre-processed wind turbine monitoring data.

[0145] The first data includes reference data and deviation data.

[0146] The second determination module 304 is used to determine a test set based on the first data.

[0147] The prediction module 305 is used to input the data in the test set into the pre-trained knowledge distillation student model to obtain the wind turbine energy efficiency status prediction result.

[0148] The pre-training module 306 is used to pre-train the knowledge distillation student model; wherein the pre-training module 306 is specifically used to: obtain wind turbine monitoring sample data; pre-process the wind turbine monitoring sample data; determine the first sample data based on the pre-processed wind turbine monitoring sample data; determine the energy efficiency status of the wind turbine based on the first sample data; based on the energy efficiency status, label the first sample data to obtain a first sample data label; based on the first sample data, pre-train the knowledge distillation teacher model; input the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result; train the knowledge distillation student model based on the first prediction result and the first sample data label to obtain a trained knowledge distillation student model.

[0149] Optionally, in some embodiments of the present application, the pre-training module 306 may implement pre-training the knowledge distillation teacher model based on the first sample data as follows:

[0150] According to the wind turbine energy efficiency status monitoring device of the embodiment of the present application, a training sample set is determined based on the first sample data; the training sample set is input into the knowledge distillation teacher model to obtain a second prediction result; according to a preset loss function, the loss value of the second prediction result and the verification set label is calculated; and the knowledge distillation teacher model is pre-trained based on the loss value.

[0151] Optionally, in some embodiments of the present application, the pre-training module 306 inputs the first sample data into the pre-trained knowledge distillation teacher model to obtain the first prediction result in the following manner:

[0152] The first sample data is input into the bidirectional long-short time calculation unit to obtain feature information of the first sample data; based on the feature information, the first sample data is input into the representation capability improvement component to obtain the first sample data with improved representation capability; according to the preset softmax classification function, the first prediction result is obtained based on the first sample data with improved representation capability.

[0153] Optionally, in some embodiments of the present application, the pre-training module 306 trains the knowledge distillation student model based on the first prediction result and the first sample data label, and the implementation method of obtaining the trained knowledge distillation student model may be as follows:

[0154] Based on the first sample data, determine a test sample set; input the test sample set into the knowledge distillation student model to obtain a second prediction result; according to a preset loss function, calculate a first loss value based on the second prediction result and the first prediction result; according to a preset loss function, calculate a second loss value based on the second prediction result and the first sample data label; train the knowledge distillation student model according to the first loss value and the second loss value to obtain a trained knowledge distillation student model.

[0155] According to the wind turbine energy efficiency status monitoring device of the embodiment of the present application, by acquiring wind turbine monitoring data, the wind turbine monitoring data is preprocessed, first data is determined based on the preprocessed wind turbine monitoring data, and a test set is determined based on the first data. The data in the test set is input into the pre-trained knowledge distillation student model to obtain a wind turbine energy efficiency status prediction result. The abnormal energy efficiency status of the wind turbine is predicted through knowledge distillation, which improves the accuracy of monitoring the abnormal energy efficiency status of the wind turbine and reduces the monitoring operation cost of the wind turbine.

[0156] According to an embodiment of the present application, the present application also provides a computer device and a readable storage medium.

[0157] like Figure 4 , is a block diagram of a computer device according to a method for monitoring the energy efficiency status of a wind turbine according to an embodiment of the present application. The computer device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The computer device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0158] like Figure 4 As shown, the computer device includes: one or more processors 401, memory 402, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as required. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 4 A processor 401 is taken as an example.

[0159] The memory 402 is a non-transitory computer-readable storage medium provided in the present application. The memory stores instructions executable by at least one processor to enable the at least one processor to perform the method for monitoring the energy efficiency status of a wind turbine provided in the present application. The non-transitory computer-readable storage medium of the present application stores computer instructions, which are used to enable a computer to perform the method for monitoring the energy efficiency status of a wind turbine provided in the present application.

[0160] The memory 402 is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, such as the program instructions / modules corresponding to the method for monitoring the energy efficiency status of a wind turbine in the embodiment of the present application (for example, the attached Figure 3 The processor 401 executes various functional applications and data processing of the server by running the non-transient software programs, instructions and modules stored in the memory 402, that is, the method for monitoring the energy efficiency status of the wind turbine in the above method embodiment is implemented.

[0161] The memory 402 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of a computer device for monitoring the energy efficiency status of a wind turbine generator set, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 402 may optionally include a memory remotely arranged relative to the processor 401, and these remote memories may be connected to the computer device for monitoring the energy efficiency status of a wind turbine generator set via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0162] The computer device of the method for monitoring the energy efficiency status of a wind turbine generator set may further include: an input device 403 and an output device 404. The processor 401, the memory 402, the input device 403 and the output device 404 may be connected via a bus or other means. Figure 4 The example of connecting through bus is taken in the following.

[0163] The input device 403 can receive input digital or character information, and generate key signal input related to user settings and function control of the computer device for wind turbine energy efficiency status monitoring, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator rod, one or more mouse buttons, a trackball, a joystick and other input devices. The output device 404 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display and a plasma display. In some embodiments, the display device may be a touch screen.

[0164] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0165] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0166] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0167] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0168] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short).

[0169] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0170] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for monitoring the energy efficiency status of a wind turbine generator system, comprising: Obtain wind turbine monitoring data; Preprocessing the wind turbine monitoring data; Determining first data based on the preprocessed wind turbine monitoring data; Based on the first data, determining a test set; Inputting the data in the test set into the pre-trained knowledge distillation student model to obtain the wind turbine energy efficiency state prediction result; The knowledge distillation student model is trained by the following method: Obtain wind turbine monitoring sample data; Preprocessing the wind turbine monitoring sample data; Determine first sample data based on the preprocessed wind turbine monitoring sample data; Determine the energy efficiency status of the wind turbine generator system based on the first sample data; Based on the energy efficiency state, labeling the first sample data to obtain a first sample data label; Pre-training the knowledge distillation teacher model based on the first sample data; Inputting the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result; The knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model.

2. The method according to claim 1, wherein: The first data includes reference degree data and deviation degree data.

3. The method according to claim 1, wherein: The pre-training of the knowledge distillation teacher model based on the first sample data includes: Based on the first sample data, determining a training sample set; Inputting the training sample set into the knowledge distillation teacher model to obtain a second prediction result; According to a preset loss function, calculate the loss value of the second prediction result and the validation set label; The knowledge distillation teacher model is pre-trained based on the loss value.

4. The method according to claim 1, wherein: The knowledge distillation student model includes a bidirectional long-short time calculation unit and a representation capability improvement component; the first sample data is input into the pre-trained knowledge distillation teacher model to obtain a first prediction result, including: The first sample data is input into the bidirectional long-short time calculation unit to obtain characteristic information of the first sample data; Based on the feature information, the first sample data is input into the characterization capability improvement component to obtain the first sample data after the characterization capability is improved; According to a preset normalized exponential function, the first prediction result is calculated based on the first sample data with improved characterization capability.

5. The method according to claim 1, wherein: The step of training the knowledge distillation student model based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model includes: Based on the first sample data, determining a test sample set; Inputting the test sample set into the knowledge distillation student model to obtain a second prediction result; Calculate a first loss value according to the second prediction result and the first prediction result according to a preset loss function; Calculate a second loss value according to the second prediction result and the first sample data label according to a preset loss function; The knowledge distillation student model is trained according to the first loss value and the second loss value to obtain the trained knowledge distillation student model.

6. A wind turbine energy efficiency status monitoring device, comprising: An acquisition module is used to acquire wind turbine monitoring data; A preprocessing module, used for preprocessing the wind turbine monitoring data; A first determination module, configured to determine first data based on the preprocessed wind turbine monitoring data; A second determination module, used to determine a test set based on the first data; A prediction module, used for inputting the data in the test set into the pre-trained knowledge distillation student model to obtain a prediction result of the energy efficiency state of the wind turbine; The knowledge distillation student model is trained by the following method: Obtain wind turbine monitoring sample data; Preprocessing the wind turbine monitoring sample data; Determine first sample data based on the preprocessed wind turbine monitoring sample data; Determine the energy efficiency status of the wind turbine generator system based on the first sample data; Based on the energy efficiency state, labeling the first sample data to obtain a first sample data label; Pre-training the knowledge distillation teacher model based on the first sample data; Inputting the first sample data into the pre-trained knowledge distillation teacher model to obtain a first prediction result; The knowledge distillation student model is trained based on the first prediction result and the first sample data label to obtain the trained knowledge distillation student model.

7. A storage medium, when instructions in the storage medium are executed by a processor of a computer device, the computer device is enabled to execute the wind turbine energy efficiency status monitoring method according to any one of claims 1 to 5.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the wind turbine energy efficiency status monitoring method according to any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the method for monitoring the energy efficiency status of a wind turbine generator set as claimed in any one of claims 1 to 5 are implemented.

Images