Fault diagnosis method and device for variable pitch system of wind driven generator
Through the combination of Kalman filter and fault diagnosis model, the faults of the wind turbine pitch system are identified and diagnosed, which solves the problem of fault diagnosis of the pitch system, improves diagnostic accuracy and efficiency, and ensures the stable operation of the wind turbine.
Patent Information
- Application Number
- CN202411013123.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-12
AI Technical Summary
The failure of the wind turbine pitch system leads to a decline in power generation efficiency, equipment damage and safety accidents, and it is difficult for the existing technology to effectively diagnose faults, affecting the stable operation and economic losses of the wind turbine.
The Kalman filter is used to process the characteristic data of the wind turbine pitch system, determine the residual between the observation vector and the measured vector, determine the abnormal state based on the residual, and input the characteristic data into the fault diagnosis model to obtain the fault type, through multiple Kals The combination of Mann filter and fault diagnosis model realizes the identification of different faults and the generation and correction of alarm information.
It improves the accuracy and efficiency of fault diagnosis of the pitch system of the wind turbine, reduces the computing burden, supports real-time fault detection and maintenance, and ensures the stable operation and sustainable development of the wind turbine.
Smart Images

Figure CN118934487A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of wind power generation, and in particular to a fault diagnosis method and device for a wind turbine pitch control system. Background Art
[0002] As the global demand for renewable energy continues to rise, wind power generation, as a clean and sustainable energy solution, has become increasingly strategic in the global energy landscape. The variable pitch system of wind turbines, as a highly integrated mechatronic core, maximizes wind energy capture by finely adjusting the blade angle and ensures safe and efficient operation of the unit under different wind speed conditions. The system integrates precise sensors, efficient actuators, advanced controllers, and complex software algorithms, and the failure of any component may have a significant impact on the overall performance. Specifically, the failure of the variable pitch system not only directly leads to a decline in power generation efficiency, but may also induce equipment damage, safety accidents, and even unplanned shutdown of the entire wind turbine, leading to huge economic losses and unstable energy supply.
[0003] Therefore, there is an urgent need for a fault diagnosis method for a wind turbine pitch system to ensure the stable supply and sustainable development of wind power generation. Summary of the invention
[0004] This application proposes a fault diagnosis method and device for a wind turbine pitch control system. The specific scheme is as follows:
[0005] In one aspect, an embodiment of the present application provides a fault diagnosis method for a wind turbine pitch control system, comprising:
[0006] Obtain characteristic data of wind turbine pitch control system;
[0007] Using a Kalman filter to process the characteristic data to determine an observation vector, and determining a residual between the observation vector and a measurement vector of a wind turbine pitch control system, so as to determine an abnormal state based on the residual;
[0008] When the abnormal state is a preset state, the characteristic data is input into the fault diagnosis model to obtain the first fault type output by the fault diagnosis model.
[0009] In a possible implementation of an embodiment of one aspect of the present application, there are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of a different fault.
[0010] In a possible implementation of an embodiment of one aspect of the present application, it also includes:
[0011] In the case where the first fault type does not match the fault corresponding to the abnormal state, generating alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state;
[0012] receiving feedback information of the alarm information, wherein the feedback information includes the second fault type confirmed by the expert;
[0013] When the second fault type is different from the first fault type, the fault diagnosis model is modified.
[0014] In a possible implementation of an embodiment of an aspect of the present application, the fault types include faults of a pitch motor, a pitch bearing, and a reduction gearbox.
[0015] In a possible implementation of an embodiment of one aspect of the present application, it also includes:
[0016] The residual is used to update the state estimate and covariance matrix of the Kalman filter.
[0017] Another aspect of the present application provides a smoke and flame detection device, comprising:
[0018] An acquisition module is used to acquire characteristic data of a wind turbine pitch control system;
[0019] A detection module, used for processing the characteristic data using a Kalman filter to determine an observation vector, and determining a residual between the observation vector and a measurement vector of a wind turbine pitch control system, so as to determine an abnormal state based on the residual;
[0020] The diagnosis module is used to input the characteristic data into the fault diagnosis model when the abnormal state is a preset state, and obtain the first fault type output by the fault diagnosis model.
[0021] In a possible implementation of an embodiment of one aspect of the present application, there are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of a different fault.
[0022] In a possible implementation manner of an embodiment of the first aspect of the present application, a first correction module is further included, which is used to:
[0023] In the case where the first fault type does not match the fault corresponding to the abnormal state, generating alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state;
[0024] receiving feedback information of the alarm information, wherein the feedback information includes the second fault type confirmed by the expert;
[0025] When the second fault type is different from the first fault type, the fault diagnosis model is modified.
[0026] In a possible implementation of an embodiment of an aspect of the present application, the fault types include faults of a pitch motor, a pitch bearing, and a reduction gearbox.
[0027] In a possible implementation of an embodiment of the first aspect of the present application, a second correction module is further included, which is used to:
[0028] The residual is used to update the state estimate and covariance matrix of the Kalman filter.
[0029] Another aspect of the present application provides a computer device, including a processor and a memory;
[0030] The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the method as described in the above embodiment.
[0031] Another aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the program implements the method of the above embodiment when executed by a processor.
[0032] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0034] Figure 1 A schematic flow chart of a fault diagnosis method for a wind turbine pitch control system provided in an embodiment of the present application;
[0035] Figure 2 A schematic flow chart of another fault diagnosis method for a wind turbine pitch control system provided in an embodiment of the present application;
[0036] Figure 3 A schematic structural diagram of a fault diagnosis device for a wind turbine pitch system provided in an embodiment of the present application. DETAILED DESCRIPTION
[0037] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.
[0038] The smoke and flame detection method according to the embodiment of the present application is described below with reference to the accompanying drawings.
[0039] Figure 1 A schematic flow chart of a fault diagnosis method for a wind turbine pitch system provided in an embodiment of the present application.
[0040] The fault diagnosis method of the wind turbine pitch system of the embodiment of the present application is executed by the fault diagnosis device of the wind turbine pitch system provided by the embodiment of the present application (hereinafter referred to as the diagnosis device), which can be configured in a computer device to realize fault diagnosis of the wind turbine pitch system.
[0041] like Figure 1 As shown, the fault diagnosis method of the wind turbine pitch control system includes:
[0042] Step 101, obtaining characteristic data of a wind turbine pitch control system.
[0043] Among them, the characteristic data may include pitch control instructions, meteorological data, power generation performance data (power generation, power output, energy efficiency ratio, power generation hours), etc., and this application does not impose any restrictions on this.
[0044] In the present application, characteristic data of a wind turbine pitch system (hereinafter referred to as the pitch system) may be acquired through a monitoring system of the pitch system.
[0045] Step 102: Process the characteristic data using a Kalman filter to determine an observation vector, and determine a residual between the observation vector and a measurement vector of the wind turbine pitch control system, so as to determine an abnormal state based on the residual.
[0046] In this application, a Kalman filter can be pre-established in the system, including the state equation and the observation equation:
[0047] x k =Ax k-1 +Bu k +w k
[0048] z k =Hx k +v k
[0049] Among them, x k is the state estimate at time step k (e.g., blade angle, speed, etc.); A is the covariance matrix, describing the relationship between states; u k is the input vector (i.e., feature data); B is the control input matrix; w k is process noise, usually assumed to be zero mean, white noise, and its covariance matrix is Q; z k is the observation vector (e.g., sensor measurements) at time step k; H is the observation matrix, describing the relationship between the state vector and the observations; v kis the observation noise, which is usually assumed to be zero-mean, white noise, and its covariance matrix is R.
[0050] The process of using the Kalman filter to process the feature data and determine the observation vector is as follows:
[0051] 1. Initialize the state estimate and covariance matrix of the Kalman filter.
[0052] 2. For each time step, perform the following operations: 1) predict the state, use the dynamic model of the system to predict the next state; 2) measure the prediction, use the Kalman filter to predict the observation vector; 3) calculate the residual, calculate the residual between the predicted observation vector and the actual measurement vector; 4) if the residual is greater than the preset threshold, determine that the abnormal state is an abnormality.
[0053] 3. Continuously monitor the system status by repeating the above process.
[0054] Optionally, the residual is used to update the state estimation and covariance matrix of the Kalman filter, thereby ensuring the accuracy of the Kalman filter and further improving the accuracy of fault diagnosis.
[0055] Optionally, there are multiple Kalman filters, and different faults correspond to different Kalman filters. For example, different Kalman filters are used to determine the abnormal state of the pitch motor, pitch bearing, reducer, and controller. Each Kalman filter is used to determine the abnormal state of different faults respectively, and when any abnormal state is a fault, the abnormal state of the pitch system is determined to be a fault.
[0056] Step 103: When the abnormal state is a preset state, the characteristic data is input into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model.
[0057] The preset state is an abnormality. The first fault type includes a pitch motor fault, a pitch bearing fault, a reduction gearbox fault, a controller fault, etc., which is not limited in this application.
[0058] In the present application, a fault diagnosis model may be pre-trained and acquired, and then the fault diagnosis model may be used to predict the first fault type of the wind turbine pitch control system.
[0059] The process of training and acquiring the fault diagnosis model is as follows:
[0060] 1. Create a data set, establish a nonlinear dynamic model or finite element model corresponding to the wind turbine pitch system, and use the nonlinear dynamic model or finite element model to simulate and generate a training data set, wherein the training data set contains multiple sample data, and the sample data contains feature data and labeled fault types.
[0061] 2. Training: Use the training data set to train the fault diagnosis model so that it can identify different fault types.
[0062] 3. Verification: During the training process, a portion of the data is used to verify and adjust the diagnostic performance of the neural network model to ensure the effectiveness of the model.
[0063] 4. Testing: Use another part of the data that was not used in the training or validation process to test the diagnostic ability of the neural network model and evaluate its generalization ability.
[0064] It can be understood that the computational overhead of the Kalman filter is relatively small. When the abnormal state is a preset state, the fault diagnosis model is used to determine the first fault type of the wind turbine pitch system, which reduces the computational burden and is conducive to improving the efficiency of fault diagnosis. Thus, real-time fault detection and diagnosis are supported, which is essential for timely maintenance and optimization of wind turbine operation.
[0065] In the present application, characteristic data of the wind turbine pitch system is obtained. The characteristic data is processed using a Kalman filter to determine an observation vector, and the residual between the observation vector and the measurement vector of the wind turbine pitch system is determined to determine the abnormal state based on the residual. In the case where the abnormal state is a fault, the characteristic data is input into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model. Thus, after the abnormal state of the wind turbine pitch system is determined using a Kalman filter, when the abnormal state is a preset state, the fault diagnosis model is used to determine the first fault type of the wind turbine pitch system, thereby reducing the computational burden, thereby improving the efficiency of fault diagnosis while realizing fault diagnosis of the wind turbine pitch system.
[0066] Figure 2 A schematic flow chart of another method for diagnosing a fault in a wind turbine pitch system provided in an embodiment of the present application.
[0067] like Figure 2 As shown, the fault diagnosis method of the wind turbine pitch control system includes:
[0068] Step 201, obtaining characteristic data of a wind turbine pitch control system.
[0069] Step 202, using a Kalman filter to process the characteristic data to determine an observation vector, and determine the residual between the observation vector and the measurement vector of the wind turbine pitch system, so as to determine the abnormal state based on the residual, wherein there are multiple Kalman filters, each Kalman filter is used to determine the abnormal state of different faults.
[0070] Step 203: when the abnormal state is a preset state, input the characteristic data into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model.
[0071] In the present application, the specific implementation process of steps 201 to 203 can be found in the detailed description of any embodiment of the present application and will not be repeated here.
[0072] Step 204 , when the first fault type does not match the fault corresponding to the abnormal state, generate alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state.
[0073] In the present application, the first fault type includes a pitch motor fault, a pitch bearing fault, a speed reducer fault, a controller fault, etc. When the first fault type is a pitch motor fault, and the abnormal state of other equipment other than the pitch motor is determined to be a fault through the Kalman filter, it is determined that the first fault type does not match the fault corresponding to the abnormal state. Similarly, when the first fault type is other fault types, and the abnormal state of equipment other than the equipment corresponding to other fault types is determined to be a fault through the Kalman filter, it is determined that the first fault type does not match the fault corresponding to the abnormal state.
[0074] In the case that the first fault type does not match the fault corresponding to the abnormal state, an alarm message can be generated based on the characteristic data, the first fault type and the fault corresponding to the abnormal state, and the alarm message can be sent to related devices. Experts can determine the actual fault in the wind turbine pitch system based on the alarm information, thereby improving the accuracy of fault diagnosis.
[0075] Step 205: receiving feedback information of the alarm information, wherein the feedback information includes the second fault type confirmed by the expert.
[0076] In this application, after receiving the alarm information, the expert determines the second fault type actually occurring in the variable pitch system through analysis, and inputs the second fault type in the client, and the client generates a feedback message of the alarm information based on the second fault type, and sends the feedback message to the diagnostic device. Thus, the diagnostic device can receive the feedback information of the alarm information.
[0077] Step 206: When the second fault type is different from the first fault type, the fault diagnosis model is modified.
[0078] In the present application, when the second fault type is different from the first fault type, it indicates that the fault diagnosis model is inaccurate. Therefore, the fault diagnosis model can be corrected based on the characteristic data and the difference between the first fault type and the second fault type, thereby improving the accuracy of the fault diagnosis model.
[0079] In the present application, characteristic data of a wind turbine pitch system is obtained. The characteristic data is processed by a Kalman filter to determine an observation vector, and the residual between the observation vector and the measurement vector of the wind turbine pitch system is determined to determine the abnormal state based on the residual, wherein there are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of different faults. When the abnormal state is a preset state, the characteristic data is input into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model. When the first fault type does not match the fault corresponding to the abnormal state, an alarm message is generated based on the characteristic data, the first fault type and the fault corresponding to the abnormal state. Feedback information of the alarm message is received, wherein the feedback information includes a second fault type confirmed by an expert. When the second fault type is different from the first fault type, the fault diagnosis model is corrected. Thus, by introducing the second fault type confirmed by the expert, the fault diagnosis model is further corrected, thereby improving the accuracy of the fault diagnosis model.
[0080] In order to implement the above embodiment, the embodiment of the present application also proposes a fault diagnosis device for a wind turbine pitch control system. Figure 3 A schematic structural diagram of a fault diagnosis device for a wind turbine pitch system provided in an embodiment of the present application.
[0081] like Figure 3 As shown, the fault diagnosis device 300 of the wind turbine pitch control system includes:
[0082] An acquisition module 310 is used to acquire characteristic data of a wind turbine pitch control system;
[0083] A detection module 320 is used to process the characteristic data using a Kalman filter to determine an observation vector, and to determine a residual between the observation vector and a measurement vector of the wind turbine pitch system, so as to determine an abnormal state based on the residual;
[0084] The diagnosis module 330 is used to input the characteristic data into the fault diagnosis model to obtain the first fault type output by the fault diagnosis model when the abnormal state is a preset state.
[0085] In a possible implementation of another aspect of the present application, there are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of a different fault.
[0086] In a possible implementation manner of another aspect of the present application, a first correction module is further included, which is used to:
[0087] In the case where the first fault type does not match the fault corresponding to the abnormal state, generating alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state;
[0088] receiving feedback information of the alarm information, wherein the feedback information includes the second fault type confirmed by the expert;
[0089] When the second fault type is different from the first fault type, the fault diagnosis model is modified.
[0090] In a possible implementation of an embodiment of another aspect of the present application, the fault types include faults of a pitch motor, a pitch bearing, and a reduction gearbox.
[0091] In a possible implementation manner of another aspect of the present application, a second correction module is further included, which is used to:
[0092] The residual is used to update the state estimate and covariance matrix of the Kalman filter.
[0093] It should be noted that the above explanation of the embodiment of the fault diagnosis method for the wind turbine pitch system is also applicable to the fault diagnosis device for the wind turbine pitch system of this embodiment, so it will not be repeated here.
[0094] In the present application, characteristic data of the wind turbine pitch system is obtained. The characteristic data is processed using a Kalman filter to determine an observation vector, and the residual between the observation vector and the measurement vector of the wind turbine pitch system is determined to determine the abnormal state based on the residual. In the case where the abnormal state is a fault, the characteristic data is input into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model. Thus, after the abnormal state of the wind turbine pitch system is determined using a Kalman filter, when the abnormal state is a preset state, the fault diagnosis model is used to determine the first fault type of the wind turbine pitch system, thereby reducing the computational burden, thereby improving the efficiency of fault diagnosis while realizing fault diagnosis of the wind turbine pitch system.
[0095] In order to implement the above embodiment, the embodiment of the present application also proposes a computer device, including a processor and a memory;
[0096] The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the fault diagnosis method of the wind turbine pitch control system as in the above-mentioned embodiment.
[0097] In order to implement the above embodiment, the embodiment of the present application further proposes a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the fault diagnosis method of the wind turbine pitch system as in the above embodiment is implemented.
[0098] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.
Claims
1. A fault diagnosis method for a wind turbine pitch control system, characterized in that: include: Obtain characteristic data of wind turbine pitch control system; Processing the characteristic data using a Kalman filter to determine an observation vector, and determining a residual between the observation vector and a measurement vector of the wind turbine pitch control system, so as to determine an abnormal state based on the residual; In a case where the abnormal state is a preset state, the characteristic data is input into a fault diagnosis model to obtain a first fault type output by the fault diagnosis model.
2. The method according to claim 1, characterized in that There are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of a different fault.
3. The method according to claim 2, characterized in that Also includes: In the case where the first fault type does not match the fault corresponding to the abnormal state, generating alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state; receiving feedback information of the alarm information, wherein the feedback information includes a second fault type confirmed by an expert; When the second fault type is different from the first fault type, the fault diagnosis model is modified.
4. The method according to claim 1, characterized in that The fault types include faults of the pitch motor, pitch bearing, and reduction gearbox.
5. The method according to claim 1, characterized in that Also includes: The residual is used to update the state estimation and covariance matrix of the Kalman filter.
6. A fault diagnosis device for a wind turbine pitch control system, characterized in that: include: An acquisition module is used to acquire characteristic data of a wind turbine pitch control system; A detection module, configured to process the characteristic data using a Kalman filter to determine an observation vector, and determine a residual between the observation vector and a measurement vector of the wind turbine pitch control system, so as to determine an abnormal state based on the residual; The diagnosis module is used to input the characteristic data into a fault diagnosis model when the abnormal state is a preset state, and obtain a first fault type output by the fault diagnosis model.
7. The device according to claim 6, characterized in that There are multiple Kalman filters, and each Kalman filter is used to determine the abnormal state of a different fault.
8. The device according to claim 7, characterized in that Also included is a first correction module, for: In the case where the first fault type does not match the fault corresponding to the abnormal state, generating alarm information based on the characteristic data, the first fault type and the fault corresponding to the abnormal state; receiving feedback information of the alarm information, wherein the feedback information includes a second fault type confirmed by an expert; When the second fault type is different from the first fault type, the fault diagnosis model is modified.
9. The device according to claim 6, characterized in that The fault types include faults of the pitch motor, pitch bearing, and reduction gearbox.
10. The device according to claim 6, characterized in that Also included is a second correction module, for: The residual is used to update the state estimation and covariance matrix of the Kalman filter.
