Fan equivalent fatigue load prediction method and system
By establishing a proxy model for wind participation in equivalent fatigue loads and a Python Flask web application, the problems of large computational complexity and low accuracy in wind turbine equivalent fatigue load prediction are solved, achieving fast and accurate fatigue load prediction, reducing computational costs and improving design efficiency.
Patent Information
- Application Number
- CN202510523211.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-24
- Publication Date
- 2025-09-19
AI Technical Summary
Existing technologies for predicting equivalent fatigue loads on wind turbines are computationally intensive, time-consuming, and labor-intensive, and are unable to accurately assess the fatigue load at each turbine location, resulting in conservative designs or excessive computational effort.
A proxy model for wind participation in equivalent fatigue loads was established, and interpolation calculations were performed using the database to quickly obtain the equivalent fatigue loads for each aircraft position. A web application was built using Python Flask to achieve convenient multi-terminal access and efficient calculations.
It achieves fast and accurate prediction of wind turbine equivalent fatigue load, reduces calculation costs, improves design accuracy and efficiency, and avoids the problems of poor cross-platform compatibility and high deployment and maintenance of ordinary GUI interfaces.
Smart Images
Figure CN120671572A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of fan load prediction, and in particular relates to a fan equivalent fatigue load prediction method and system. Background Art
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] Wind turbines (abbreviated as wind turbines) are slender structures. Their cyclical operation and constantly changing external loads often produce periodic vibrations, which accelerate structural fatigue damage. Generally, wind turbines have a design lifespan exceeding twenty years, spending the vast majority of their time in operation and generating electricity. Consequently, they are subject to significant fatigue loads. Therefore, the size, weight, and cost of key wind turbine components are closely related to the fatigue loads. However, wind turbine fatigue loads are not determined by a single factor but rather by the combined effects of numerous factors. These factors, such as rotor diameter, tower height, air density, turbulence, wind shear, inflow angle, average wind speed, pitch control, and yaw, all significantly influence wind turbine fatigue. This multiplicity of factors significantly increases the difficulty of accurately assessing wind turbine fatigue loads, presenting unprecedented challenges to engineering design. Existing simulation software derives the equivalent fatigue loads of wind turbines by simulating numerous operating conditions and performing rainflow statistics on the load time history curves. To accurately calculate the equivalent fatigue load of the fan, designers often need to invest a lot of time and energy to calculate hundreds or thousands of working conditions before they can more accurately evaluate the fatigue design load of the fan, which brings great inconvenience to the designers.
[0004] The existing method for calculating equivalent fatigue loads on wind turbines involves enveloping the wind parameters at the wind turbine site (air density, turbulence, inflow angle, wind shear, and average annual wind speed). These enveloping wind parameters are then used to calculate hundreds or thousands of load conditions. Rainflow statistics are then applied to the load time history curves, combined with the wind speed Weibull distribution to ultimately derive the equivalent fatigue loads on the wind turbines. While this method uses enveloping wind parameters at the wind turbine site to calculate, it ensures the safety of the entire turbine. However, the calculated equivalent fatigue loads are excessive, leading to a conservative turbine design and high component costs. Furthermore, it is impossible to accurately determine the fatigue loads at each turbine site. Obtaining the exact fatigue loads at each turbine site requires performing hundreds or thousands of load condition calculations for each site, which is computationally intensive, time-consuming, and labor-intensive. Summary of the Invention
[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention proposes a method and system for predicting equivalent fatigue loads of wind turbines. First, a proxy model of wind parameters and equivalent fatigue loads is established and stored in a database. Then, the wind parameters of each machine position in the wind farm are used as input, and interpolation calculations are performed using the proxy model in the database, so as to quickly obtain the equivalent fatigue load of each machine position, thereby realizing rapid and accurate prediction of the equivalent fatigue load of the wind turbine.
[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: In a first aspect, a method for predicting equivalent fatigue load of a wind turbine is disclosed, comprising: Obtain the complete simulation model of the wind turbine generator set, wind turbine operating condition table, and environmental parameters; According to the fan operating condition table and environmental parameters, the whole machine simulation model is pre-processed to obtain the time series load of each channel under different operating conditions; The time series load of each channel under different working conditions is post-processed to obtain the first fatigue load; Fit the first fatigue load to obtain the life cycle equivalent fatigue load proxy model; An equivalent fatigue load is predicted for each wind speed of the wind turbine to be predicted based on the life cycle equivalent fatigue load proxy model to obtain a second fatigue load; The equivalent fatigue load of the wind turbine to be predicted is calculated based on the second fatigue load.
[0007] Furthermore, after fitting the first fatigue load to obtain the life cycle equivalent fatigue load proxy model, the method further includes: The first fatigue load is used to verify the fitted life cycle equivalent fatigue load proxy model. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
[0008] In a second aspect, a wind turbine equivalent fatigue load prediction system is disclosed, comprising: A front-end interface input module is configured to obtain a complete simulation model of a wind turbine generator set, a wind turbine operating condition table, and environmental parameters; The back-end calculation and model training module is configured to: pre-process the whole machine simulation model according to the wind turbine operating condition table and environmental parameters to obtain the time series load of each channel under different operating conditions; post-process the time series load of each channel under different operating conditions to obtain the first fatigue load; and fit the first fatigue load to obtain the life cycle equivalent fatigue load proxy model; A site wind parameter analysis module is configured to: perform equivalent fatigue load prediction for each wind speed of the wind turbine to be predicted based on a life cycle equivalent fatigue load proxy model to obtain a second fatigue load; The site point fatigue load prediction module is configured to calculate the equivalent fatigue load of the wind turbine to be predicted based on the second fatigue load.
[0009] Furthermore, the wind turbine equivalent fatigue load prediction system further includes: The model verification and comparison module is configured to: use the first fatigue load to verify the life cycle equivalent fatigue load proxy model obtained by fitting. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
[0010] In a third aspect, an electronic device is disclosed, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps of the above-mentioned wind turbine equivalent fatigue load prediction method are completed.
[0011] In a fourth aspect, a computer-readable storage medium is disclosed for storing computer instructions. When the computer instructions are executed by a processor, the steps of the above-mentioned wind turbine equivalent fatigue load prediction method are completed.
[0012] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a method and system for predicting equivalent fatigue loads of wind turbines. By using Python Flask to build a Web application framework, a fatigue load prediction model training user input interface in the form of a website is created. This method avoids the shortcomings of ordinary GUI interfaces, such as poor cross-platform compatibility, high deployment and maintenance costs, and insufficient network support, while achieving the advantages of convenient multi-terminal access, centralized version management, and high scalability.
[0013] The present invention creates a table database containing environmental parameters, unit simulation model information, and fatigue loads. It generates equivalent fatigue loads for life cycle samples and stores them in a MySQL database, improving the security, reliability, and efficiency of data storage.
[0014] The present invention uses a timed task mode to automatically submit HPC pre-processing or post-processing, and realizes automatic and efficient calculation and interactive connection between processing tasks and HPC.
[0015] The present invention obtains site point wind parameters as input. Usually, the amount of data provided by wind resources, such as the aircraft turbulence matrix, wind frequency matrix, wind shear, inflow angle, and air density, is very large. For ease of management, pkl files are used to serialize and deserialize Python objects, making them easy to store, transmit, and share.
[0016] The present invention performs linear fitting on DEL using turbulence, wind shear and inflow angle data, and uses air density to perform piecewise linear function modeling on DEL, thereby improving the accuracy of the prediction model.
[0017] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0019] Figure 1 This is a flow chart of the method for predicting equivalent fatigue load of a wind turbine described in Example 1 of the present invention. DETAILED DESCRIPTION
[0020] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.
[0021] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.
[0022] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.
[0023] Example 1 In one or more embodiments, a method for predicting equivalent fatigue load of a wind turbine is disclosed, such as Figure 1 As shown, the following steps are included: Step 1: Obtain the whole machine simulation model, wind turbine operating condition table, and environmental parameters of the wind turbine generator set.
[0024] Step 101: In this embodiment, Python Flask is used to build a Web application framework, which includes an input port for the whole-machine simulation model of the wind turbine generator set and an input port for the wind turbine operating condition table; The input ports of the wind turbine simulation model include the Bladed model input and the model controller input. The Bladed model includes the modeling of various components of the wind turbine, such as blades, towers, transmission chains, hubs, etc., and is a file with the extension .prj. The model controller is a file that controls the operation of the wind turbine during the wind turbine simulation process, and is usually composed of two files: .dll and .xml.
[0025] The wind turbine operating table is a preset operating table that defines the combination of wind turbines under different wind conditions, yaw angles, inflow angles, wind shear, etc. Each row in the operating table represents a sub-operating condition. Each row of data includes: sub-operating condition number Counter, operating condition level Case, simulation name Run Name, hub center wind speed Vhub_turb, turbulence standard deviation , X-direction turbulence intensity Ti_x, Y-direction turbulence intensity Ti_y, Z-direction turbulence intensity Ti_z, inflow angle In_flow, yaw error yaw_turb, wind file name Wind name, wind shear Shear, simulation output filtering time t out , simulation time t Sim .
[0026] Step 102: Obtain preset environmental parameters. Environmental parameters that typically affect unit fatigue loads include turbulence, wind shear, inflow angle, and air density. Based on the predicted wind field environmental parameters, the approximate ranges of turbulence, wind shear, inflow angle, and air density can be determined, along with baseline levels for the environmental parameters. In this embodiment, turbulence is set to 3 levels, wind shear is set to 3 levels, inflow angle is set to 3 levels, and air density is set to 7 levels. This results in a total of 3*3*3*7 = 189 operating condition groups to be simulated.
[0027] Step 103: In this embodiment, the entire wind turbine simulation model is simplified to determine the wind turbine simulation model. As can be seen from step 102, a single wind turbine model may require hundreds of simulated operating conditions, requiring significant storage space. Therefore, this embodiment simplifies the model before calculation, removing output variables unrelated to fatigue loads to reduce storage space required for calculations.
[0028] Specifically, the simplified unit simulation model means that the load output only includes five key positions: blade root, rotating hub, fixed hub, tower top, and tower bottom.
[0029] Step 2: Pre-process the whole machine simulation model according to the fan operating condition table and environmental parameters to obtain the time series load of each channel under different operating conditions.
[0030] Step 201: Create a table database of environmental parameters, unit simulation model information, and fatigue loads.
[0031] Among them, each row in the unit simulation model information database stores a simulation model, a simulation model represents a unit, and different rows represent different units; the first column represents the model ID number, and the second column represents the model path. After the Web application framework obtains the whole machine simulation model file, it stores the simulation model ID number and path in the database.
[0032] The environmental parameters and fatigue loads form a separate database called the fatigue load database. Initially, a blank table with a header is created, and data is stored in this table after the simulation is complete. Each row in the fatigue load database represents the equivalent fatigue load del for different values of turbulence Iref, air density rho, wind shear, inflow angle inflow, wind speed ws, and equivalent fatigue load m.
[0033] Simulating hundreds of working conditions requires storing a large amount of data. Using Excel to store data will significantly reduce the running speed, and may even cause freezes and crashes, which directly affects work efficiency. Excel is difficult to meet the needs of situations where real-time processing and querying of data are required. Since Excel files can be easily copied, tampered with, and deleted, data security cannot be effectively guaranteed. Excel has hidden dangers in terms of data storage security. When sensitive data is involved, using Excel for storage may lead to risks such as data leakage. This embodiment uses a more suitable data storage solution such as MySQL database to improve the security, reliability, and efficiency of data storage.
[0034] Step 202: Generate a pre-processing DLC1.2 working condition and a corresponding first job list joblist1.
[0035] Based on the environmental parameter levels set in step 102, DLC1.2 operating conditions (.$PJ and .in files) are generated, and then joblist1 is generated based on the DLC1.2 operating conditions. Generating operating conditions refers to generating .$PJ model files and .in files for unit simulation under different wind conditions.
[0036] The IEC61400-1 standard stipulates that the operating conditions that affect unit fatigue loads are DLC1.2, DLC2.4, DLC3.1, DLC4.1, and DLC6.4. DLC1.2, representing normal power generation, accounts for the highest percentage of time and is the primary condition affecting unit fatigue loads. To reduce the nonlinear effects of DLC2.4, DLC3.1, DLC4.1, and DLC6.4 on the final fatigue loads and to improve simulation efficiency, this example only considers the impact of DLC1.2 on unit fatigue loads.
[0037] According to different working conditions composed of different wind conditions specified in IEC61400-1, the first working list is a file including the model folder path and the Bladed pre-processing program path under different wind conditions.
[0038] In this embodiment, the generated working conditions and generated work list are generated by running a python script.
[0039] Step 203: Submit the first worklist to the high-performance computing scheduling system (HPC) to automatically identify the calculation path. The Bladed program automatically calculates these working condition files. The calculation results are the load data at each time step, that is, the time series load (time series load) calculation results of each channel of each working condition automatically generate pre-processing result files (.% and .$ files) for storage.
[0040] In this embodiment, a Python scheduled task is used to automatically submit joblist1 to the HPC for calculation, thereby solving the technical difficulties of submitting joblist1 to the HPC for calculation and interactively connecting with the HPC. Specifically, the difficulty of interactive connection lies in the fact that hundreds of joblists usually need to be manually submitted to the HPC for calculation one by one, and manual observation is required after the pre-processing calculation is completed before post-processing calculation can be performed, and then the calculated fatigue load can be manually extracted to the database. The scheduled task can automatically determine when the pre-processing calculation is completed, and after confirming that the pre-processing calculation is completed, it can automatically submit the post-processing calculation after waiting for a preset time (for example, 10 seconds), and automatically extract the equivalent fatigue load to the database after confirming that the post-processing calculation is completed.
[0041] Step 204: Check the completion status of the pre-processing task calculations. During the HPC calculation process, some conditions may fail to complete the calculation or return incomplete data due to hardware failure or software interaction failure. In such cases, it is necessary to regenerate a new joblist for these incomplete conditions or incomplete data and submit it to the HPC for calculation again until all tasks are completed.
[0042] Step 3: Post-process the time series load of each channel under different working conditions to obtain the first fatigue load.
[0043] Step 301: Generate a post-processing file and a corresponding second job list joblist2.
[0044] The generating of the post-processing file specifically includes: using Python to automatically generate the Bladed post-processing file, the post-processing file contains the contents specified by the Bladed software, such as the occurrence time of each working condition required for Bladed rainflow counting, the path of each working condition, the load channel to be calculated, etc.
[0045] The second work list is a file including a post-processing folder path and a Bladed post-processing program path.
[0046] In this embodiment, this scheme only considers the fatigue loads at five key positions: blade root, rotating hub, stationary hub, yaw (tower top), and tower bottom, and focuses on predicting and evaluating the equivalent fatigue loads of m=4 and m=10.
[0047] Step 302: Submit the second worklist to the high-performance computing scheduling system (HPC) to automatically identify the calculation path. The Bladed model software automatically calculates these post-processing files and performs rain flow counting on the time series load obtained from the completed pre-processing to obtain the first fatigue load.
[0048] In this embodiment, a Python scheduled task is used to automatically submit joblist2 to the HPC for calculation, which also solves the technical difficulties of submitting joblist2 to the HPC for calculation and interacting with the HPC.
[0049] Step 303: Check the completion of post-processing task calculations. Similar to step 204, ensure that all tasks are fully calculated.
[0050] Step 4: Fit the first fatigue load to obtain a life cycle equivalent fatigue load proxy model.
[0051] Step 401: Extract the first fatigue load for each wind speed, each turbulence, each air density, each wind shear, each inflow angle, and each load channel, and save it to the table database constructed in step 201 (the equivalent fatigue load column for the Wöhler index m value of the material fatigue property). Generate the life cycle sample equivalent fatigue load DEL|(v, rho, channel), where DEL represents the equivalent fatigue load, v represents the wind speed, rho represents the air density, and channel represents the load channel. The life cycle sample equivalent fatigue load includes the equivalent fatigue load del for different turbulence Iref, air density rho, wind shear, inflow angle inflow, wind speed ws, and equivalent fatigue load m values.
[0052] Step 402: Delete the completed task. After the task has been completed and all data has been stored in the database, the calculation results can be deleted to save disk capacity.
[0053] Step 403: Perform model training to build a life cycle equivalent fatigue load proxy model. Specifically: Extract turbulence, wind shear, and inflow angle data from the database; The life cycle sample equivalent fatigue load is linearly regressed using turbulence, wind shear, and inflow angle to obtain the coefficients a, b, c, and d of the life cycle sample equivalent fatigue load DEL|(v, rho, channel). The life cycle equivalent fatigue load proxy model is constructed as shown in the following formula: DEL| (v, rho, channel)_proxy=a*Ti+b*inflow+c*shear+d Step 5: Use the first fatigue load to verify the fitted life cycle equivalent fatigue load proxy model. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
[0054] In this embodiment, the verification load, i.e., the first fatigue load DEL|(v, rho, channel) stored in the database after detailed calculation in step 401, is compared with the equivalent fatigue load DEL|(v, rho, channel)_proxy of the proxy model. If the error is controlled within 1%, it is considered that the DEL proxy model can meet the requirements.
[0055] Step 6: Perform equivalent fatigue load prediction on each wind speed of the wind turbine to be predicted based on the life cycle equivalent fatigue load proxy model to obtain a second fatigue load.
[0056] Step 601: Obtain the environmental parameters of the wind turbine site to be predicted as input. Wind resources typically provide wind parameters that affect fatigue loads, including the turbulence matrix, wind frequency matrix, wind shear, inflow angle, and air density. These wind parameters are stored in separate Excel spreadsheets or txt files. The volume of data corresponding to these wind parameters is large when there are many wind turbine sites. In this embodiment, a PKL-formatted file is used as a cache mechanism to store wind parameter data for each wind turbine site, achieving persistent storage and improving data loading speed.
[0057] Step 602: Input the site wind parameters (turbulence, wind shear, inflow angle) into the life cycle equivalent fatigue load proxy model to obtain the proxy model equivalent fatigue load DEL|(v, rho, channel)_proxy at each wind speed and air density. Step 603: Perform quadratic polynomial fitting on three adjacent air densities at different levels using Lagrange or Newton interpolation method to obtain an analytical solution.
[0058] Then, piecewise quadratic polynomial modeling is adopted to obtain the analytical proxy model according to different air density segments at each wind speed, and the second fatigue load is calculated, that is, the equivalent fatigue load DEL|(v,channel)_proxy of the analytical proxy model at each wind speed. This can finely characterize the nonlinear relationship between air density and equivalent fatigue load, and improve the accuracy of proxy model prediction.
[0059] The specific modeling of piecewise quadratic polynomial is: Setting the number of segments according to the air density level in step 102; DEL|(v, rho, channel) performs Lagrange interpolation or Newton interpolation on three sets of adjacent air densities rho to obtain the analytical solutions e, f, and g: DEL|(v, channel)_proxy=e*rho^2+f*rho+g where DEL|(v, channel)_proxy is the analytical proxy model equivalent fatigue load for each wind speed, and rho is the air density.
[0060] The piecewise quadratic polynomial fit has a strong ability to capture nonlinear features. Each discrete interval is fitted independently, avoiding the smoothing loss of local features caused by the global linear model. It has the characteristics of high model accuracy, strong local approximation ability, good numerical stability, and high computational efficiency.
[0061] Step 604: Substitute the wind parameter (air density) of the site to be predicted into step 603, find the corresponding segment and perform interpolation calculation to obtain the equivalent fatigue load DEL|(v, channel)_temp of the analytical proxy model for each wind speed.
[0062] Step 7: Calculate the equivalent fatigue load of the wind turbine to be predicted based on the second fatigue load.
[0063] Step 701: Accumulate the equivalent fatigue load of the analytical proxy model considering the wind speed probability at each wind speed, as shown in the following formula: DEL|(v, channel)=(sum(DEL|(v, channel)_temp^m*probability))^(1 / m) Where DEL|(v, channel) is the wind speed fatigue load, m is the Wöhler index of the material fatigue property, and probability is the wind speed probability.
[0064] Since the probability of occurrence of each wind speed segment is different, the probability of each wind speed segment is considered to calculate the equivalent fatigue load to ensure the accuracy of the calculation.
[0065] Step 702: Accumulate all wind speed fatigue loads of each channel, and the expression is: DEL|(channel)=(sum(DEL|(v, channel)^m))^(1 / m) Where DEL|(channel) is the predicted value of the equivalent fatigue load.
[0066] Step 703: Output the equivalent fatigue load prediction values of all points and all channels.
[0067] Among them, the load channel includes the blade root, rotating hub, fixed hub, tower top and tower bottom as key positions.
[0068] In this embodiment, the DEL prediction values under all load channels of all machine sites are output to an Excel spreadsheet for user viewing.
[0069] Example 2 In one or more embodiments, a wind turbine equivalent fatigue load prediction system is disclosed, specifically comprising: A front-end interface input module is configured to obtain a complete simulation model of a wind turbine generator set, a wind turbine operating condition table, and environmental parameters; The back-end calculation and model training module is configured to: pre-process the whole machine simulation model according to the wind turbine operating condition table and environmental parameters to obtain the time series load of each channel under different operating conditions; post-process the time series load of each channel under different operating conditions to obtain the first fatigue load; and fit the first fatigue load to obtain the life cycle equivalent fatigue load proxy model; A site wind parameter analysis module is configured to: perform equivalent fatigue load prediction for each wind speed of the wind turbine to be predicted based on a life cycle equivalent fatigue load proxy model to obtain a second fatigue load; The site point fatigue load prediction module is configured to calculate the equivalent fatigue load of the wind turbine to be predicted based on the second fatigue load.
[0070] Furthermore, the wind turbine equivalent fatigue load prediction system further includes: The model verification and comparison module is configured to: use the first fatigue load to verify the life cycle equivalent fatigue load proxy model obtained by fitting. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
[0071] Example 3 This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps of the above-mentioned wind turbine equivalent fatigue load prediction method are completed.
[0072] Example 4 This embodiment provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the above-mentioned wind turbine equivalent fatigue load prediction method are completed.
[0073] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0074] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.
[0075] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide the functions for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.
[0076] The descriptions of the various embodiments in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0077] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.
Claims
1. A method for predicting equivalent fatigue load of a wind turbine, characterized in that: include: Obtain the complete simulation model of the wind turbine generator set, wind turbine operating condition table, and environmental parameters; According to the fan operating condition table and environmental parameters, the whole machine simulation model is pre-processed to obtain the time series load of each channel under different operating conditions; The time series load of each channel under different working conditions is post-processed to obtain the first fatigue load; Fit the first fatigue load to obtain the life cycle equivalent fatigue load proxy model; An equivalent fatigue load is predicted for each wind speed of the wind turbine to be predicted based on the life cycle equivalent fatigue load proxy model to obtain a second fatigue load; The equivalent fatigue load of the wind turbine to be predicted is calculated based on the second fatigue load.
2. A method for predicting equivalent fatigue load of a wind turbine according to claim 1, characterized in that: After fitting the first fatigue load to obtain the life cycle equivalent fatigue load proxy model, the method further includes: The first fatigue load is used to verify the fitted life cycle equivalent fatigue load proxy model. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
3. A method for predicting equivalent fatigue load of a wind turbine according to claim 1, characterized in that: A Web application framework is used to obtain the whole machine simulation model, fan operating condition table, and environmental parameters, and to build a unit simulation model information database and a fatigue load database.
4. A method for predicting equivalent fatigue load of a wind turbine according to claim 1, characterized in that: The step of fitting the first fatigue load to obtain a life cycle equivalent fatigue load proxy model includes: Extract turbulence, wind shear, and inflow angle data from the fatigue load database; The first fatigue load in the fatigue load database is linearly regressed using turbulence, wind shear, and inflow angle to obtain the coefficient of the equivalent fatigue load of the life cycle sample and construct a life cycle equivalent fatigue load proxy model.
5. The method for predicting equivalent fatigue load of a wind turbine according to claim 1, wherein: The method of performing equivalent fatigue load prediction for each wind speed of the wind turbine to be predicted based on the life cycle equivalent fatigue load proxy model to obtain a second fatigue load includes: Obtain site point wind parameters and input them into the life cycle equivalent fatigue load proxy model to obtain the proxy model equivalent fatigue load at each wind speed and air density; The Lagrange or Newton interpolation method is used to fit the air density at three different levels to obtain the analytical solution. The piecewise quadratic polynomial modeling is adopted to obtain the analytical proxy model according to different air density segments at each wind speed, and the second fatigue load is calculated.
6. A method for predicting equivalent fatigue load of a wind turbine according to claim 1, characterized in that: The step of calculating the equivalent fatigue load of the wind turbine to be predicted based on the second fatigue load specifically includes: The equivalent fatigue load of the analytical proxy model considering the wind speed probability at each wind speed is accumulated to obtain the wind speed fatigue load; The wind speed fatigue load of each channel is accumulated to obtain the equivalent fatigue load prediction value.
7. A wind turbine equivalent fatigue load prediction system, characterized in that: include: A front-end interface input module is configured to obtain a complete simulation model of a wind turbine generator set, a wind turbine operating condition table, and environmental parameters; The back-end calculation and model training module is configured to: pre-process the whole machine simulation model according to the wind turbine operating condition table and environmental parameters to obtain the time series load of each channel under different operating conditions; post-process the time series load of each channel under different operating conditions to obtain the first fatigue load; and fit the first fatigue load to obtain the life cycle equivalent fatigue load proxy model; A site wind parameter analysis module is configured to: perform equivalent fatigue load prediction for each wind speed of the wind turbine to be predicted based on a life cycle equivalent fatigue load proxy model to obtain a second fatigue load; The site point fatigue load prediction module is configured to calculate the equivalent fatigue load of the wind turbine to be predicted based on the second fatigue load.
8. A wind turbine equivalent fatigue load prediction system according to claim 7, characterized in that: Also includes: The model verification and comparison module is configured to: use the first fatigue load to verify the life cycle equivalent fatigue load proxy model obtained by fitting. When the error between the first fatigue load and the equivalent fatigue load obtained by the proxy model is within a preset range, the life cycle equivalent fatigue load proxy model meets the prediction requirements.
9. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the method for predicting equivalent fatigue load of a wind turbine according to any one of claims 1 to 6 is completed.
10. A computer-readable storage medium, characterized in that Used to store computer instructions, which, when executed by a processor, complete the wind turbine equivalent fatigue load prediction method according to any one of claims 1 to 6.