Wind turbine blade fatigue damage prediction method, computer device and readable storage medium

Abstract

The present application provides a method for predicting fatigue damage of wind turbine blades, a computer device and a readable storage medium, and relates to the field of wind power generation technology. The present application obtains the actual unit operation data of each wind turbine to be tested in the target wind farm in the current detection cycle, and then calls the target damage prediction model corresponding to the actual working condition type of the wind turbine to be tested in the current detection cycle among the multiple blade fatigue damage prediction models matched with the target wind farm, and predicts blade fatigue damage based on the actual unit operation data of the wind turbine to be tested, and obtains the fatigue damage estimate of the blade fatigue damage concentration area of ​​the wind turbine to be tested in the current detection cycle, thereby effectively reducing the cost of monitoring the health status of wind turbine blades, and achieving high-precision blade fatigue damage real-time prediction effects for each wind turbine in the same wind farm.

Description

Technical Field

[0001] The present application relates to the technical field of wind power generation, and in particular to a method for predicting fatigue damage of wind turbine blades, a computer device and a readable storage medium. Background Art

[0002] With the continuous development of science and technology, the application of wind power generation technology has become more and more extensive. Wind farms are often set up in remote mountainous areas, plateaus or coastal areas. In the actual production process of wind farms, wind turbine blades are the core components of wind turbines to capture wind energy. They will be subjected to the coupling effects of bending, torsion, shear and other loads for a long time, and blade fatigue damage and cumulative damage are prone to occur, which seriously affects the production capacity efficiency of the corresponding wind turbines. Therefore, real-time monitoring of the health status of wind turbine blades is an important means to ensure the normal and stable operation of the corresponding wind turbines.

[0003] At present, the mainstream real-time monitoring solution for the health of wind turbine blades in the industry is to install load sensors, audio sensors and other sensor devices for each wind turbine in the wind farm, so as to analyze the wind turbine blade status of the load data or audio data monitored in real time by the installed sensor devices, and determine the health of the wind turbine blades of the corresponding wind turbine. However, it is worth noting that this solution requires the use of a large number of expensive sensor devices, and the overall cost of monitoring the health of wind turbine blades is high. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a wind turbine blade fatigue damage prediction method, computer equipment and readable storage medium, which can predict blade fatigue damage based on wind turbine operating data through a blade fatigue damage prediction model adapted to the wind farm, without spending a lot of cost to install expensive sensor devices for each wind turbine in the wind farm. Therefore, on the basis of effectively reducing the cost of monitoring the health status of wind turbine blades, high-precision real-time prediction of blade fatigue damage can be achieved for each wind turbine in the same wind farm.

[0005] In order to achieve the above purpose, the technical solution adopted in the embodiment of the present application is as follows:

[0006] In a first aspect, the present application provides a method for predicting fatigue damage of a wind turbine blade, the method comprising:

[0007] For each wind turbine to be tested in the target wind farm, actual unit operation data of the wind turbine to be tested in the current detection cycle is obtained;

[0008] According to the actual unit operation data of the wind turbine unit to be tested, identifying the actual operating condition type of the wind turbine unit to be tested in the current detection cycle;

[0009] Selecting a target damage prediction model corresponding to the actual operating condition type of the wind turbine to be tested from a plurality of blade fatigue damage prediction models matching the target wind farm, wherein each blade fatigue damage prediction model corresponds to a turbine operating condition type alone;

[0010] The target damage prediction model is called to predict blade fatigue damage based on actual unit operation data of the wind turbine to be tested, and an estimated fatigue damage value of a blade fatigue damage concentration area of ​​the wind turbine to be tested within a current detection cycle is obtained.

[0011] In an optional embodiment, the method further comprises:

[0012] Calculating the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine in the target wind farm according to the wind speed frequency distribution data of the target wind farm;

[0013] Acquire respectively measured unit operation data of the sample wind turbines in a plurality of historical detection periods under different unit operating conditions in the target wind farm and measured load data related to a blade fatigue damage concentration area;

[0014] For each historical detection cycle, blade fatigue damage calculation is performed according to the load-stress conversion coefficient and the measured load data of the historical detection cycle to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection cycle;

[0015] For each unit operating condition type, the measured unit operation data and actual fatigue damage values ​​corresponding to multiple historical detection cycles corresponding to the unit operating condition type are stored in a fatigue damage database corresponding to the unit operating condition type, and a deep neural network model is trained based on the fatigue damage database corresponding to the unit operating condition type to obtain a blade fatigue damage prediction model that matches the target wind farm and corresponds to the unit operating condition type.

[0016] In an optional embodiment, the step of calculating the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine in the target wind farm according to the wind speed frequency distribution data of the target wind farm comprises:

[0017] According to the wind speed frequency distribution data of the target wind farm, wind power generation simulation is performed on the sample wind turbine in a normal power generation state to obtain load simulation data corresponding to the blade fatigue damage concentration area of ​​the sample wind turbine in multiple simulation detection cycles;

[0018] For each simulation detection cycle, rain flow counting processing is performed on the load simulation data of the simulation detection cycle, and the equivalent load value of the simulation detection cycle is calculated according to the obtained load rain flow counting result;

[0019] According to the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in normal power generation conditions within a single detection cycle, and the equivalent load values ​​corresponding to each of the multiple simulation detection cycles, the load-stress conversion coefficient is solved based on the fatigue damage accumulation principle to obtain the load-stress conversion coefficient of the sample wind turbine in the target wind farm.

[0020] In an optional embodiment, the process of solving the load stress conversion coefficient is expressed by the following equation:

[0021] ;

[0022] in, It is used to represent the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in a normal power generation condition within a single detection cycle, It is used to represent the total number of wind speed segments of the target wind farm, It is used to indicate the target wind farm The probability of occurrence of the wind speed segment is The blade fatigue damage concentration area of ​​the sample wind turbine generator set in the normal power generation condition is in the first The simulated average fatigue damage value in a single test cycle under the action of different wind speed ranges is Used to indicate that the average wind speed is The total number of simulation detection cycles in the wind speed range is Used to indicate that the corresponding average wind speed is in The first wind speed range The simulated fatigue damage value of the simulation test cycle is Used to indicate that the corresponding average wind speed is in The first wind speed range The equivalent load value of a simulation test cycle, It is used to represent the load stress conversion coefficient of the sample wind turbine in the target wind farm. It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to indicate that the fatigue damage concentration area of ​​the blades of the sample wind turbine is The number of stress cycles under action, Used to indicate the duration of a single detection cycle. It is used to represent the load cycle frequency of the sample wind turbine. and are all constants.

[0023] In an optional embodiment, for each historical detection cycle, the step of performing blade fatigue damage calculation according to the load-stress conversion coefficient and the measured load data of the historical detection cycle to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection cycle includes:

[0024] Performing stress conversion processing on the measured load data of the historical detection period according to the load-stress conversion coefficient to obtain actual stress data of the historical detection period;

[0025] Perform rain flow counting processing on the actual stress data of the historical detection period, and perform Markov matrix conversion on the obtained stress rain flow counting results to obtain a target Markov matrix related to stress in the historical detection period;

[0026] Performing mean stress correction on the target Markov matrix according to the stress rainflow counting result to obtain effective stress values ​​corresponding to all matrix elements in the target Markov matrix;

[0027] For each matrix element in the target Markov matrix, the stress cycle life number is calculated according to the effective stress value corresponding to the matrix element to obtain the target stress cycle life number matching the matrix element;

[0028] Cumulative damage calculation is performed based on the actual stress cycle times and target stress cycle life times corresponding to all matrix elements in the target Markov matrix to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine set within the historical detection period.

[0029] In an optional embodiment, the target Markov matrix Line The effective stress value corresponding to the matrix element of the column is calculated using the following formula:

[0030] ;

[0031] The target Markov matrix Line The target stress cycle life corresponding to the matrix element of the column is calculated using the following formula:

[0032] ;

[0033] in, It is used to represent the first Line The effective stress value corresponding to the matrix element of the column is It is used to represent the first The row matrix elements correspond to the stress amplitudes in the stress rainflow counting results, It is used to represent the first The stress mean of the column matrix elements in the corresponding stress rainflow counting results, Used to indicate reference stress values, It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is and are all constants.

[0034] In an optional implementation manner, the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine set in a single historical detection cycle is calculated using the formula:

[0035] ;

[0036] in, It is used to represent the actual fatigue damage value of the fatigue damage concentration area of ​​the blade of the sample wind turbine in a single historical detection cycle. The matrix order used to represent the target Markov matrix, It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is It is used to represent the first Line The actual number of stress cycles corresponding to the matrix elements in the column.

[0037] In an optional embodiment, the method further comprises:

[0038] Predicting the current blade fatigue life of the wind turbine to be tested based on the estimated fatigue damage value of the wind turbine to be tested in the current detection cycle and the estimated fatigue damage value of the wind turbine to be tested in all historical detection cycles before the current detection cycle;

[0039] Among them, a single wind turbine to be tested is The blade fatigue life of a test cycle is calculated using the following formula:

[0040] ;

[0041] in, It is used to indicate that the wind turbine to be tested is in The fatigue life of the blades in the test cycle, It is used to indicate the blade design life of the wind turbine to be tested. It is used to indicate that the wind turbine to be tested is in The accumulated full-send duration of the detection cycle, It is used to indicate the total full-power duration of the wind turbine to be tested in its entire life cycle. It is used to indicate that the wind turbine to be tested is in Fatigue damage estimation for each test cycle, Used to indicate the blade material reinforcement factor of the wind turbine to be tested.

[0042] In a second aspect, the present application provides a computer device, including a processor and a memory, wherein the memory stores a computer program that can be executed by the processor, and the processor can execute the computer program to implement the wind turbine blade fatigue damage prediction method described in any one of the aforementioned embodiments.

[0043] In a third aspect, the present application provides a readable storage medium having a computer program stored thereon, and when the computer program is executed by a computer device, the method for predicting fatigue damage of wind turbine blades as described in any one of the aforementioned embodiments is implemented.

[0044] In this case, the beneficial effects of the embodiments of the present application may include the following:

[0045] The present application targets each wind turbine to be tested in a target wind farm, obtains the actual unit operation data of the wind turbine to be tested in a current detection cycle, and identifies the actual operating condition type of the wind turbine to be tested in the current detection cycle based on the actual unit operation data of the wind turbine to be tested, and then screens out a target damage prediction model corresponding to the actual operating condition type of the wind turbine to be tested from a variety of blade fatigue damage prediction models that match the target wind farm, and calls the screened target damage prediction model to perform blade fatigue damage prediction based on the actual unit operation data of the wind turbine to be tested, so as to obtain an estimated fatigue damage value of the blade fatigue damage concentration area of ​​the wind turbine to be tested in the current detection cycle, thereby predicting blade fatigue damage based on the wind turbine operation data using a blade fatigue damage prediction model adapted to the wind farm, without the need to incur a large amount of cost to separately install expensive sensor devices for each wind turbine in the wind farm, so as to achieve a high-precision real-time prediction effect of blade fatigue damage for each wind turbine in the same wind farm on the basis of effectively reducing the cost of monitoring the health status of wind turbine blades.

[0046] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A schematic diagram of the composition of a computer device provided in an embodiment of the present application;

[0049] Figure 2 One of the flow charts of the method for predicting fatigue damage of wind turbine blades provided in the embodiment of the present application;

[0050] Figure 3 A second flow chart of a method for predicting fatigue damage of wind turbine blades provided in an embodiment of the present application;

[0051] Figure 4 for Figure 3 A schematic flow chart of the sub-steps included in step S250;

[0052] Figure 5 for Figure 3 A flowchart of the sub-steps included in step S270;

[0053] Figure 6 The third flowchart of the method for predicting fatigue damage of wind turbine blades provided in the embodiment of the present application.

[0054] Icons: 10 - computer device; 11 - memory; 12 - processor; 13 - communication unit. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0056] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0057] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0058] In the description of the present application, it should be understood that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, or are the orientations or positional relationships in which the product of the application is conventionally placed when in use, or are the orientations or positional relationships conventionally understood by those skilled in the art. They are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0059] Furthermore, the terms “first”, “second”, “third”, etc. are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0060] In the description of this application, it should also be noted that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0061] In addition, in the description of the present application, it is understood that the relational terms such as the term "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the term "include", "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements that are not clearly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of more restrictions, the elements defined by the sentence "comprise one..." do not exclude the existence of other identical elements in the process, method, article or equipment including the elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood by specific circumstances.

[0062] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0063] Please refer to Figure 1 , Figure 11 is a schematic diagram of the composition of the computer device 10 provided in the embodiment of the present application. In the embodiment of the present application, the computer device 10 can communicate and connect with all wind turbines in a certain target wind farm to obtain the unit operation data of each wind turbine in different detection cycles, and identify the specific unit operating condition type (for example, unit startup condition, unit shutdown condition, unit standby condition and normal power generation condition, etc.) of each wind turbine in the corresponding detection cycle according to the unit operation data of each wind turbine obtained, and then call the blade fatigue damage prediction model adapted to the target wind farm and corresponding to the specific unit operating condition type, and perform blade fatigue damage prediction based on the unit operation data of the corresponding wind turbine, so as to directly achieve high-precision blade fatigue damage real-time prediction effect for each wind turbine in the same wind farm, without spending a lot of cost to install expensive sensor devices for each wind turbine in the wind farm, so as to reduce the cost of monitoring the health status of wind turbine blades. Among them, the cycle duration of the detection cycle can be 10 minutes, 15 minutes, or 20 minutes, and the specific cycle duration can be flexibly configured according to needs; the unit operation data can be SCADA data collected by the SCADA (Supervisory Control And Data Acquisition) system of the corresponding wind turbine unit, and the unit operation data includes the average unit speed, average wind speed, average blade angle, and average power generation data of the corresponding wind turbine unit within a certain detection cycle; the computer device 10 can be, but is not limited to, a tablet computer, a laptop computer, a personal computer, a server, etc.

[0064] In the embodiment of the present application, the computer device 10 may include a memory 11, a processor 12, and a communication unit 13, wherein the memory 11, the processor 12, and the communication unit 13 are electrically connected to each other directly or indirectly to achieve data transmission or interaction. For example, the memory 11, the processor 12, and the communication unit 13 may be electrically connected to each other via one or more communication buses or signal lines.

[0065] In this embodiment, the memory 11 may be, but not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc. The memory 11 is used to store a computer program, and the processor 12 may execute the computer program accordingly after receiving an execution instruction.

[0066] In addition, the memory 11 can also be used to store multiple blade fatigue damage prediction models matching the target wind farm, each blade fatigue damage prediction model corresponds to a unit operating condition type, and each blade fatigue damage prediction model is used to mine the potential mapping relationship between the unit operation data and the blade fatigue damage amount of any wind turbine in the target wind farm under the corresponding unit operating condition type, so as to achieve a high-precision real-time prediction effect of blade fatigue damage; the memory 11 can also be used to store multiple fatigue damage databases matching the target wind farm, wherein each fatigue damage database corresponds to a unit operating condition type, and each fatigue damage database records the data association relationship between the unit operation data and the blade fatigue damage value of at least one wind turbine in at least one detection cycle under the corresponding unit operating condition type, so as to obtain a blade fatigue damage prediction model matching the target wind farm and corresponding to the unit operating condition type by calling the corresponding fatigue damage database for deep neural network model training.

[0067] In this embodiment, the processor 12 may be an integrated circuit chip with signal processing capability. The processor 12 may be a general-purpose processor, including a central processing unit (CPU), a graphics processing unit (GPU), a network processor (NP), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or at least one of other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc., which may implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present application.

[0068] In this embodiment, the communication unit 13 is used to establish a communication connection between the computer device 10 and other electronic devices through a network, and to send and receive data through the network, wherein the network includes a wired communication network and a wireless communication network. For example, the computer device 10 can obtain the unit operation data of a wind turbine in different detection cycles from a wind turbine through the communication unit 13.

[0069] In this embodiment, the computer device 10 may pre-store a specific computer program related to the wind turbine blade fatigue damage prediction function in the memory 11, and by driving the processor 12 to execute the specific computer program accordingly, call the blade fatigue damage prediction model adapted to the wind farm to perform blade fatigue damage prediction based on the wind turbine operating data, without spending a lot of cost to install expensive sensor devices for each wind turbine in the wind farm. Therefore, on the basis of effectively reducing the cost of monitoring the health status of wind turbine blades, high-precision real-time prediction of blade fatigue damage can be achieved for each wind turbine in the same wind farm.

[0070] Understandably, Figure 1 The block diagram shown is only a schematic diagram of a composition of the computer device 10. The computer device 10 may also include Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown. Figure 1 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0071] In the present application, in order to ensure that the computer device 10 can achieve high-precision real-time prediction of blade fatigue damage for each wind turbine in the same wind farm on the basis of effectively reducing the cost of monitoring the health of wind turbine blades, the present embodiment of the application achieves the above-mentioned purpose by providing a method for predicting fatigue damage of wind turbine blades. The method for predicting fatigue damage of wind turbine blades provided by the present application is described in detail below.

[0072] Please refer to Figure 2 , Figure 2 This is one of the flow charts of the wind turbine blade fatigue damage prediction method provided in the embodiment of the present application. In the embodiment of the present application, Figure 2 The wind turbine blade fatigue damage prediction method shown may include steps S210 to S240.

[0073] Step S210: for each wind turbine to be tested in the target wind farm, actual unit operation data of the wind turbine to be tested in the current detection cycle is obtained.

[0074] Step S220: identifying the actual operating condition type of the wind turbine to be tested in the current detection cycle according to the actual unit operation data of the wind turbine to be tested.

[0075] In this embodiment, after obtaining the unit operation data of any wind turbine to be tested in the target wind farm within a certain detection period, the computer device 10 can numerically match the specific numerical distribution conditions of various unit parameters included in the unit operation data with the unit parameter distribution standards corresponding to each of the multiple pre-stored unit operating condition types, so as to effectively determine which unit operating condition type's unit parameter distribution standard the unit operation data actually complies with, and further determine the specific unit operating condition type of the corresponding wind turbine to be tested within the detection period.

[0076] Step S230, selecting a target damage prediction model corresponding to the actual operating condition type of the wind turbine to be tested from a plurality of blade fatigue damage prediction models matching the target wind farm, wherein each blade fatigue damage prediction model corresponds to a turbine operating condition type.

[0077] In this embodiment, the various blade fatigue damage prediction models are all adapted to the wind speed frequency distribution data of the target wind farm, and the wind speed frequency distribution data can be used to describe the occurrence probability values ​​of the various wind speed segments involved in the target wind farm (for example, 21 wind speed segments 3~4m / s, 4~5m / s, 5~6m / s,..., 23~24m / s and 24~25m / s).

[0078] Step S240, calling the target damage prediction model to perform blade fatigue damage prediction based on the actual unit operation data of the wind turbine to be tested, and obtaining a fatigue damage estimate value of the blade fatigue damage concentration area of ​​the wind turbine to be tested in the current detection cycle.

[0079] In this embodiment, the blade fatigue damage concentration area is used to represent the main fatigue accumulation location of the corresponding wind turbine blade, and is usually represented by an area near the blade root of the corresponding wind turbine blade (for example, a local area 1 meter away from the blade root of the corresponding wind turbine blade), which can be determined using expert experience and / or bladed simulation.

[0080] After determining the target damage prediction model corresponding to a certain wind turbine to be tested in the target wind farm in the current detection cycle, the computer device 10 will extract the unit characteristic parameters with the highest correlation with blade fatigue damage from the actual unit operation data of the wind turbine to be tested in the current detection cycle based on expert experience and Spearman correlation analysis, and then substitute the extracted unit characteristic parameters into the target damage prediction model, so as to use the target damage prediction model to predict the blade fatigue damage prediction value matching the unit characteristic parameters, and use the blade fatigue damage prediction value as the fatigue damage estimation value of the corresponding wind turbine to be tested in the current detection cycle, thereby achieving a high-precision real-time prediction effect of blade fatigue damage for each wind turbine to be tested in the target wind farm, without spending a lot of cost to install expensive sensor devices for each wind turbine in the target wind farm, so as to reduce the cost of monitoring the health status of wind turbine blades.

[0081] Therefore, the present application can execute the above steps S210 to S240, and call the blade fatigue damage prediction model adapted to the wind farm to predict blade fatigue damage based on the wind turbine operation data, without spending a lot of cost to install expensive sensor devices for each wind turbine in the wind farm. In this way, on the basis of effectively reducing the cost of monitoring the health status of wind turbine blades, high-precision real-time prediction of blade fatigue damage can be achieved for each wind turbine in the same wind farm.

[0082] Optionally, see Figure 3 , Figure 3 This is the second flow chart of the wind turbine blade fatigue damage prediction method provided in the embodiment of the present application. Figure 2 Compared with the wind turbine blade fatigue damage prediction method shown in the Figure 3 The wind turbine blade fatigue damage prediction method shown may further include steps S250 to S280, so as to train blade fatigue damage prediction models respectively adapted to different unit operating conditions for the target wind farm.

[0083] Step S250, calculating the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine in the target wind farm according to the wind speed frequency distribution data of the target wind farm.

[0084] In this embodiment, since the wind turbine blade design drawings and the load-stress conversion tables of different components of the wind turbine blade are all wind turbine design parameters, the overall acquisition is difficult, and it is impossible to directly obtain the load-stress conversion coefficient of the blade fatigue damage concentration area of ​​any wind turbine in the target wind farm. It is necessary to use the bladed simulation method to fit the load-stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine (i.e., any wind turbine in the target wind farm) in the target wind farm.

[0085] Optionally, see Figure 4 , Figure 4 yes Figure 3 In the embodiment of the present application, the step S250 may include sub-steps S251 to S253 to simulate and solve the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine when it is adapted to the wind speed frequency distribution data in the target wind farm.

[0086] Sub-step S251, based on the wind speed frequency distribution data of the target wind farm, performs wind power generation simulation on a sample wind turbine in a normal power generation condition to obtain load simulation data corresponding to the blade fatigue damage concentration area of ​​the sample wind turbine in multiple simulation detection cycles.

[0087] In this embodiment, the average wind speed simulated by the sample wind turbines in any simulation detection cycle is within any one of the multiple wind speed segments involved in the target wind farm, and each wind speed segment involved in the target wind farm corresponds to at least one simulation detection cycle.

[0088] Sub-step S252, for each simulation detection cycle, rain flow counting processing is performed on the load simulation data of the simulation detection cycle, and the equivalent load value of the simulation detection cycle is calculated according to the obtained load rain flow counting result.

[0089] In this embodiment, for a single simulation detection cycle, the load rain flow counting result of the simulation detection cycle can be calculated using the load rain flow matrix To express:

[0090] ;

[0091] in, Used to indicate the The mean load value of the load cycle, Used to indicate the The load amplitude of the load cycle, Used to indicate the The actual number of cycles for a load cycle.

[0092] For a single simulation test cycle, the equivalent load value of the simulation test cycle can be calculated using the following formula:

[0093] ;

[0094] in, Used to represent the equivalent load value of the simulation detection cycle, Used to indicate the number of The actual number of cycles of a load cycle, Used to indicate the number of The load amplitude of the load cycle, Used to indicate the number of load cycle types in this simulation test period. It is used to indicate the duration of a single simulation detection cycle. It is used to represent the load cycle frequency of the sample wind turbine. Used to represent the blade material reinforcement factor of the sample wind turbine.

[0095] In sub-step S253, according to the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in normal power generation conditions within a single detection cycle, and the equivalent load values ​​corresponding to each of multiple simulation detection cycles, the load-stress conversion coefficient is solved based on the fatigue damage accumulation principle to obtain the load-stress conversion coefficient of the sample wind turbine in the target wind farm.

[0096] In this embodiment, the process of solving the load stress conversion coefficient is expressed by the following equation:

[0097] ;

[0098] in, It is used to represent the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in a normal power generation condition within a single detection cycle, It is used to represent the total number of wind speed segments of the target wind farm, It is used to indicate the target wind farm The probability of occurrence of the wind speed segment is The blade fatigue damage concentration area of ​​the sample wind turbine generator set in the normal power generation condition is in the first The simulated average fatigue damage value in a single test cycle under the action of different wind speed ranges is Used to indicate that the average wind speed is The total number of simulation detection cycles in the wind speed range is Used to indicate that the corresponding average wind speed is in The first wind speed range The simulated fatigue damage value of the simulation test cycle is Used to indicate that the corresponding average wind speed is in The first wind speed range The equivalent load value of a simulation test cycle, It is used to represent the load stress conversion coefficient of the sample wind turbine in the target wind farm. It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to indicate that the fatigue damage concentration area of ​​the blades of the sample wind turbine is The number of stress cycles under action, Used to indicate the duration of a single detection cycle. It is used to represent the load cycle frequency of the sample wind turbine. and are all constants.

[0099] In this process, ,in It is used to indicate the distribution ratio of normal power generation conditions to all unit conditions. It is used to indicate the design life of the blades of the sample wind turbine. Used to indicate the duration of a single detection cycle.

[0100] Therefore, the present application can simulate and solve the load stress conversion coefficient when the blade fatigue damage concentration area of ​​the sample wind turbine is adapted to the wind speed frequency distribution data in the target wind farm by executing the above sub-steps S251 to S253.

[0101] Step S260, obtaining respectively measured unit operation data of a plurality of historical detection cycles of the sample wind turbines in different unit operating conditions in the target wind farm and measured load data related to the blade fatigue damage concentration area.

[0102] In this embodiment, a load sensor can be added to the sample wind turbine to obtain multiple groups of measured data corresponding to the sample wind turbine under different unit operating conditions, wherein each group of measured data corresponds to a unit operating condition type, and each group of measured data corresponds to a historical detection cycle, and each group of measured data includes the measured unit operation data of the sample wind turbine in the corresponding historical detection cycle, and the measured load data of the blade fatigue damage concentration area of ​​the sample wind turbine in the corresponding historical detection cycle.

[0103] Step S270, for each historical detection cycle, blade fatigue damage calculation is performed based on the load stress conversion coefficient and the measured load data of the historical detection cycle to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection cycle.

[0104] Optionally, see Figure 5 , Figure 5 yes Figure 3 Schematic diagram of the flow of sub-steps included in step S270. In the embodiment of the present application, step S270 may include sub-steps S271 to S275 to solve the actual blade fatigue damage value that matches the actual measured load data for any historical detection cycle.

[0105] Sub-step S271 , performing stress conversion processing on the measured load data of the historical detection period according to the load-stress conversion coefficient to obtain the actual stress data of the historical detection period.

[0106] The actual stress data of a single historical detection period is obtained by multiplying the actual load data of the historical detection period by a load-stress conversion coefficient adapted to the target wind farm.

[0107] Sub-step S272, performing rainflow counting processing on the actual stress data of the historical detection period, and performing Markov matrix conversion on the obtained stress rainflow counting results to obtain a target Markov matrix related to stress of the historical detection period.

[0108] Sub-step S273, performing mean stress correction on the target Markov matrix according to the stress rainflow counting result, and obtaining effective stress values ​​corresponding to all matrix elements in the target Markov matrix.

[0109] In this embodiment, the first Line The effective stress value corresponding to the matrix element of the column is calculated using the following formula:

[0110] ;

[0111] in, It is used to represent the first Line The effective stress value corresponding to the matrix element of the column is It is used to represent the first The row matrix elements correspond to the stress amplitudes in the stress rainflow counting results, It is used to represent the first The stress mean of the column matrix elements in the corresponding stress rainflow counting results, Used to indicate reference stress values.

[0112] Sub-step S274, for each matrix element in the target Markov matrix, the stress cycle life number is calculated according to the effective stress value corresponding to the matrix element to obtain the target stress cycle life number matching the matrix element.

[0113] In this embodiment, the first Line The target stress cycle life corresponding to the matrix element of the column is calculated using the following formula:

[0114] ;

[0115] in, It is used to represent the first Line The effective stress value corresponding to the matrix element of the column is It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is and are all constants.

[0116] Sub-step S275, performing cumulative damage calculation based on the actual stress cycle times and the target stress cycle life times corresponding to all matrix elements in the target Markov matrix, and obtaining the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection period.

[0117] In this embodiment, the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in a single historical detection cycle is calculated using the formula:

[0118] ;

[0119] in, It is used to represent the actual fatigue damage value of the fatigue damage concentration area of ​​the blade of the sample wind turbine in a single historical detection cycle. the matrix order used to represent the target Markov matrix (e.g., 128), It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is It is used to represent the first Line The actual number of stress cycles corresponding to the matrix elements in the column.

[0120] Therefore, the present application can solve the actual blade fatigue damage value that matches the actual measured load data for any historical detection cycle by executing the above sub-steps S271 to S275.

[0121] Step S280: For each unit operating condition type, the measured unit operation data and actual fatigue damage values ​​corresponding to multiple historical detection cycles corresponding to the unit operating condition type are stored in a fatigue damage database corresponding to the unit operating condition type, and a deep neural network model is trained based on the fatigue damage database corresponding to the unit operating condition type to obtain a blade fatigue damage prediction model that matches the target wind farm and corresponds to the unit operating condition type.

[0122] In this embodiment, for each unit operating condition type, expert experience and Spearman correlation analysis can be used to extract the unit characteristic parameters with the highest correlation with blade fatigue damage from the measured unit operation data of each historical detection cycle recorded in the fatigue damage database corresponding to the unit operating condition type, and then a deep neural network model is trained based on the unit characteristic parameters and actual fatigue damage values ​​of each historical detection cycle to ensure that the blade fatigue damage prediction model finally trained can effectively mine the potential mapping relationship between the unit operation data and the blade fatigue damage amount of any wind turbine in the target wind farm under the corresponding unit operating condition type.

[0123] In addition, it can be understood that after the computer device 10 executes the above-mentioned step S240, the fatigue damage estimate value and the actual unit operation data of the corresponding wind turbine to be tested in the current detection cycle can be stored in a fatigue damage database adapted to the unit operating condition type, so that the computer device 10 can continuously call the fatigue damage database to optimize the adapted target damage prediction model to improve the blade fatigue damage prediction accuracy of the target damage prediction model.

[0124] Therefore, the present application can train blade fatigue damage prediction models that are respectively adapted to different unit operating conditions for the target wind farm by executing the above steps S250 to S280.

[0125] Optionally, see Figure 6 , Figure 6 This is the third flow chart of the wind turbine blade fatigue damage prediction method provided in the embodiment of the present application. Figure 2or Figure 3 Compared with the wind turbine blade fatigue damage prediction method shown in the Figure 6 The wind turbine blade fatigue damage prediction method shown may further include step S290 to perform a high-precision prediction of the current blade fatigue life of any wind turbine in the target wind farm, so as to achieve a blade life warning effect.

[0126] Step S290, predicting the current blade fatigue life of the wind turbine to be tested based on the fatigue damage estimated value of the wind turbine to be tested in the current detection cycle and the fatigue damage estimated value of the wind turbine to be tested in all historical detection cycles before the current detection cycle.

[0127] In this embodiment, a single wind turbine to be tested is The blade fatigue life of a test cycle is calculated using the following formula:

[0128] ;

[0129] in, It is used to indicate that the wind turbine to be tested is in The fatigue life of the blades in the test cycle, It is used to indicate the blade design life of the wind turbine to be tested. It is used to indicate that the wind turbine to be tested is in The accumulated full-send duration of the detection cycle, It is used to indicate the total full-power duration of the wind turbine to be tested in its entire life cycle. It is used to indicate that the wind turbine to be tested is in Fatigue damage estimation for each test cycle, Used to indicate the blade material reinforcement factor of the wind turbine to be tested.

[0130] The computer device 10 can make a high-precision blade fatigue life prediction for any wind turbine in the target wind farm based on the above blade fatigue life calculation formula, and then confirm whether the wind turbine currently needs a blade life warning by detecting whether the blade fatigue life of the corresponding wind turbine is less than or equal to the preset life threshold. Among them, when the blade fatigue life of a wind turbine is less than or equal to the preset life threshold, the computer device 10 can send a blade life warning message about the wind turbine to the production management terminal of the target wind farm, so as to inform the wind farm management personnel through the blade life warning message that the wind turbine needs to be replaced and repaired in time, so as to avoid serious safety accidents and production capacity losses.

[0131] Therefore, the present application can perform a high-precision prediction of the current blade fatigue life of any wind turbine in the target wind farm by executing the above step S290, thereby achieving a blade life warning effect.

[0132] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to the embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and a module, a program segment or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0133] In addition, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. If the various functions provided by the present application are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a readable storage medium, including several instructions for the computer device 10 to execute all or part of the steps of the method recorded in each embodiment of the present application. The aforementioned readable storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0134] The above are only various implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for predicting fatigue damage of wind turbine blades, characterized in that: The method comprises: For each wind turbine unit to be tested in the target wind farm, actual unit operation data of the wind turbine unit to be tested in the current detection cycle is obtained; According to the actual unit operation data of the wind turbine unit to be tested, identifying the actual operating condition type of the wind turbine unit to be tested in the current detection cycle; Selecting a target damage prediction model corresponding to the actual operating condition type of the wind turbine to be tested from a plurality of blade fatigue damage prediction models matching the target wind farm, wherein each blade fatigue damage prediction model corresponds to a turbine operating condition type alone; Calling the target damage prediction model to perform blade fatigue damage prediction based on actual unit operation data of the wind turbine to be tested, and obtaining a fatigue damage estimate of a blade fatigue damage concentration area of ​​the wind turbine to be tested within a current detection cycle; Predicting the current blade fatigue life of the wind turbine to be tested based on the estimated fatigue damage value of the wind turbine to be tested in the current detection cycle and the estimated fatigue damage value of the wind turbine to be tested in all historical detection cycles before the current detection cycle; Among them, a single wind turbine to be tested is The blade fatigue life of a test cycle is calculated using the following formula: ； in, It is used to indicate that the wind turbine to be tested is in The fatigue life of the blades in the test cycle, It is used to indicate the blade design life of the wind turbine to be tested. It is used to indicate that the wind turbine to be tested is in The accumulated full-send duration of the detection cycle, It is used to indicate the total full-power duration of the wind turbine to be tested in its entire life cycle. It is used to indicate that the wind turbine to be tested is in Fatigue damage estimation for each test cycle, Used to indicate the blade material reinforcement factor of the wind turbine to be tested.

2. The method according to claim 1, characterized in that The method further comprises: Calculating the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine in the target wind farm according to the wind speed frequency distribution data of the target wind farm; Acquire respectively measured unit operation data of the sample wind turbines in a plurality of historical detection periods under different unit operating conditions in the target wind farm and measured load data related to a blade fatigue damage concentration area; For each historical detection cycle, blade fatigue damage calculation is performed according to the load-stress conversion coefficient and the measured load data of the historical detection cycle to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection cycle; For each unit operating condition type, the measured unit operation data and actual fatigue damage values ​​corresponding to multiple historical detection cycles corresponding to the unit operating condition type are stored in a fatigue damage database corresponding to the unit operating condition type, and a deep neural network model is trained based on the fatigue damage database corresponding to the unit operating condition type to obtain a blade fatigue damage prediction model that matches the target wind farm and corresponds to the unit operating condition type.

3. The method according to claim 2, characterized in that The step of calculating the load stress conversion coefficient of the blade fatigue damage concentration area of ​​the sample wind turbine in the target wind farm according to the wind speed frequency distribution data of the target wind farm comprises: According to the wind speed frequency distribution data of the target wind farm, wind power generation simulation is performed on the sample wind turbine in a normal power generation state to obtain load simulation data corresponding to the blade fatigue damage concentration area of ​​the sample wind turbine in multiple simulation detection cycles; For each simulation detection cycle, rain flow counting processing is performed on the load simulation data of the simulation detection cycle, and the equivalent load value of the simulation detection cycle is calculated according to the obtained load rain flow counting result; According to the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in normal power generation conditions within a single detection cycle, and the equivalent load values ​​corresponding to each of the multiple simulation detection cycles, the load-stress conversion coefficient is solved based on the fatigue damage accumulation principle to obtain the load-stress conversion coefficient of the sample wind turbine in the target wind farm.

4. The method according to claim 3, characterized in that The solution process of the load stress conversion coefficient is expressed by the following equation: ； in, It is used to represent the theoretical average fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in a normal power generation condition within a single detection cycle, It is used to represent the total number of wind speed segments of the target wind farm, It is used to indicate the target wind farm The probability of occurrence of the wind speed segment is The blade fatigue damage concentration area of ​​the sample wind turbine generator set in the normal power generation condition is in the first The simulated average fatigue damage value in a single test cycle under the action of different wind speed ranges is Used to indicate that the average wind speed is The total number of simulation detection cycles in the wind speed range is Used to indicate that the corresponding average wind speed is in The first wind speed range The simulated fatigue damage value of the simulation test cycle is Used to indicate that the corresponding average wind speed is in The first wind speed range The equivalent load value of a simulation test cycle, It is used to represent the load stress conversion coefficient of the sample wind turbine in the target wind farm. It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to indicate that the fatigue damage concentration area of ​​the blades of the sample wind turbine is The number of stress cycles under action, Used to indicate the duration of a single detection cycle. It is used to represent the load cycle frequency of the sample wind turbine. and are all constants.

5. The method according to claim 2, characterized in that: For each historical detection cycle, the step of calculating blade fatigue damage according to the load stress conversion coefficient and the measured load data of the historical detection cycle to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in the historical detection cycle includes: Performing stress conversion processing on the measured load data of the historical detection period according to the load-stress conversion coefficient to obtain actual stress data of the historical detection period; Perform rain flow counting processing on the actual stress data of the historical detection period, and perform Markov matrix conversion on the obtained stress rain flow counting results to obtain a target Markov matrix related to stress in the historical detection period; Performing mean stress correction on the target Markov matrix according to the stress rainflow counting result to obtain effective stress values ​​corresponding to all matrix elements in the target Markov matrix; For each matrix element in the target Markov matrix, the stress cycle life number is calculated according to the effective stress value corresponding to the matrix element to obtain the target stress cycle life number matching the matrix element; Cumulative damage calculation is performed based on the actual stress cycle times and target stress cycle life times corresponding to all matrix elements in the target Markov matrix to obtain the actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine set within the historical detection period.

6. The method according to claim 5, characterized in that The target Markov matrix Line The effective stress value corresponding to the matrix element of the column is calculated using the following formula: ； The target Markov matrix Line The target stress cycle life corresponding to the matrix element of the column is calculated using the following formula: ； in, It is used to represent the first Line The effective stress value corresponding to the matrix element of the column is It is used to represent the first The row matrix elements correspond to the stress amplitudes in the stress rainflow counting results, It is used to represent the first The column matrix element corresponds to the stress mean value in the stress rainflow counting result, Used to indicate the reference stress value, It is used to indicate the minimum stress that the fatigue damage concentration area of ​​the blade of the sample wind turbine receives when the fatigue damage is accumulated. It is used to indicate the maximum stress in the fatigue damage concentration area of ​​the blade of the sample wind turbine when the fatigue damage is accumulated. It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is and are all constants.

7. The method according to claim 5, characterized in that The actual fatigue damage value of the blade fatigue damage concentration area of ​​the sample wind turbine in a single historical detection cycle is calculated using the formula: ； in, It is used to represent the actual fatigue damage value of the fatigue damage concentration area of ​​the blade of the sample wind turbine in a single historical detection cycle. The matrix order used to represent the target Markov matrix, It is used to represent the first Line The target stress cycle life corresponding to the matrix elements of the column is It is used to represent the first Line The actual number of stress cycles corresponding to the matrix elements in the column.

8. A computer device, characterized in that: It comprises a processor and a memory, wherein the memory stores a computer program executable by the processor, and the processor can execute the computer program to implement the method for predicting fatigue damage of wind turbine blades as described in any one of claims 1 to 7.

9. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a computer device, the method for predicting fatigue damage of a wind turbine blade as claimed in any one of claims 1 to 7 is implemented.

Images