Load determination method and device for wind turbine generator system, wind turbine generator system

Abstract

This disclosure provides a method and apparatus for determining the load of a wind turbine generator set, and the wind turbine generator set itself. The load determination method includes: estimating the load of a target generator set at a target time based on a preset load estimation model to obtain target load estimation data; and correcting the target load estimation data based on a load correction relationship corresponding to the load estimation model to determine the load data of the target generator set. The load correction relationship characterizes the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data. This disclosure solves the problem that configuring hardware sensors to measure the load of a generator set leads to a large workload, high cost, and difficult maintenance. By obtaining the load data of the generator set through a model, the workload of sensor installation and maintenance is reduced, thus lowering costs.

Description

Technical Field

[0001] This disclosure relates to the field of wind power generation, and more specifically, to a method and apparatus for determining the load of a wind turbine generator set, and a wind turbine generator set. Background Technology

[0002] Wind turbine generators are typically designed for a service life of 20 years. However, during actual operation, wind resources, generator operating status, generator model, and control strategies are likely to differ from the initial design. Therefore, it is necessary to obtain actual damage data to ensure generator safety. Furthermore, real-time load data is required for precise control of generator operation.

[0003] For example, hardware load sensors can be installed on different components or locations of the unit to obtain the actual load conditions of the unit and thus deduce the actual damage to the unit. However, if such hardware load sensors are configured for each unit, such as configuring sensors for load measurement on all units in the wind farm, it will result in a large workload for installation, high cost, and difficult maintenance. Summary of the Invention

[0004] Given that configuring hardware sensors to measure loads for wind turbine generator sets would result in a large workload, high cost, and difficult maintenance, this disclosure proposes a method and apparatus for determining the load of wind turbine generator sets, as well as a wind turbine generator set, to solve or at least alleviate the above problems.

[0005] The first aspect of this disclosure provides a method for determining the load of a wind turbine generator set. The method includes: estimating the load of a target generator set at a target time based on a preset load estimation model to obtain target load estimation data; and correcting the target load estimation data based on a load correction relationship corresponding to the load estimation model to determine the load data of the target generator set, wherein the load correction relationship characterizes the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

[0006] Optionally, the load correction relationship is determined by: acquiring load measurement data actually measured for the benchmark unit at the target time; estimating the load of the benchmark unit at the target time based on the load estimation model to obtain benchmark load estimation data; and determining the load correction relationship based on the load measurement data and the benchmark load estimation data.

[0007] Optionally, the step of determining the load correction relationship based on the load measurement data and the benchmark load estimation data includes: determining the benchmark load distribution range based on the benchmark load estimation data and a preset confidence level, wherein the confidence level characterizes the error range of the load estimation model in estimating the load; selecting multiple benchmark load distribution data from the benchmark load distribution range according to a preset selection method based on the benchmark load distribution range; and determining multiple load correction relationships based on each benchmark load distribution data and the load measurement data.

[0008] Optionally, the step of correcting the target load estimation data based on the load correction relationship corresponding to the load estimation model to determine the load data of the target unit includes: determining the target load distribution range based on the target load estimation data and the confidence level; selecting multiple target load distribution data from the target load distribution range according to the selection method, wherein the multiple target load distribution data corresponds one-to-one with the multiple load correction relationships; determining the corrected load data corresponding to each target load distribution data based on each target load distribution data and the corresponding load correction relationship; and determining the load data of the target unit by statistically analyzing each corrected load data.

[0009] Optionally, the step of determining the load data of the target unit by statistically analyzing each modified load data includes: determining the modified load data that is at a preset quantile among each modified load data as the load data of the target unit, wherein the selection method includes selecting data corresponding to each preset probability density level according to multiple preset probability density levels.

[0010] Optionally, the load correction relationship is determined by: using multiple fitting methods to fit the load measurement data at multiple times and the benchmark load estimation data at the corresponding times, obtaining candidate correction relationships corresponding to each fitting method; and using the candidate correction relationship whose goodness of fit meets the preset conditions as the load correction relationship.

[0011] Optionally, the input to the load estimation model includes the operating data of the wind turbine generator set whose load is to be estimated at the time of load estimation. The operating data includes wind resource data and generator set operating data. The target generator set includes one or more wind turbine generator sets located in the same wind farm as the benchmark generator set.

[0012] A second aspect of this disclosure provides a load determination device for a wind turbine generator set. The load determination device includes: an estimation unit configured to estimate the load of a target generator set at a target time based on a preset load estimation model to obtain target load estimation data; and a correction unit configured to correct the target load estimation data based on a load correction relationship corresponding to the load estimation model to determine the load data of the target generator set, wherein the load correction relationship characterizes the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

[0013] A third aspect of this disclosure provides an electronic device comprising: a processor; and a memory for storing processor-executable instructions, wherein, when executed by the processor, the processor causes the processor to perform a load determination method for a wind turbine generator according to this disclosure.

[0014] A fourth aspect of this disclosure provides a wind turbine generator set, the wind turbine generator set including electronic equipment according to this disclosure, or the control system of the wind turbine generator set being connected to electronic equipment according to this disclosure.

[0015] A fifth aspect of this disclosure provides a computer-readable storage medium that, when instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform a load determination method for a wind turbine generator according to this disclosure.

[0016] A sixth aspect of this disclosure provides a computer program product including computer-executable instructions that, when executed by at least one processor, implement the load determination method for a wind turbine generator set according to this disclosure.

[0017] According to the load determination scheme for wind turbine generators disclosed herein, the load of the target generator at a target time can be estimated based on a preset load estimation model, yielding target load estimation data. This target load estimation data can then be corrected based on the model's load correction relationship to obtain the final load data. The load correction relationship characterizes the difference between the load estimation data from the load estimation model and the corresponding actual load data. In this way, the load data of the generator can be obtained through a model without requiring separate hardware sensors for each generator, thus reducing the workload of sensor installation and maintenance and lowering costs. Furthermore, this method allows for the correction of the load estimation data output by the model based on the load correction relationship, enabling the acquisition of relatively accurate load data even without using hardware sensors to measure the load. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart illustrating a method for determining the load of a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0019] Figure 2 This is a flowchart illustrating the steps of determining load correction relationships in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0020] Figure 3 This is a flowchart illustrating the steps of determining load correction relationships based on benchmark load distribution data and load measurement data in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0021] Figure 4 This is a schematic diagram illustrating the model error of the load estimation model in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0022] Figure 5 This is a schematic diagram illustrating the frequency of model error occurrence in a load estimation model of a wind turbine generator load determination method according to an exemplary embodiment of the present disclosure.

[0023] Figure 6 This is a schematic diagram illustrating the obtained load distribution range in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0024] Figure 7 This is a flowchart illustrating the steps of determining load data of a target turbine in a load determination method for a wind turbine according to an exemplary embodiment of the present disclosure.

[0025] Figure 8 This is a schematic diagram illustrating corrected load data under different probability densities in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0026] Figure 9 This is a flowchart illustrating the step of determining the load transfer function in a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0027] Figure 10 A schematic block diagram illustrating an example implementation of a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0028] Figure 11 This is a schematic block diagram illustrating a load determination device for a wind turbine generator set according to an exemplary embodiment of the present disclosure. Detailed Implementation

[0029] The following detailed embodiments are provided to aid the reader in gaining a comprehensive understanding of the methods, apparatus, and / or systems described herein. However, various changes, modifications, and equivalents of the methods, apparatus, and / or systems described herein will become apparent upon understanding this disclosure. For example, the order of operations described herein is merely illustrative and is not limited to those orders set forth herein, but may be changed as will become clear upon understanding this disclosure, except for operations that must occur in a specific order. Furthermore, for clarity and conciseness, descriptions of features known in the art may be omitted.

[0030] The features described herein may be implemented in different forms and should not be construed as limited to the examples described herein. Rather, the examples described herein are provided only to illustrate some of the many feasible ways of implementing the methods, apparatus, and / or systems described herein, which will become clear upon understanding the disclosure of this application.

[0031] As used herein, the term “and / or” includes any one of the associated listed items and any combination of any two or more.

[0032] Although terms such as “first,” “second,” and “third” may be used herein to describe various components, assemblies, regions, layers, or parts, these components, assemblies, regions, layers, or parts should not be limited by these terms. Rather, these terms are used only to distinguish one component, assembly, region, layer, or part from another. Thus, without departing from the teaching of the examples described herein, the first component, first assembly, first region, first layer, or first part referred to as the first component, first assembly, first region, first layer, or first part may also be referred to as the second component, second assembly, second region, second layer, or second part.

[0033] In the specification, when an element (such as a layer, region, or substrate) is described as being "on" another element, "connected to," or "bonded to" another element, the element may be directly "on" another element, directly "connected to," or "bonded to" the other element, or one or more other elements may be present in between. Conversely, when an element is described as being "directly on" another element, "directly connected to," or "directly bonded to" another element, no other elements may be present in between.

[0034] The terminology used herein is for the purpose of describing various examples only and is not intended to limit disclosure. Unless the context clearly indicates otherwise, the singular form is intended to include the plural form as well. The terms “comprising,” “including,” and “having” indicate the presence of the described features, quantities, operations, components, elements, and / or combinations thereof, but do not preclude the presence or addition of one or more other features, quantities, operations, components, elements, and / or combinations thereof.

[0035] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains upon understanding this disclosure. Unless expressly defined herein, terms (such as those defined in a general dictionary) shall be interpreted as having a meaning consistent with their meaning in the context of the relevant field and in this disclosure, and shall not be interpreted in an idealized or overly formalistic manner.

[0036] Furthermore, in the description of the examples, detailed descriptions of well-known related structures or functions will be omitted when it is believed that such detailed descriptions would lead to a vague interpretation of this disclosure.

[0037] As mentioned earlier, configuring hardware sensors to measure loads for multiple units would result in a large workload, high cost, and difficult maintenance.

[0038] In some cases, virtual sensors can be used to output the load of wind turbine generators at various points in the wind farm. However, in this method, there will be differences between the load predicted by the virtual sensor and the actual load measured by the generator. The consistency of simulation cannot be avoided. Therefore, the load predicted by the virtual sensor cannot fully reflect the actual load of the generator.

[0039] In view of these problems, this disclosure provides a method for determining the load of a wind turbine generator set, a device for determining the load of a wind turbine generator set, a wind turbine generator set, a computer-readable storage medium, and a computer program product, which can solve or at least alleviate the above-mentioned problems.

[0040] According to a first aspect of an exemplary embodiment of the present disclosure, a method for determining the load of a wind turbine generator is provided. This load determination method can be performed by an electronic device with computing capabilities. The electronic device may be, for example, a terminal device or a server, wherein the terminal device may be such as a tablet computer, a laptop computer, a digital assistant, etc.; the server may be a standalone server, a server cluster, a cloud computing platform, or a virtualization center.

[0041] Here, electronic devices can be installed at wind turbine generators or wind farms, and can be communicatively connected to measuring devices or data centers at wind turbine generators or wind farms, thereby obtaining the data required to perform the above methods, such as load measurement data.

[0042] According to embodiments of this disclosure, such as Figure 1 As shown, the load determination method may include the following steps:

[0043] In step S110, the load of the target unit at the target time can be estimated based on a preset load estimation model to obtain target load estimation data.

[0044] The target turbine can be any wind turbine in the wind farm, and the target load estimation data for the target turbine at any target time can be output through the load estimation model. Here, the target turbine can include one or more wind turbines. The target time can include multiple time points, such as multiple time points within the target time period.

[0045] Load estimation models can be used, for example, to estimate loads at target locations on wind turbine generators. As an example, target locations may include, but are not limited to, the base of the wind turbine tower, the top of the tower, and the blade roots. Load estimation models can also be considered as software sensors or virtual sensors, which can be determined through calculation or simulation.

[0046] As an example, a load estimation model can be obtained through simulation calculations or testing. This model can be, for example, a machine learning model trained using training or a statistical model constructed using statistical methods. The input to the load estimation model can include, for example, the operating data of the wind turbine generator to be estimated at the time of the load estimate. This operating data can include wind resource data and generator operating data. For instance, the operating data of the target generator at the target time can be input into the load estimation model to obtain the target load estimation data.

[0047] Specifically, taking a wind farm as an example, within the same wind farm, the load exhibited at the same target location of a wind turbine generator at different points can be different. This load is related to two main factors: wind resources and the operating status of the turbine itself. Wind resource data may include, but is not limited to, wind speed and direction; turbine operating data may include, but is not limited to, wind turbine generator information such as speed, torque, pitch angle, and azimuth angle. Here, operating data can be, for example, SCADA signals, which can contain the aforementioned wind resource information and wind turbine generator information. Therefore, operating data can reflect the current environment and operating status of the wind turbine generator.

[0048] Load estimation models can be constructed by analyzing the relationship between operational data such as SCADA signals and unit loads. For example, the relationship between operational data and unit loads can be represented by the following equation (1):

[0049] M = G(v, w, r, p, b, ...) (1)

[0050] Where M represents the load, G() represents the transfer function between the operating data and the load, v represents the nacelle wind speed, w represents the wind direction angle, r represents the rotational speed, p represents the power, and b represents the pitch angle. This transfer function can be obtained by testing or simulating calculations in a pre-set load database, followed by data analysis or machine learning. The above equation (1) is only an example; the operating data used may vary depending on the target location.

[0051] In the same wind farm, different turbines are located in different flow fields and receive different incoming winds, resulting in different turbine states. This means that the operating data of different turbines may be different at the same time. By inputting the operating data of the turbines at various points in the entire wind farm into the load estimation model, the load level of the turbine at that moment can be obtained.

[0052] In step S120, the target load estimation data can be corrected based on the load correction relationship corresponding to the load estimation model to determine the load data of the target unit.

[0053] Here, the load correction relation can characterize the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

[0054] Specifically, considering that the load estimation data output by the load estimation model may differ from the actual load, a load correction relationship can be introduced to correct the load estimation data output by the model, making the corrected data closer to the actual load situation.

[0055] In one example, the load correction relationship can be obtained based on historical data, such as by comparing historical load estimates and historical actual load data over a historical period. For instance, an expression for the load correction relationship can be fitted. In actual estimation, the target load estimate data can be corrected using the preset load correction relationship.

[0056] In another example, the load correction relationship can be determined in real time by selecting a benchmark unit, thereby dynamically updating the relationship and making the correction of the load estimation data more accurate.

[0057] Specifically, such as Figure 2 As shown, the load correction relationship can be determined in the following way:

[0058] In step S210, load measurement data actually measured for the benchmark unit at the target time can be obtained.

[0059] In this step, load measurement data of the benchmark unit at the target time can be acquired in real time. The load measurement data can be obtained, for example, through a hardware load sensor, which can be, but is not limited to, a resistance strain gauge sensor or a fiber optic sensor such as a fiber optic grating. The hardware load sensor can emit a sensing signal, which can be sent, for example, to the unit's control system or monitoring platform. This sensing signal can be converted into measured load data. As an example, the sensing signal of the hardware load sensor can be an electrical signal such as a voltage signal; however, it is not limited to this. Depending on the type of sensor and the sensing method, the sensing signal can also be other forms of signal such as an optical signal. As an example, other forms of signal can first be converted into electrical signals, and then the measured load data can be obtained based on the electrical signals.

[0060] Benchmark units can include any one or more wind turbines, for example, they can be located in the same wind farm as the target unit. Target units can include one or more wind turbines located in the same wind farm as the benchmark unit, which can more accurately determine the load correction relationship; however, it is not limited to this. The load correction relationship is intended to reflect the error of the load estimation model in load estimation, so benchmark units from different wind farms can also be used.

[0061] As an example, the type of wind turbine in the target wind farm where the load of the turbine to be measured can be determined; for each type of wind turbine, a benchmark turbine and a target turbine are determined. Here, the type of wind turbine can refer to, for example, the model of the turbine, and the wind farm can include a single model of turbine or multiple models of turbines (i.e., a mixed-model wind farm).

[0062] Specifically, for existing wind farms, benchmark turbines can be determined for each type in the following way: Within a group of wind turbines of the same type, the turbines that meet preset conditions can be used as benchmark turbines. These preset conditions may include: the turbine that generates the most electricity within the group during a predetermined time period. For example, the predetermined time period can be longer than one year to collect data on the impact of seasonal changes on the turbines.

[0063] Specifically, since different turbine models may have different loads, one turbine from the same model group can be selected as a benchmark turbine. For existing wind farms, operating data for each turbine in the same model group should be collected over the same period, ideally for at least one year, to account for the impact of seasonal changes. Then, the total generating time of each turbine during this period can be calculated, and the turbine with the longest generating time can be selected as the benchmark turbine, with hardware load sensors installed. This is because, compared to idling and other operating conditions, the generating condition corresponds to the highest load. In other words, the benchmark turbine is also considered the most vulnerable turbine in the entire farm, and precise strategies can be implemented for it, thus enabling risk control for the entire wind farm.

[0064] For the initial design of wind farms, the adaptability assessment of the entire wind farm in the early stage can be used as a reference to select the unit with the largest load as the benchmark unit.

[0065] In step S220, the load of the benchmark unit at the target time can be estimated based on the load estimation model to obtain benchmark load estimation data.

[0066] For example, the operating data of the benchmark unit at the target time can be input into the load estimation model, and the benchmark load estimation data can be obtained using the load estimation model.

[0067] In step S230, the load correction relationship can be determined based on the load measurement data and the benchmark load estimation data.

[0068] In this step, the load correction relationship can be determined by comparing the load measurement data and the benchmark load estimation data. For example, the difference between the two can be fitted to obtain the load correction relationship. In one example, the load correction relationship can be obtained by fitting the difference between the load measurement data and the benchmark load estimation data at multiple time points.

[0069] In another example, the load estimation data output by the load estimation model may include a load distribution range, or a possible load band. This load distribution range or possible load band represents the range of loads that the model estimates the actual loads might fall into. By setting such a range, the load state can be assessed more accurately compared to estimating a single-point load.

[0070] In this example, the load distribution range output by the model can be considered. Based on multiple benchmark load distribution data in the benchmark load distribution range of the benchmark unit, multiple load correction relationships can be determined. Based on the target load distribution data in the target load distribution range of the target unit that corresponds to each benchmark load distribution data, the corresponding load correction relationships are used to correct the loads respectively, resulting in multiple corrected load data. Based on the multiple corrected load data, the final load data of the target unit can be obtained.

[0071] Specifically, such as Figure 3 As shown, step S230 may include:

[0072] In step S310, the range of the benchmark load distribution can be determined based on the benchmark load estimation data and the preset confidence level.

[0073] Here, the confidence level can characterize the error range of the load estimation model in estimating the load.

[0074] Specifically, considering that during the construction of the load estimation model, errors may occur in the load estimation data output by the load estimation model due to algorithm or testing errors, sample data such as historical data can be used to compare the output of the load estimation model with the corresponding true load value to obtain the model error of the load estimation model, and the frequency of error occurrence can be statistically analyzed.

[0075] As an example, the model error can be represented as shown in equation (2):

[0076] Error = Measure - Output (2)

[0077] Where Error represents the model error, Output represents the load estimation data output by the load estimation model, and Measure represents the corresponding true load value.

[0078] The model error can be calculated at multiple points within a time period, resulting in a schematic diagram of the model error, such as... Figure 4 As shown, by statistically analyzing the various values ​​of the model error within this time period and the frequency corresponding to each value, we can also obtain the following: Figure 5 The diagram shows the distribution of model errors.

[0079] Based on, for example Figure 5 The error distribution shown allows for the selection of the confidence level of the load estimation model to determine its error range. Different confidence levels can be chosen according to actual needs; for example, a confidence level of 90% can be used. Therefore, a 90% confidence interval for the output signal of the load estimation model can be determined, meaning the probability that the true signal value falls within this interval is 90%.

[0080] As an example, based on the unit's operating data, the load estimation model can output different load distribution ranges according to different confidence levels, and each load distribution range can contain loads under different probability densities.

[0081] For example, such as Figure 6 As shown, the upper and lower limits of the load distribution range corresponding to each load estimation data can be determined based on the confidence level, and the actual load data can fall within the load distribution range at the corresponding time. Here, Figures 4 to 6This is intended to illustrate an example of a defined load distribution range; therefore, specific values ​​and units are not shown.

[0082] Here, as an example, the confidence level can be used as a configurable parameter of the load estimation model. After obtaining the load estimation data, the load estimation model can further determine the load distribution range based on the confidence level. In this example, the load distribution range can be directly output from the load estimation model. For example, the load distribution range can output the upper and lower limits of the load distribution range.

[0083] Furthermore, although the examples of determining the confidence level have been described above, the embodiments of this disclosure are not limited thereto. The error range of the load estimation model can also be determined by statistical methods, thereby pre-determining the confidence level.

[0084] In step S310, the load estimation model can be used to determine the range of the benchmark load distribution based on the benchmark load estimation data and the preset confidence level.

[0085] In step S320, multiple benchmark load distribution data can be selected from the benchmark load distribution range according to a preset selection method, based on the benchmark load distribution range.

[0086] Here, the preset selection method can be arbitrary, which means selecting multiple data points within the range of the benchmark load distribution.

[0087] As an example, the selection method can include selecting data corresponding to each data point according to preset data points, such as selecting data corresponding to each of multiple preset probability density levels. The probability density level can be, for example, the boundary value of the cumulative probability density. For instance, the range of the probability density level can be 0.1 to 0.9, with an interval or step size of 0.1. Multiple benchmark load distribution data can be selected within the benchmark load distribution range according to such probability density levels. Taking a benchmark load distribution range of [0,10] as an example, the multiple benchmark load distribution data selected in the above manner could be [1,2,……,9]. However, the selection method and the number of benchmark load distribution data selected are not limited to the above example; other selection methods can also be used, such as selecting data points at different intervals.

[0088] In step S330, multiple load correction relationships can be determined based on the load distribution data and load measurement data of each benchmark.

[0089] In this step, the corresponding load correction relationship can be determined for each benchmark load distribution data. Taking the above probability density level (i.e., data point) range of 0.1 to 0.9 as an example, the following table 1 can be obtained:

[0090] Table 1

[0091]

[0092] As an example, the load correction relationship with each benchmark load distribution data point can be determined based on the benchmark load distribution data M1 to M9 at multiple times and the corresponding load measurement data Ms., such as the load transfer function in Table 1 above, where f1(M1,Ms) to f9(M9,Ms) represent load transfer functions at different probability density levels. The process of determining the load transfer function will be described in detail below.

[0093] Here, given the measured load Ms of the benchmark unit, it is difficult to determine a strong correlation between the estimated load (or predicted load) and the measured load at a certain probability density or data point. Therefore, the above method can more accurately determine multiple load correction relationships between the measured load and the estimated load of the benchmark unit. Furthermore, when there are multiple benchmark units, the load correction relationship corresponding to the data point can also be determined based on the benchmark load distribution data and corresponding load measurement data of each benchmark unit at the same data point.

[0094] In this example, a similar method can be used to determine multiple target load distribution data, and the corrected load data can be obtained by using the above-mentioned load correction relationships.

[0095] Specifically, such as Figure 7 As shown, in step S120, the load data of the target unit can be determined in the following way:

[0096] In step S710, the target load distribution range can be determined based on the target load estimation data and confidence level.

[0097] In this step, similar to determining the benchmark load distribution range, the target load distribution range can be determined based on the target load estimation data and the aforementioned confidence level. Here, the same confidence level can be used when determining the benchmark load distribution range and the target load distribution range.

[0098] In step S720, multiple target load distribution data can be selected from the target load distribution range according to the selection method, based on the target load distribution range.

[0099] In this step, the target load distribution data can be selected in the same way as the benchmark load distribution data.

[0100] For turbines without hardware load sensors, the estimated load time series for each turbine location within the same wind farm can be obtained using the same probability density level. For example, for the i-th turbine, taking the probability density level range of 0.1 to 0.9 as an example, the target load distribution data can be obtained as shown in Table 2 below:

[0101] Table 2

[0102]

[0103] Here, multiple target load distribution data correspond one-to-one with multiple benchmark load distribution data, and multiple target load distribution data also correspond one-to-one with multiple load correction relationships. For example, for the i-th target unit, we can obtain target load distribution data M1 corresponding to load correction relationships f1(M1,Ms) to f9(M9,MS), respectively. i To M9 i .

[0104] In step S730, the corrected load data corresponding to each target load distribution data can be determined based on the target load distribution data and the corresponding load correction relationship.

[0105] For example, the target load distribution data of the i-th unit can be substituted into the corresponding load transfer function obtained from the benchmark unit, such as M1. i *f1(M1,Ms) yields the target load distribution data M1. i Corresponding corrected load data Ms i Similarly, a series of possible measured loads for the i-th unit can be obtained.

[0106] In step S740, the load data of the target unit can be determined by statistically analyzing each corrected load data.

[0107] In this step, the final load data is obtained by statistically analyzing the various corrected load data, which allows for a more accurate determination of the load data.

[0108] As an example, step S740 may include: determining the modified load data that is at a preset quantile among the modified load data as the load data of the target unit.

[0109] Here, the preset quantile can be determined based on the application of the load data. For example, if the load data is used for the safety assessment of the unit, the preset quantile can be the 90th percentile.

[0110] For example, such as Figure 8As shown, the modified load data at three probability density levels over a certain time period are displayed. Multiple modified load data at each time point can be statistically analyzed according to different quantiles to obtain the load data at each time point.

[0111] The following will describe an example of the process for determining the load transfer function.

[0112] As an example, the load correction relationship can be determined as follows: multiple fitting methods are used to fit the load measurement data at multiple times and the benchmark load estimation data at the corresponding times to obtain the candidate correction relationship corresponding to each fitting method; the candidate correction relationship whose goodness of fit meets the preset conditions is taken as the load correction relationship.

[0113] Specifically, such as Figure 9 As shown, in step S901, load measurement data of the benchmark unit over a period of time can be collected, and benchmark load estimation data of the benchmark unit for the corresponding time period can be obtained. Since it is impossible to determine whether the transfer function changes with time or the unit's operating state, a preset method pool can be established in advance. This method pool can include various applicable fitting methods, such as linear regression and least squares methods. In step S902, the load transfer function can be solved using methods from the method pool. In step S903, it is determined whether the currently obtained load transfer function meets preset conditions. If the preset conditions are met, in step S904, the optimal solution can be output as the final load transfer function.

[0114] By using the above method, the best fitting method that meets the goodness-of-fit requirements can be found by traversing and optimizing the methods in the method pool, and the fitting parameters can be obtained, thereby obtaining a more accurate load transfer function.

[0115] Furthermore, as an example, if a fitting method that meets the preset conditions cannot be obtained after traversal, it may indicate that the current amount of data is insufficient to form a transfer function. Data can be collected and the process can wait for the next round of transfer function solving. For example, a round of solving can be performed every predetermined time interval (e.g., 10 minutes) to obtain the load transfer function.

[0116] According to embodiments of this disclosure, the fitting effect obtained using a Gaussian function is better. Specifically, the Gaussian function can be represented by the following equation (3):

[0117]

[0118] Where a, mu, and sigma represent the parameters of the Gaussian function, i.e., the parameters to be solved. The initial values ​​are set as a = 0, and mu = the average value of the estimated load data. sigma = the standard deviation of the estimated load data. The Levenberg-Marquardt curve fitting method can be used, combined with the gradient descent method and the Gauss-Newton method for minimization. The optimal fitting parameters are obtained by minimizing the sum of squares of the errors between the load data of the benchmark unit and the output of the fitting function. Taking the first data point (e.g., the probability density 0.1 mentioned above) as an example, the parameters a1, mu1 and sigma1 can be determined. The load transfer function f1 obtained based on the benchmark load distribution data M1 and the load measurement data Ms can be expressed by the following equation (4):

[0119]

[0120] Among them, M1 i Ms represents the load estimation data. i This indicates that the load data has been corrected.

[0121] Figure 10 An example implementation of a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure is shown. Figure 10 As shown, operational data such as SCADA signals from wind turbine generators at the target time can be input into the load estimation model 1010 (also known as a "load calculator"). The load estimation model 1010 can estimate the load for benchmark and target turbine generators (e.g., other locations in the wind farm), obtaining benchmark load distribution data and target load distribution data. The load processing device 1020 can obtain multiple load correction relationships based on the benchmark load distribution data and the actual load measurement data of the benchmark turbine generators at the target time. Based on these load correction relationships, the target load distribution data can be corrected to obtain the final load, such as the load for the entire wind farm.

[0122] According to the load determination method of the embodiments of this disclosure, a synchronous load measurement scheme for the entire wind farm based on a benchmark turbine can be provided for the wind power field. By selecting a benchmark turbine in the wind farm, installing a hardware load sensor, and acquiring the load measurement value of the benchmark turbine in real time, the load transfer function between the measured load and the predicted load of the benchmark turbine can be obtained as a load correction relationship. Thus, the predicted load of each target turbine in the entire field can be corrected using the load correction relationship, and finally the load of each point in the entire field can be obtained. In this way, synchronous real-time measurement of all points in the entire field can be achieved without installing hardware sensors on all turbines.

[0123] Compared to methods that rely solely on SCADA data or wind resource data as input to estimate the load on wind turbines at various locations across the entire wind farm, the methods of embodiments of this disclosure can determine the load more accurately. For example, load measurement sensors can be installed on only one benchmark turbine to derive the measured loads at other locations at the same time, providing a reference for the overall evaluation or control of wind turbines across the entire farm at a lower cost.

[0124] According to a second aspect of an exemplary embodiment of the present disclosure, a load determination device for a wind turbine generator set is provided, such as... Figure 11 As shown, the load determination device may include an estimation unit 1110 and a correction unit 1120.

[0125] The estimation unit 1110 is configured to estimate the load of the target unit at the target time based on a preset load estimation model, and obtain target load estimation data.

[0126] The correction unit 1120 is configured to correct the target load estimation data based on the load correction relationship corresponding to the load estimation model, and determine the load data of the target unit, wherein the load correction relationship characterizes the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

[0127] As an example, the correction unit 1120 is also configured to determine the load correction relationship by: acquiring load measurement data actually measured for the benchmark unit at the target time; estimating the load of the benchmark unit at the target time based on the load estimation model to obtain benchmark load estimation data; and determining the load correction relationship based on the load measurement data and the benchmark load estimation data.

[0128] As an example, the correction unit 1120 is further configured to: determine the range of benchmark load distribution based on benchmark load estimation data and a preset confidence level, wherein the confidence level characterizes the error range of the load estimation model in estimating the load; select multiple benchmark load distribution data from the range of benchmark load distribution according to a preset selection method based on the range of benchmark load distribution; and determine multiple load correction relationships based on each benchmark load distribution data and load measurement data.

[0129] As an example, the correction unit 1120 is further configured to: determine the target load distribution range based on the target load estimation data and the confidence level; select multiple target load distribution data from the target load distribution range according to the selection method, wherein the multiple target load distribution data correspond one-to-one with multiple load correction relationships; determine the corrected load data corresponding to each target load distribution data based on each target load distribution data and the corresponding load correction relationship; and determine the load data of the target unit by statistically analyzing each corrected load data.

[0130] As an example, the correction unit 1120 is also configured to: determine the correction load data that is at a preset quantile among the correction load data as the load data of the target unit, wherein the selection method includes selecting data corresponding to each probability density level according to a preset plurality of probability density levels.

[0131] As an example, the correction unit 1120 is also configured to: use multiple fitting methods to fit the load measurement data at multiple times and the benchmark load estimation data at the corresponding times to obtain candidate correction relationships corresponding to each fitting method; and use the candidate correction relationships whose goodness of fit meets the preset conditions as load correction relationships.

[0132] As an example, the input to the load estimation model includes the operating data of the wind turbine generator set whose load is to be estimated at the time of the load being estimated. The operating data includes wind resource data and generator set operating data. The target generator set includes one or more wind turbine generator sets located in the same wind farm as the benchmark generator set.

[0133] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0134] According to a third aspect of an exemplary embodiment of the present disclosure, an electronic device is provided, the electronic device comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor-executable instructions, when executed by the processor, cause the processor to perform a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0135] As an example, an electronic device can be a PC, tablet, personal digital assistant, smartphone, or other device capable of executing the aforementioned set of instructions. Here, the electronic device is not necessarily a single device; it can be any collection of devices or circuits capable of executing the aforementioned instructions (or instruction sets) individually or in combination. The electronic device can also be part of an integrated control system or system manager, or can be configured to interconnect locally or remotely (e.g., via wireless transmission) through an interface.

[0136] In electronic devices, a processor may include a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller, or a microprocessor. By way of example and not limitation, a processor may also include analog processors, digital processors, microprocessors, multi-core processors, processor arrays, network processors, etc.

[0137] The processor can execute instructions or code stored in memory, which can also store data. Instructions and data can also be sent and received over a network via a network interface device, which can employ any known transport protocol.

[0138] Memory can be integrated with the processor; for example, RAM or flash memory can be housed within an integrated circuit microprocessor. Alternatively, memory can comprise a separate device, such as an external disk drive, storage array, or other storage device that can be used by any database system. Memory and processor can be operatively coupled, or can communicate with each other, for example, via I / O ports, network connections, etc., enabling the processor to read files stored in the memory.

[0139] In addition, electronic devices may include video displays (such as liquid crystal displays) and user interaction interfaces (such as keyboards, mice, touch input devices, etc.). All components of the electronic device may be interconnected via buses and / or networks.

[0140] According to a fourth aspect of exemplary embodiments of the present disclosure, a wind turbine generator set is provided, which may include electronic equipment described in exemplary embodiments of the present disclosure, or the control system of the wind turbine generator set may be connected to electronic equipment described in exemplary embodiments of the present disclosure.

[0141] According to a fifth aspect of an exemplary embodiment of the present disclosure, a computer-readable storage medium is provided such that, when instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform a load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0142] Specifically, the load determination method for a wind turbine generator set according to embodiments of this disclosure can be programmed into a computer program and stored on a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor causes the processor to perform the pitch control method for the wind turbine generator set according to exemplary embodiments of this disclosure. Examples of computer-readable storage media include: read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disc storage, hard disk drive (HDD), solid-state drive (SSD), card storage (such as multimedia cards, secure digital (SD) cards, or ultra-fast digital (XD) cards), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state drive, and any other device configured to store a computer program and any associated data, data files, and data structures in a non-transitory manner and to provide the computer program and any associated data, data files, and data structures to a processor or computer so that the processor or computer can execute the computer program. In one example, the computer program and any associated data, data files, and data structures are distributed across a networked computer system, such that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed manner through one or more processors or computers.

[0143] According to a sixth aspect of an exemplary embodiment of the present disclosure, a computer program product is provided, including computer-executable instructions that, when executed by at least one processor, implement the load determination method for a wind turbine generator set according to an exemplary embodiment of the present disclosure.

[0144] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0145] Furthermore, it should be noted that although several examples of each step have been described above with reference to the specific accompanying drawings, it should be understood that the embodiments of this disclosure are not limited to the combinations given in the examples. The steps appearing in different drawings can be combined, and the execution order of each step can be changed, which will not be exhaustive here.

[0146] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

[0147] The specific embodiments of this disclosure have been described in detail above. Although some embodiments have been shown and described, those skilled in the art should understand that modifications and variations can be made to these embodiments without departing from the principles and spirit of this disclosure, which are defined by the claims and their equivalents. Such modifications and variations should also be within the protection scope of the claims of this disclosure.

Claims

1. A method for determining the load of a wind turbine generator set, characterized in that, The load determination method includes: Based on the preset load estimation model, the load of the target unit at the target time is estimated to obtain the target load estimation data; Based on the load correction relationship corresponding to the load estimation model, the target load estimation data is corrected to determine the load data of the target unit. The load correction relationship represents the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

2. The load determination method according to claim 1, characterized in that, The load correction relationship is determined in the following manner: Obtain the actual load measurement data for the benchmark unit at the target time; Based on the load estimation model, the load of the benchmark unit at the target time is estimated to obtain benchmark load estimation data; Based on the load measurement data and the benchmark load estimation data, the load correction relationship is determined.

3. The load determination method according to claim 2, characterized in that, The step of determining the load correction relationship based on the load measurement data and the benchmark load estimation data includes: Based on the benchmark load estimation data and the preset confidence level, the range of benchmark load distribution is determined, wherein the confidence level characterizes the error range of the load estimation model in estimating the load; Based on the benchmark load distribution range, multiple benchmark load distribution data are selected from the benchmark load distribution range according to a preset selection method; Based on the load distribution data of each benchmark and the load measurement data, multiple load correction relationships are determined.

4. The load determination method according to claim 3, characterized in that, The step of correcting the target load estimation data based on the load correction relationship corresponding to the load estimation model to determine the load data of the target unit includes: Based on the target load estimation data and the confidence level, the target load distribution range is determined; Based on the target load distribution range, and according to the selection method, multiple target load distribution data are selected from the target load distribution range, wherein the multiple target load distribution data correspond one-to-one with the multiple load correction relationships; Based on the target load distribution data and the corresponding load correction relationship, determine the corrected load data corresponding to each target load distribution data. The load data of the target unit is determined by statistically analyzing the various corrected load data.

5. The load determination method according to claim 4, characterized in that, The step of determining the load data of the target unit by statistically analyzing each corrected load data includes: The corrected load data that falls within a preset quantile among the corrected load data are determined as the load data of the target unit. The selection method includes selecting data corresponding to each probability density level according to a preset plurality of probability density levels.

6. The load determination method according to claim 2, characterized in that, The load correction relationship is determined in the following manner: Multiple fitting methods were used to fit the load measurement data at multiple times and the benchmark load estimation data at the corresponding times, and candidate correction relationships were obtained for each fitting method. Candidate correction relationships that meet the preset conditions for goodness of fit are used as the load correction relationships.

7. The load determination method according to claim 2, characterized in that, The input to the load estimation model includes the operating data of the wind turbine generator set whose load is to be estimated at the time of load estimation. The operating data includes wind resource data and generator set operating data. The target unit includes one or more wind turbine generators located in the same wind farm as the benchmark unit.

8. A load determination device for a wind turbine generator set, characterized in that, The load determination device includes: The estimation unit is configured to estimate the load of the target unit at the target time based on a preset load estimation model, and obtain target load estimation data. The correction unit is configured to correct the target load estimation data based on the load correction relationship corresponding to the load estimation model, thereby determining the load data of the target unit. The load correction relationship represents the difference between the load estimation data estimated by the load estimation model and the corresponding actual load data.

9. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store processor-executable instructions. Wherein, when the processor executes the executable instructions, it causes the processor to perform the load determination method for a wind turbine generator set according to any one of claims 1 to 7.

10. A wind turbine generator set, characterized in that, The wind turbine generator set includes the electronic equipment according to claim 9, or the control system of the wind turbine generator set is connected to the electronic equipment according to claim 9.

11. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the load determination method for a wind turbine generator set according to any one of claims 1 to 7.

12. A computer program product comprising computer-executable instructions, characterized in that, When the computer-executable instructions are executed by at least one processor, they implement the load determination method for a wind turbine generator set according to any one of claims 1 to 7.

Images