Adaptability evaluation method of wind turbine generator, medium and computer equipment

Abstract

The invention provides an adaptability evaluation method of a wind turbine generator, a medium and computer equipment. The adaptability evaluation method comprises the steps of obtaining design parameters of a to-be-developed unit and target wind parameters of a predetermined site; calculating predicted load data of various loads of the to-be-developed unit at the predetermined site based on the design parameters of the to-be-developed unit and the target wind parameters of the predetermined site; based on the design parameters of the to-be-developed unit, calculating an optimal load boundary of various loads of the to-be-developed unit; and comparing the predicted load data with the corresponding optimal load boundary, and evaluating the adaptability of the to-be-developed wind turbine generator in the predetermined site according to a comparison result.

Description

Technical Field

[0001] The present disclosure generally relates to the field of wind power technology, and more specifically, to a method for evaluating the adaptability of a wind turbine generator set. Background Art

[0002] In the relevant technologies used to develop wind turbines (abbreviated as wind turbines) / units, as the competition for the development of wind turbines / units becomes increasingly fierce, in order to ensure that the wind turbines / units to be developed can meet demand expectations, it is necessary to take into account the adaptability of the wind turbines / units at specific sites during the development of the wind turbines / units. The design wind parameters of the wind turbines / units will be involved in the development of the wind turbines / units, and the design wind parameters can represent the external environment in which the wind turbines can operate. In the actual use of the wind turbines / units, the actual wind parameters in the specific area where the wind turbines / units are located are somewhat different from the above-mentioned design wind parameters. This difference in the external environment may result in the wind turbines / units not being applicable in that specific area.

[0003] Therefore, a technical means capable of solving the above problems is needed. Summary of the invention

[0004] An exemplary embodiment of the present disclosure provides a method, medium and computer equipment for evaluating the adaptability of a wind turbine generator set, which can accurately evaluate the adaptability of a wind turbine / unit to be developed at a specific site.

[0005] According to one aspect of an embodiment of the present disclosure, a method for evaluating the adaptability of a wind turbine is provided, the method comprising: obtaining design parameters of a turbine to be developed and target wind parameters of a predetermined site; calculating predicted load data of various loads of the turbine to be developed at the predetermined site based on the design parameters of the turbine to be developed and the target wind parameters of the predetermined site; calculating optimal load boundaries for various loads of the turbine to be developed based on the design parameters of the turbine to be developed; comparing the predicted load data with the corresponding optimal load boundaries, and evaluating the adaptability of the wind turbine to be developed at the predetermined site based on the comparison result.

[0006] Optionally, the step of calculating the predicted load data of various loads of the unit to be developed at the predetermined site may include: obtaining reference wind parameters; calculating first load data of various loads of the unit to be developed under the reference wind parameters and the target wind parameters by inputting the design parameters of the unit to be developed, the target wind parameters of the predetermined site and the reference wind parameters into a load proxy model; calculating second load data of various loads of the unit to be developed at the predetermined site by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into a load trend prediction model; and calculating the predicted load data based on the first load data and the second load data.

[0007] Optionally, the load trend prediction model can be constructed by the following steps: obtaining unit parameters and corresponding modeling parameters of multiple known wind turbines; calculating the load data under various working conditions based on the unit parameters of the multiple known wind turbines, the corresponding modeling parameters and the reference wind parameters; based on the reference wind parameters and the load data under the various working conditions, constructing a load trend prediction model under various working conditions through a predetermined algorithm as the load trend prediction model.

[0008] Optionally, the design parameters of the unit to be developed may include the design wind parameters and target pass rate of the unit to be developed, and the step of calculating the optimal load boundary for various loads of the unit to be developed may include: obtaining historical wind parameters associated with the predetermined site; calculating the fourth load data of each point of the predetermined site under the design wind parameters by inputting the design wind parameters, the target pass rate and the historical wind parameters into the load proxy model; and calculating the optimal load boundary based on the design wind parameters, the target pass rate and the fourth load data.

[0009] Optionally, the step of calculating the optimal load boundary may include: determining multiple groups of candidate load boundaries including all loads based on the baseline values ​​of various loads corresponding to the design wind parameters and the maximum values ​​of various loads in the fourth load data; determining several groups of candidate load boundaries that meet the target pass rate among the multiple groups of candidate load boundaries; calculating cost values ​​of the several groups of candidate load boundaries, and taking a group of candidate load boundaries among the several groups of candidate load boundaries that have cost values ​​that meet preset conditions as the optimal load boundary, wherein the cost values ​​of the several groups of candidate load boundaries are determined based on weight parameters of various loads, candidate load boundaries of various loads and baseline values ​​of various loads.

[0010] Optionally, the step of determining multiple groups of candidate load boundaries that meet the target pass rate among the multiple groups of candidate load boundaries may include: determining N groups of candidate load boundaries that meet the following conditions among the multiple groups of candidate load boundaries as several groups of candidate load boundaries that meet the target pass rate: for each group of candidate load boundaries among the N groups of candidate load boundaries, the number of points among all points that meet the requirement that each load is less than the corresponding benchmark value exceeds a predetermined threshold, wherein the predetermined threshold is the product of the target pass rate and the total number of all points, and N is a natural number greater than 1.

[0011] Optionally, the load proxy model and the load trend prediction model may be constructed respectively by predetermined algorithms based on the reference wind parameter and at least one known unit parameter of the unit.

[0012] Optionally, the load trend model and the load proxy model may respectively include at least one of a multivariate linear regression algorithm and a neural network algorithm.

[0013] Optionally, the step of calculating the first load data of various loads of the unit to be developed under the reference wind parameters and the target wind parameters may include: calculating the third load data of various loads of the unit to be developed under the reference wind parameters by inputting the design parameters of the unit to be developed and the reference wind parameters into the load proxy model; calculating the fourth load data of various loads of the unit to be developed under the target wind parameters by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load proxy model, wherein the first load data includes the third load data and the fourth load data.

[0014] Optionally, the load proxy model can be constructed by the following steps: obtaining modeling parameters of a known wind turbine that is closest to the design parameters of the unit to be developed; calculating the load data under various working conditions based on the modeling parameters and the reference wind parameters; and constructing a load proxy model under various working conditions through a predetermined algorithm based on the reference wind parameters and the load data under various working conditions as the load proxy model.

[0015] According to another aspect of an embodiment of the present disclosure, there is provided an adaptability assessment device for a wind turbine set, the adaptability assessment device comprising: an acquisition module, configured to: acquire design parameters of the set to be developed and target wind parameters of a predetermined site; a load prediction module, configured to: calculate predicted load data of various loads of the set to be developed at the predetermined site based on the design parameters of the set to be developed and the target wind parameters of the predetermined site; a load boundary determination module, configured to: calculate optimal load boundaries for various loads of the set to be developed based on the design parameters of the set to be developed; and an adaptability determination module, configured to: compare the predicted load data with the corresponding optimal load boundary, and evaluate the adaptability of the wind turbine set to be developed at the predetermined site based on the comparison result.

[0016] Optionally, the operation of the load prediction module calculating the predicted load data of various loads of the unit to be developed at the predetermined site may include: obtaining reference wind parameters; calculating first load data of various loads of the unit to be developed under the reference wind parameters and the target wind parameters by inputting the design parameters of the unit to be developed, the target wind parameters of the predetermined site and the reference wind parameters into a load proxy model; calculating second load data of various loads of the unit to be developed at the predetermined site by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into a load trend prediction model; and calculating the predicted load data based on the first load data and the second load data.

[0017] Optionally, the load trend prediction model can be constructed by the following operations: obtaining unit parameters and corresponding modeling parameters of multiple known wind turbines; calculating the load data under various working conditions based on the unit parameters of the multiple known wind turbines, the corresponding modeling parameters and the reference wind parameters; based on the reference wind parameters and the load data under the various working conditions, constructing a load trend prediction model under various working conditions through a predetermined algorithm as the load trend prediction model.

[0018] Optionally, the design parameters of the unit to be developed may include the design wind parameters and target pass rate of the unit to be developed, and the operation of the load boundary determination module calculating the optimal load boundary for various loads of the unit to be developed may include: obtaining historical wind parameters associated with the predetermined site; calculating the fourth load data of each point of the predetermined site under the design wind parameters by inputting the design wind parameters, the target pass rate and the historical wind parameters into the load proxy model; and calculating the optimal load boundary based on the design wind parameters, the target pass rate and the fourth load data.

[0019] Optionally, the operation of the load boundary determination module calculating the optimal load boundary may include: determining multiple groups of candidate load boundaries including all loads based on the baseline values ​​of various loads corresponding to the design wind parameters and the maximum values ​​of various loads in the fourth load data; determining several groups of candidate load boundaries that meet the target pass rate among the multiple groups of candidate load boundaries; calculating cost values ​​of the several groups of candidate load boundaries, and taking a group of candidate load boundaries among the several groups of candidate load boundaries that have cost values ​​that meet preset conditions as the optimal load boundary, wherein the cost values ​​of the several groups of candidate load boundaries are determined based on weight parameters of various loads, candidate load boundaries of various loads and baseline values ​​of various loads.

[0020] Optionally, the operation of determining multiple groups of candidate load boundaries that meet the target pass rate among the multiple groups of candidate load boundaries may include: determining N groups of candidate load boundaries that meet the following conditions among the multiple groups of candidate load boundaries as several groups of candidate load boundaries that meet the target pass rate: for each group of candidate load boundaries among the N groups of candidate load boundaries, the number of points among all points that meet the requirement that each load is less than the corresponding benchmark value exceeds a predetermined threshold, wherein the predetermined threshold is the product of the target pass rate and the total number of all points, and N is a natural number greater than 1.

[0021] Optionally, the load proxy model and the load trend prediction model may be constructed respectively by predetermined algorithms based on the reference wind parameter and at least one known unit parameter of the unit.

[0022] Optionally, the load trend model and the load proxy model may respectively include at least one of a multivariate linear regression algorithm and a neural network algorithm.

[0023] Optionally, the operation of the load prediction module calculating the first load data of various loads of the unit to be developed under the reference wind parameters and the target wind parameters may include: calculating the third load data of various loads of the unit to be developed under the reference wind parameters by inputting the design parameters of the unit to be developed and the reference wind parameters into the load proxy model; calculating the fourth load data of various loads of the unit to be developed under the target wind parameters by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load proxy model, wherein the first load data includes the third load data and the fourth load data.

[0024] Optionally, the load proxy model can be constructed by the following operations: obtaining modeling parameters of a known wind turbine that is closest to the design parameters of the unit to be developed; calculating the load data under various working conditions based on the modeling parameters and the reference wind parameters; and constructing a load proxy model under various working conditions through a predetermined algorithm based on the reference wind parameters and the load data under various working conditions as the load proxy model.

[0025] According to another aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to perform the adaptability evaluation method as described above.

[0026] According to another aspect of an embodiment of the present disclosure, a computer device is provided, comprising: at least one processor; and at least one memory storing computer executable instructions, wherein the computer executable instructions, when executed by the at least one processor, cause the at least one processor to execute the adaptability assessment method as described above.

[0027] According to the adaptability assessment method, medium and computer equipment of the wind turbine set of the exemplary embodiment of the present disclosure, it is proposed to construct a load proxy model in the absence of an accurate wind turbine simulation model and use the load trend prediction model to correct the constructed load proxy model to obtain the final accurate load prediction scheme, thereby realizing a rapid assessment scheme for accurately assessing the adaptability of the wind turbine / set to be developed at a specific site.

[0028] In addition, the rapid evaluation scheme disclosed herein takes into account the impact of the target area requirements on the load boundary. The load boundary considered during the unit design represents the size of the margin that needs to be reserved when the development of the model to be developed is completed. The designed load boundary allows developers to more accurately obtain the cost / price required for the unit to meet the design requirements during the unit development stage. Since different target area ranges may correspond to different load boundaries, the load boundary can be used to compare the cost / price differences of different target areas, which enables developers to adjust the development direction in a timely manner according to the evaluation results during the development stage and improve the competitiveness of the unit.

[0029] Additional aspects and / or advantages of the present general inventive concept will be set forth in part in the following description and in part will be apparent from the description or may be learned through practice of the present general inventive concept. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The above and other objects and features of the exemplary embodiments of the present disclosure will become more apparent through the following description in conjunction with the accompanying drawings which exemplarily illustrate the embodiments, in which:

[0031] Figure 1 A flow chart showing a method for evaluating the adaptability of a wind turbine generator system according to an exemplary embodiment of the present disclosure;

[0032] Figure 2 A flowchart showing a method for constructing a load proxy model according to an exemplary embodiment of the present disclosure;

[0033] Figure 3 A flowchart showing a method for constructing a load trend prediction model according to an exemplary embodiment of the present disclosure;

[0034] Figure 4 is a schematic diagram showing optimal load boundaries for multiple loads according to an exemplary embodiment of the present disclosure;

[0035] Figure 5 is a schematic diagram showing an application interface for evaluating the adaptability of a wind turbine using an adaptability evaluation method of a wind turbine according to an exemplary embodiment of the present disclosure;

[0036] Figure 6 A structural block diagram showing an adaptability assessment device for a wind turbine generator system according to an exemplary embodiment of the present disclosure is shown;

[0037] Figure 7 A structural block diagram of a computer device according to an exemplary embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0038] Reference will now be made in detail to the embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to like parts throughout. The embodiments will be described below with reference to the drawings in order to explain the present disclosure.

[0039] As mentioned in the above description, in order to ensure that the wind turbine / unit to be developed can meet the expected regional demand, it is necessary to have a certain grasp of the adaptability of the wind turbine / unit at each specific site during the development of the wind turbine / unit. The original intention of this disclosure is to solve how to accurately evaluate the adaptability of the wind turbine / unit to be developed at a specific site without an accurate wind turbine simulation model.

[0040] In order to achieve the above intention, it is necessary to determine the adaptability of the wind turbine / unit at the specific site by determining the various loads of the wind turbine / unit at the characteristic site.

[0041] As an example, the adaptability of a wind turbine can be accurately evaluated in the following manner: using wind parameters as input, establishing a wind turbine simulation model, calculating various loads on the wind turbine under the input (for example, loads on wind turbine components), and verifying the safety of various components of the wind turbine based on the calculated loads, thereby obtaining the overall safety of the unit.

[0042] As another example, in order to quickly determine the safety of a specific type of wind turbine in a specific area, there are currently many methods for quickly evaluating loads based on wind parameters in a specific area. In order to obtain a rapid load evaluation model for this type of method, engineers need an accurate simulation model to calculate a large number of load samples as input to train the model. After the model training is completed, the wind parameters are input into the trained model to obtain the wind turbine load in a specific area, and the safety of the unit is determined by comparing the obtained wind turbine load with the corresponding load boundary.

[0043] Since there is no accurate wind turbine simulation model for the wind turbine to be developed (the model is not determined), the load cannot be accurately evaluated, so it is difficult to use the methods in the above two examples to evaluate the adaptability of the wind turbine. Therefore, before proposing the solution of this application, we first consider the following two aspects to propose an adaptability evaluation for the wind turbine / unit to be developed, namely, the comparison method of the wind parameter of a specific site with the design wind parameter, and the rapid evaluation method with reference to similar units.

[0044] Specifically, in terms of the comparison between the wind parameters of a specific site and the design wind parameters, a wind parameter index calculation formula can be designed based on experience, which comprehensively considers wind parameter characteristics such as air density, wind shear, turbulence (such as turbulence intensity), annual average wind speed, and inflow angle. For example, the formula can be expressed as: indicator = f (air density, wind shear, turbulence, annual average wind speed, inflow angle), where indicator represents the wind parameter index, and f represents the functional relationship between the wind parameter index and the wind parameter characteristics in brackets in the formula. By comparing the indicator of a specific site with the indicator of the design wind parameters, the applicability of the wind turbine at the specific site can be judged. However, the calculation results of this wind parameter comparison method are relatively rough and cannot accurately reflect the load characteristics of the wind turbine. In addition, this method is highly dependent on the experience of engineers, so it has poor scalability.

[0045] In terms of rapid assessment with reference to similar units, we can use the rule that models with similar unit characteristics such as impeller diameter and rated speed also have similar load characteristics. We can use fans of models similar to the wind turbine to be developed for rapid load assessment. By comparing the load boundaries of fans of similar models with the differences in loads at a specific site, we can determine the applicability of the wind turbine at the specific site.

[0046] However, this rapid assessment method cannot fully consider the impact of design differences on wind turbine loads, and will lead to large deviations when the load patterns between different models are inconsistent. In addition, this method mainly judges the adaptability of wind turbines by comparing the loads at a specific site with the loads corresponding to the design wind parameters. However, due to positioning differences such as target area requirements, the loads of the design wind parameters for wind turbines of similar models often cannot well reflect the boundaries of the wind turbines to be developed.

[0047] Therefore, the present disclosure proposes a rapid assessment scheme for the adaptability of a wind turbine / unit to be developed at a predetermined site, which can accurately assess the adaptability of the wind turbine / unit without an accurate wind turbine simulation model. Furthermore, the assessment scheme according to the present disclosure can accurately reflect the performance of the wind turbine / unit, take into account the impact of technological progress, and can also be used to further study the differences in unit development costs under different target area requirements.

[0048] Refer to the following Figures 1 to 6 The adaptability assessment method and device of a wind turbine generator system according to the present disclosure are specifically described.

[0049] Figure 1 A flow chart of a method 100 for evaluating the adaptability of a wind turbine according to an exemplary embodiment of the present disclosure is shown.

[0050] As an example, a wind farm may include a plurality of wind turbine generator sets (referred to as turbines for short), and each wind turbine generator set may include one or more wind turbines (referred to as wind turbines for short).

[0051] Reference Figure 1 In step S101, the design parameters of the unit to be developed and the target wind parameters of the predetermined site are obtained.

[0052] As an example, the design parameters of the unit to be developed may include unit-related parameters, design wind parameters and target pass rate of the unit to be developed. As an example, the wind parameters mainly considered in the present disclosure include wind speed, turbulence intensity, wind shear, inflow angle, and air density. For example, unit-related parameters may include impeller diameter, rated speed, rated power, model development time, etc. Here, the target pass rate is an indicator of the adaptability of the model of the unit to be developed to the point locations in the target area (or target site). For example, assuming that the total number of points in the target area is M, and the value of the target pass rate is set to 80%, then the target pass rate of 80% means that the model of the unit to be developed needs to be applicable at 80%×M points.

[0053] In step S102, based on the design parameters of the unit to be developed and the target wind parameters of the predetermined site, the predicted load data of various loads of the unit to be developed at the predetermined site are calculated.

[0054] According to an embodiment of the present disclosure, the step of calculating predicted load data of various loads of the unit to be developed at a predetermined site may include the following steps A1) to A4):

[0055] A1) Obtain reference wind parameters;

[0056] A2) calculating first load data of various loads of the unit to be developed under the reference wind parameter and the target wind parameter by inputting the design parameters of the unit to be developed, the target wind parameter and the reference wind parameter of the predetermined site into the load proxy model;

[0057] A3) calculating second load data of various loads of the unit to be developed at the predetermined site by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load trend prediction model;

[0058] A4) Calculating predicted load data based on the first load data and the second load data.

[0059] According to an embodiment of the present disclosure, the step A2) of calculating the first load data may further include the following steps:

[0060] A21) calculating third load data of various loads of the unit to be developed under the reference wind parameter by inputting the design parameters of the unit to be developed and the reference wind parameter into the load proxy model;

[0061] A22) calculating fourth load data of various loads of the unit to be developed under the target wind parameter by inputting the design parameters of the unit to be developed and the target wind parameter of the predetermined site into the load proxy model;

[0062] A23) The third payload data and the fourth payload data are taken together as the first payload data.

[0063] As an example, the load proxy model and the load trend prediction model may be constructed by a predetermined algorithm based on a reference wind parameter and a unit parameter of at least one known unit, respectively. For example, the predetermined algorithm may be at least one of a multivariate linear regression algorithm and a neural network algorithm, but the present disclosure is not limited thereto.

[0064] Here, the predetermined algorithms of the load proxy model and the load trend prediction model may be the same or different. For example, the load proxy model may be constructed by a multivariate linear regression algorithm based on reference wind parameters and at least one known unit parameter, and the load trend prediction model may be constructed by a neural network algorithm based on reference wind parameters and at least one known unit parameter.

[0065] Refer to the following Figure 2 and Figure 3 The construction methods of load proxy model and load trend prediction model are explained in detail respectively.

[0066] Figure 2 A flow chart of a method 200 for constructing a load proxy model according to an exemplary embodiment of the present disclosure is shown.

[0067] Reference Figure 2 In step S201, the modeling parameters of the known wind turbine generator set closest to the design parameters of the turbine generator set to be developed are obtained.

[0068] Here, the known wind turbine that is closest to the unit design parameters in the design parameters of the unit to be developed is a wind turbine that has been developed and has accurate modeling information (for example, including a determined model, etc.). For example, the closest parameters may include model parameters, etc., but are not limited thereto.

[0069] In step S202, based on the modeling parameters and the reference wind parameters, the load data under each working condition is calculated.

[0070] As an example, the reference wind parameters used here can be multiple wind parameter combinations (for example, at least 50 wind parameter combinations) generated within the range (for example, upper and lower limits) of historical project wind parameters based on various data generation methods (for example, random number sequences, orthogonal sampling, etc.). In addition, the working conditions considered in the present disclosure comply with the IEC standard working condition design, and relevant details are not repeated here.

[0071] In step S203, based on the reference wind parameters and the load data under various working conditions, a load proxy model under various working conditions is constructed by a predetermined algorithm as the load proxy model.

[0072] As an example, a multivariate linear regression algorithm can be used as a predetermined algorithm to train the load proxy model for each working condition, the input of the algorithm is wind speed, turbulence intensity, wind shear, inflow angle, air density, and the output is the equivalent load value (for calculating the equivalent fatigue load) and / or load value (for calculating the limit load) of various loads. As an example, the post-processing algorithm specified in the IEC standard can be used to process the load of each working condition (for example, the output described above) to obtain the final output result, that is, the equivalent fatigue load and / or limit load of various loads.

[0073] For example, the above post-processing algorithm is an algorithm for further post-processing the output data obtained by the multiple linear regression algorithm to obtain the final output result. In other words, the post-processing algorithm can be used together with the multiple linear regression algorithm as an algorithm required for constructing a load proxy model.

[0074] By constructing a load proxy model in the above manner, it is possible to construct a suitable load proxy model for the unit to be developed in the absence of an accurate wind turbine simulation model, so as to be used for subsequent reasonable prediction of the load under various wind parameter conditions.

[0075] Figure 3 A flow chart of a method 300 for constructing a load trend prediction model according to an exemplary embodiment of the present disclosure is shown. According to an embodiment of the present disclosure, the load trend prediction model is used to correct the load predicted by the load proxy model to more accurately predict the load of the aircraft model to be developed.

[0076] Reference Figure 3 In step S301, the unit parameters and corresponding modeling parameters of a plurality of known wind turbine units are obtained.

[0077] Here, since the unit parameters of multiple known wind turbines are taken into consideration, the load data corresponding to multiple models can be obtained, thereby obtaining the trend of random changes in the values ​​of the same load component and more accurately predicting the values ​​of the load components of the models to be developed.

[0078] In step S302, based on a plurality of known wind turbine parameters, corresponding modeling parameters and reference wind parameters, load data under various working conditions are calculated.

[0079] As an example, since the impeller diameter, rated speed, and rated power usually determine the load magnitude of the model, the input variables that need to be considered include at least: impeller diameter, rated speed, rated power, air density, turbulence, and model development time. The values ​​of the above input variables are arranged and combined to obtain the input condition combination. The values ​​of different load components of each working condition under the above input condition combination are calculated through load simulation to obtain training samples.

[0080] For example, the impeller diameter and model development time need to be determined based on historical data and do not need to be changed. The rated speed and rated power have a certain correlation with the impeller diameter, and the range of variation is small, but they can still be arranged and combined with the impeller diameter within a certain numerical range. Air density and turbulence can be determined by referring to the range and step size of the load proxy model.

[0081] In step S303, based on the reference wind parameters and the load data under various working conditions, a load trend prediction model under various working conditions is constructed by a predetermined algorithm as a load trend prediction model.

[0082] As an example, a neural network algorithm can be used as a predetermined algorithm to train a load trend prediction model for each operating condition, and the algorithm input is the impeller diameter, rated speed, rated power, air density, wind speed, turbulence and model development time of different training samples, and the output is the equivalent load value (for calculating the equivalent fatigue load) and / or load value (for calculating the limit load) of various loads. As an example, the post-processing algorithm specified in the IEC standard can be used to process the load of each operating condition (the above-mentioned output) to obtain the final output result, that is, the equivalent fatigue load and / or limit load of various loads.

[0083] For example, the above-mentioned post-processing algorithm is an algorithm for further post-processing the output data obtained by using the neural network algorithm to obtain the final output result. That is to say, the post-processing algorithm can be used together with the neural network algorithm as an algorithm required for constructing a load trend prediction model. The following example illustrates the specific process of realizing load prediction using the constructed load proxy model and load trend prediction model.

[0084] In step a), the input wind parameters (corresponding to the specific site), the impeller diameter, rated speed, rated power, and development time of the model to be developed are determined.

[0085] In step b), the predicted load TPL (Trend Prediction Load) of each component is calculated using the load trend prediction model according to the air density, turbulence in the input wind parameters and the impeller diameter, rated speed, rated power and development time of the model to be developed.

[0086] In step c), a set of another reference wind parameter is generated, including wind parameters that are configured according to the aforementioned reference wind parameter configuration except that the air density and turbulence are the same as the air density and turbulence of the input wind parameter, and the load SPL (Surrogate Prediction Load) under the other reference wind parameter is predicted using the load proxy model. Therefore, the correction coefficients of different load components can be calculated, that is, the correction coefficients of different load components can be equal to the quotient of the TPL of the corresponding load divided by the SPL.

[0087] In step d), for each point of the specific site, the load proxy model is used to calculate the load of the benchmark model under the input wind parameters, and the load value of the model to be developed at the current point can be calculated, that is, the load value of the model to be developed at the current point is equal to the load of the benchmark model multiplied by the corresponding correction coefficient.

[0088] By constructing a load trend prediction model in the above manner, the load proxy model constructed above for the unit to be developed can be further modified, and the impact of different machine models on load prediction can be further considered to make a more accurate prediction of the load under various wind parameter conditions.

[0089] Reference Figure 1 In step S103, based on the design parameters of the unit to be developed, the optimal load boundaries for various loads of the unit to be developed are calculated.

[0090] According to an embodiment of the present disclosure, the step of calculating the optimal load boundary for various loads of the unit to be developed may include:

[0091] B1) Obtain historical wind parameters associated with the intended site;

[0092] B2) calculating the fourth load data of each point of the predetermined site under the design wind parameters by inputting the design wind parameters, the target passing rate and the historical wind parameters into the load proxy model;

[0093] B3) Calculate the optimal load boundary based on the design wind parameter, target pass rate and the fourth load data.

[0094] As an example, the step B3) of calculating the optimal load boundary may further include:

[0095] B31) determining a plurality of groups of candidate load boundaries including all loads based on reference values ​​of various loads corresponding to the design wind parameters and maximum values ​​of various loads in the fourth load data;

[0096] B32) determining a plurality of groups of candidate load boundaries that meet a target pass rate from among the plurality of groups of candidate load boundaries;

[0097] B33) Calculate cost values ​​of several groups of candidate load boundaries, and use a group of candidate load boundaries with cost values ​​that meet preset conditions among the several groups of candidate load boundaries as the optimal load boundary. Here, the cost values ​​of the several groups of candidate load boundaries are determined based on weight parameters of various loads, candidate load boundaries of various loads, and reference values ​​of various loads.

[0098] For example, the cost value satisfying the preset condition may be a minimum cost value.

[0099] As an example, the step (B32) of determining multiple groups of candidate load boundaries that meet the target pass rate among multiple groups of candidate load boundaries may further include: determining N groups of candidate load boundaries that meet the following conditions among multiple groups of candidate load boundaries as several groups of candidate load boundaries that meet the target pass rate: for each group of candidate load boundaries among the N groups of candidate load boundaries, the number of points that meet the requirement that each load is less than the corresponding reference value among all points exceeds a predetermined threshold. Here, the predetermined threshold is the product of the target pass rate and the total number of all points, and N is a natural number greater than 1.

[0100] The following example illustrates the specific process of calculating the optimal load boundary. The load components that need to be considered in this example can be found in Table 1.

[0101] C1) Determine the design input of the aircraft to be developed, such as design wind parameters, target pass rate, etc.;

[0102] C2) according to the preset target area requirements, obtain the wind parameters of existing projects in the target area. For example, if the target area is the south, select the wind parameters of all existing wind farms in the southern provinces;

[0103] C3) using the load proxy model to calculate various load components corresponding to the wind parameters of each point of the site of the existing wind farm in the target area;

[0104] C4) Use the above load proxy model to calculate the various load components under the design wind parameters;

[0105] C5) One approach to the problem of calculating the optimal load boundary is to interpret the problem as solving the following optimization problem.

[0106] miny=p 1 L 1 / L base1+p 2 L 2 / L base2 +…+p n L n / L basen (1)

[0107]

[0108] Among them, L n is the optimal load boundary to be solved. n represents the number of the load component to be considered (i.e., the serial number in Table 1). base1 is the reference load corresponding to the design wind parameter. n Represents the weight of different load components (determined by the ratio of the impact of a 5% change in different load components on the cost. The greater the impact on the overall machine cost, the higher the weight). t,n represents the nth load value at the tth point. passrate represents the target pass rate for the model to be developed. y represents the cost value. The optimal solution of formula (1) can be understood as the optimal load boundary L when the minimum cost value (i.e., y reaches the minimum value) is required to meet the constraints. n The constraint condition means that the number of points among all points that satisfy that all loads are less than the predetermined boundary L must exceed passrate multiplied by m, where m represents the total number of points.

[0109] One of the solutions is as follows:

[0110] D1) defines the upper and lower limits of the load boundary of each load component, the lower limit is the load corresponding to the design wind parameter, and the upper limit is the maximum load within the wind parameter range selected in C2). The value between the upper and lower limits is evenly divided into a parts as the sampling range. Here, since the larger a is, the longer the algorithm execution time is, the value of a can be determined based on the time requirement and accuracy requirement.

[0111] D2) Randomly select a value in the range of each load component to form a set of boundary values, and randomly generate b groups of boundaries.

[0112] D3) Select the boundary in group b that meets the target pass rate condition.

[0113] D4) Calculate the cost value y (for example, the cost value y can be represented by a score value here) of each group of boundaries that meet the target pass rate condition, and select a group of boundaries with the smallest cost value (for example, the lowest score value) from the several groups of boundaries that meet the pass rate condition as the optimal load boundary.

[0114] Table 1

[0115]

[0116]

[0117] By fully considering the impact of design wind parameters, pass rate and other indicators on the load boundary according to the needs of the target area, for example, a higher load boundary means better adaptability, but the unit design cost will be higher; a lower load boundary means that the unit is limited in scope of use and may not have good universality. According to the rapid evaluation scheme of the embodiment of the present disclosure, it is possible to strike a good balance between the adaptability of the wind turbine in the target area and the design cost / scope of use, so that the optimal load boundary is well matched with the actual situation of the unit to be developed, effectively avoiding the greater risks brought about by mismatches.

[0118] Reference Figure 1 In step S104, the predicted load data is compared with the corresponding optimal load boundary, and the adaptability of the wind turbine to be developed at the predetermined site is evaluated according to the comparison result. Specifically, if the predicted load data is within the range of the corresponding optimal load boundary, the adaptability of the wind turbine to be developed at the predetermined site can be evaluated as available.

[0119] The following is referenced Figure 4 and Figure 5 To illustrate the effect of the solution according to the present application.

[0120] Figure 4 is a schematic diagram illustrating optimal load boundaries for multiple loads according to an exemplary embodiment of the present disclosure. Figure 5 is a schematic diagram showing an application interface for evaluating the adaptability of a wind turbine using the adaptability evaluation method of the wind turbine according to an exemplary embodiment of the present disclosure.

[0121] Reference Figure 4 The black solid line represents the optimal load boundary, the black dotted line represents the load baseline (i.e., the load corresponding to the design wind parameter), and the box plot represents the distribution of the change proportion of each load component (the load components can be found in Table 1) relative to the corresponding benchmark load.

[0122] Reference Figure 5 , shows the result of evaluating the model to be developed using the evaluation method according to the present disclosure. It can be seen from the figure that the applicability of the model in the figure is evaluated as "passed" at five points.

[0123] According to the adaptability assessment method and device of a wind turbine set of the exemplary embodiment of the present disclosure, it is proposed to construct a load proxy model in the absence of an accurate wind turbine simulation model and use a load trend prediction model to correct the constructed load proxy model to obtain a final accurate load prediction scheme, thereby realizing a rapid assessment scheme for accurately assessing the adaptability of the wind turbine / set to be developed at a specific site.

[0124] In addition, the rapid evaluation scheme disclosed herein takes into account the impact of the target area requirements on the load boundary. The load boundary considered during the unit design represents the size of the margin that needs to be reserved when the development of the model to be developed is completed. The designed load boundary allows developers to more accurately obtain the cost / price required for the unit to meet the design requirements during the unit development stage. Since different target area ranges may correspond to different load boundaries, the load boundary can be used to compare the cost / price differences of different target areas, which enables developers to adjust the development direction in a timely manner according to the evaluation results during the development stage and improve the competitiveness of the unit.

[0125] Figure 6 The structural block diagram of the wind turbine adaptability assessment device according to the exemplary embodiment of the present disclosure is shown. According to the embodiment of the present disclosure, the wind turbine adaptability assessment device 600 can be used to assess the adaptability of the wind turbine at a predetermined site.

[0126] Reference Figure 6 The adaptability assessment device 600 for a wind turbine generator system includes: an acquisition module 610 , a load prediction module 620 , a load boundary determination module 630 and an adaptability determination module 640 .

[0127] According to an embodiment of the present disclosure, the acquisition module 610 is configured to: acquire the design parameters of the unit to be developed and the target wind parameters of the predetermined site.

[0128] For example, the design parameters of the unit to be developed may include unit-related parameters, design wind parameters and target pass rate of the unit to be developed.

[0129] According to an embodiment of the present disclosure, the load prediction module 620 is configured to calculate predicted load data of various loads of the unit to be developed at the predetermined site based on the design parameters of the unit to be developed and the target wind parameters of the predetermined site.

[0130] As an example, the operation of the load prediction module 620 calculating the predicted load data of various loads of the unit to be developed at the predetermined site may include: (1) obtaining reference wind parameters; (2) calculating first load data of various loads of the unit to be developed under the reference wind parameters and the target wind parameters by inputting the design parameters of the unit to be developed, the target wind parameters of the predetermined site and the reference wind parameters into the load proxy model; (3) calculating second load data of various loads of the unit to be developed at the predetermined site by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load trend prediction model; (4) calculating the predicted load data based on the first load data and the second load data.

[0131] Optionally, the operation of the load prediction module 620 calculating the first load data of various loads of the unit to be developed under reference wind parameters and target wind parameters may include: (1) calculating third load data of various loads of the unit to be developed under the reference wind parameters by inputting the design parameters of the unit to be developed and the reference wind parameters into the load proxy model; (2) calculating fourth load data of various loads of the unit to be developed under the target wind parameters by inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load proxy model; (3) taking the third load data and the fourth load data together as the first load data for calculating the predicted load data.

[0132] Optionally, the load proxy model and the load trend prediction model may be constructed respectively by predetermined algorithms based on reference wind parameters and at least one known unit parameter of the unit.

[0133] Optionally, the load trend model and the load proxy model may respectively include at least one of a multivariate linear regression algorithm and a neural network algorithm.

[0134] As an example, the load proxy model can be constructed by the following operations: obtain the modeling parameters of a known wind turbine that is closest to the design parameters of the unit to be developed; calculate the load data under various working conditions based on the modeling parameters and reference wind parameters; and construct a load proxy model under various working conditions through a predetermined algorithm based on the reference wind parameters and the load data under various working conditions as the load proxy model.

[0135] As an example, a load trend prediction model can be constructed by the following operations: obtain the unit parameters and corresponding modeling parameters of multiple known wind turbines; calculate the load data under various working conditions based on the unit parameters, corresponding modeling parameters and reference wind parameters of multiple known wind turbines; based on the reference wind parameters and the load data under various working conditions, construct a load trend prediction model under various working conditions through a predetermined algorithm as a load trend prediction model.

[0136] According to an embodiment of the present disclosure, the load boundary determination module 630 is configured to calculate the optimal load boundaries for various loads of the unit to be developed based on the design parameters of the unit to be developed.

[0137] As an example, the operation of the load boundary determination module 630 to calculate the optimal load boundary for various loads of the unit to be developed may include: obtaining historical wind parameters associated with the predetermined site; calculating the fourth load data of each point of the predetermined site under the design wind parameters by inputting the design wind parameters, target pass rate and historical wind parameters into the load proxy model; and calculating the optimal load boundary based on the design wind parameters, target pass rate and fourth load data.

[0138] Optionally, the operation of calculating the optimal load boundary by the load boundary determination module 630 may include: determining multiple groups of candidate load boundaries including all loads based on the reference values ​​of various loads corresponding to the design wind parameters and the maximum values ​​of various loads in the fourth load data; determining several groups of candidate load boundaries that meet the target pass rate among the multiple groups of candidate load boundaries; calculating the cost values ​​of the several groups of candidate load boundaries, and taking one group of candidate load boundaries with the cost values ​​that meet the preset conditions among the several groups of candidate load boundaries as the optimal load boundary. Here, the cost values ​​of the several groups of candidate load boundaries are determined based on the weight parameters of various loads, the candidate load boundaries of various loads, and the reference values ​​of various loads.

[0139] Optionally, the operation of determining multiple groups of candidate load boundaries that meet the target pass rate from multiple groups of candidate load boundaries may include: determining N groups of candidate load boundaries that meet the following conditions from multiple groups of candidate load boundaries as several groups of candidate load boundaries that meet the target pass rate: for each group of candidate load boundaries in the N groups of candidate load boundaries, the number of points that meet the requirement that each load is less than the corresponding reference value in all points exceeds a predetermined threshold. Here, the predetermined threshold is the product of the target pass rate and the total number of all points, and N is a natural number greater than 1.

[0140] According to an embodiment of the present disclosure, the adaptability determination module 640 is configured to compare the predicted load data with the corresponding optimal load boundary, and evaluate the adaptability of the wind turbine to be developed at the predetermined site according to the comparison result.

[0141] It should be understood that the specific processing performed by the adaptability evaluation device of the wind turbine generator set according to the exemplary embodiment of the present disclosure has been referred to. Figures 1 to 5 It has been described in detail and will not be repeated here.

[0142] It should be understood that each module in the adaptability assessment device of a wind turbine according to the exemplary embodiment of the present disclosure may be implemented as a hardware component and / or a software component. Those skilled in the art may implement each module, for example, using a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC) according to the processing performed by each defined module.

[0143] An exemplary embodiment of the present disclosure provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to execute the adaptability assessment method of a wind turbine as described in the above exemplary embodiment. The computer-readable storage medium is any data storage device that can store data read by a computer system. Examples of computer-readable storage media include: read-only memory, random access memory, read-only optical disk, magnetic tape, floppy disk, optical data storage device, and carrier wave (such as data transmission through the Internet via a wired or wireless transmission path).

[0144] Figure 7 A structural block diagram of a computer device according to an exemplary embodiment of the present disclosure is shown.

[0145] The computer device 700 according to the exemplary embodiment of the present disclosure includes: a processor 710 and a memory 720. The memory 720 stores a computer program, and when the computer program is executed by the processor 710, the processor 710 is prompted to execute the adaptability assessment method of the wind turbine generator system as described in the above exemplary embodiment.

[0146] Although some exemplary embodiments of the present disclosure have been shown and described, it will be appreciated by those skilled in the art that modifications may be made to these embodiments without departing from the principles and spirit of the present disclosure, the scope of which is defined by the claims and their equivalents.

Claims

1. A wind turbine adaptability assessment method, It is characterized in that The adaptability assessment method includes: Obtain the design parameters of the units to be developed and the target wind parameters of the intended site; Calculating predicted load data of various loads of the unit to be developed at the predetermined site based on the design parameters of the unit to be developed and the target wind parameters of the predetermined site; Calculating the optimal load boundaries for various loads of the unit to be developed based on the design parameters of the unit to be developed; The predicted load data is compared with the corresponding optimal load boundary, and the adaptability of the wind turbine to be developed at the predetermined site is evaluated according to the comparison result.

2. The adaptability assessment method according to claim 1, It is characterized in that The step of calculating the predicted load data of various loads of the unit to be developed at the predetermined site includes: Get reference wind parameters; By inputting the design parameters of the unit to be developed, the target wind parameter of the predetermined site and the reference wind parameter into the load proxy model, first load data of various loads of the unit to be developed under the reference wind parameter and the target wind parameter are calculated; By inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load trend prediction model, second load data of various loads of the unit to be developed at the predetermined site are calculated; The predicted load data is calculated based on the first load data and the second load data.

3. The adaptability assessment method according to claim 2, It is characterized in that The load trend prediction model is constructed by the following steps: Obtaining unit parameters and corresponding modeling parameters of multiple known wind turbine units; Calculating load data under various working conditions based on the unit parameters of the plurality of known wind turbine units, the corresponding modeling parameters and the reference wind parameters; Based on the reference wind parameter and the load data under the various working conditions, a load trend prediction model under the various working conditions is constructed by a predetermined algorithm as the load trend prediction model.

4. The adaptability assessment method according to claim 2, It is characterized in that The design parameters of the unit to be developed include the design wind parameter and the target pass rate of the unit to be developed, and the step of calculating the optimal load boundary for various loads of the unit to be developed includes: Obtaining historical wind parameters associated with the predetermined site; By inputting the designed wind parameter, the target passing rate and the historical wind parameter into the load proxy model, fourth load data of each point of the predetermined site under the designed wind parameter is calculated; The optimal load boundary is calculated based on the design wind parameter, the target pass rate and the fourth load data.

5. The adaptability assessment method according to claim 4, It is characterized in that The step of calculating the optimal load boundary includes: Determine multiple groups of candidate load boundaries including all loads based on the reference values ​​of various loads corresponding to the design wind parameters and the maximum values ​​of various loads in the fourth load data; Determining a plurality of groups of candidate load boundaries that meet the target pass rate from among the plurality of groups of candidate load boundaries; Calculating cost values ​​of the plurality of groups of candidate load boundaries, and taking a group of candidate load boundaries having cost values ​​satisfying a preset condition among the plurality of groups of candidate load boundaries as an optimal load boundary, The cost values ​​of the plurality of groups of candidate load boundaries are determined based on weight parameters of various loads, candidate load boundaries of various loads and reference values ​​of various loads.

6. The adaptability assessment method according to claim 5, It is characterized in that The step of determining a plurality of groups of candidate load boundaries satisfying the target pass rate from the plurality of groups of candidate load boundaries comprises: Determine N groups of candidate load boundaries that satisfy the following conditions among the multiple groups of candidate load boundaries as several groups of candidate load boundaries that satisfy the target pass rate: for each group of candidate load boundaries among the N groups of candidate load boundaries, the number of points in all points that satisfy the condition that each load is less than the corresponding benchmark value exceeds a predetermined threshold, wherein the predetermined threshold is the product of the target pass rate and the total number of all points, and N is a natural number greater than 1.

7. The adaptability assessment method according to claim 2, It is characterized in that The load proxy model and the load trend prediction model are respectively constructed by a predetermined algorithm based on the reference wind parameter and at least one known unit parameter of the unit.

8. The adaptability assessment method according to claim 7, It is characterized in that The predetermined algorithm includes at least one of a multivariate linear regression algorithm and a neural network algorithm.

9. The adaptability assessment method according to claim 2, It is characterized in that The step of calculating first load data of various loads of the unit to be developed under the reference wind parameter and the target wind parameter comprises: By inputting the design parameters of the unit to be developed and the reference wind parameter into the load proxy model, third load data of various loads of the unit to be developed under the reference wind parameter is calculated; By inputting the design parameters of the unit to be developed and the target wind parameters of the predetermined site into the load proxy model, fourth load data of various loads of the unit to be developed under the target wind parameters are calculated, The first payload data includes the third payload data and the fourth payload data.

10. The adaptability assessment method according to claim 2, It is characterized in that The load proxy model is constructed by the following steps: Acquire modeling parameters of a known wind turbine generator set that is closest to the design parameters of the turbine generator set to be developed; Based on the modeling parameters and the reference wind parameters, calculating load data under various working conditions; Based on the reference wind parameters and the load data under the various working conditions, a load proxy model under the various working conditions is constructed by a predetermined algorithm as the load proxy model.

11. A computer-readable storage medium, It is characterized in that When the instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to perform the adaptability assessment method according to any one of claims 1 to 10.

12. A computer device, It is characterized in that include: at least one processor; at least one memory storing computer executable instructions, Wherein, when the computer executable instructions are executed by the at least one processor, the at least one processor is prompted to perform the adaptability assessment method according to any one of claims 1 to 10.

Images