Wind resource assessment method and system based on optimized energy mode factor

Abstract

The invention discloses a wind resource assessment method and system based on an optimized energy mode factor. According to the method, data preprocessing, parameter estimation, model verification and final evaluation are integrated into a standardized and repeatable process, so that the consistency and reliability of evaluation results are ensured; an energy mode factor directly related to wind energy is introduced to serve as a core target, a trustworthy region broken line algorithm is adopted to optimize parameter solution for solving, and it is ensured that the finally obtained Weibull parameter can reflect the energy characteristics of wind most truly instead of just the statistical distribution form of the wind speed; due to the improvement of the Weibull parameter estimation precision, the average wind energy power density calculated on the basis of the distribution model is closer to a true value, scientific support can be provided for wind power plant site selection, fan type selection and wind energy potential evaluation, and fine development of wind power projects is promoted.

Description

Technical Field

[0001] This invention relates to the field of renewable energy technology, and specifically to a wind resource assessment method and system based on optimized energy pattern factors. Background Technology

[0002] Wind energy is a green, clean, and environmentally friendly renewable energy source. Accurately assessing wind resource potential requires defining the probability distribution function parameters suitable for wind speed data. Among these, the Weibull distribution has become the mainstream model for describing wind speed distribution due to its simplicity, flexibility, and ability to effectively express wind speed fluctuations. The shape and scale parameters of the Weibull distribution are crucial for defining wind speed distribution, supporting wind turbine design and site selection, improving energy production efficiency, and promoting the sustainable development of renewable energy.

[0003] Currently, the industry commonly estimates Weibull parameters using methods based on Energy Pattern Factors (EPFs). Existing methods include the traditional Energy Pattern Factor Method (EPFM), the Sathyajith-modified Energy Pattern Factor Method (EPFMS), and the novel Energy Pattern Factor Method (NEPFM). However, these methods have significant limitations: traditional EPFM and EPFMS rely on fixed methods to calculate parameters, failing to adequately adapt to the unique distribution characteristics of wind speed data in different regions, and are prone to errors; while NEPFM attempts to optimize parameter estimation capabilities, its fitting effect in practical applications remains unsatisfactory; furthermore, existing methods have limited application of empirical methods aided by mathematical or metaheuristic optimization, and most lack efficient optimization solution mechanisms, making it difficult to balance computational accuracy and robustness.

[0004] Meanwhile, existing research has shortcomings in method validation and data processing: some studies only use a few goodness-of-fit indices to measure performance, which cannot fully reflect the fitting effect between the model and the measured wind speed data; some studies have not fully considered the variation of wind speed with height, or have not corrected the data for different regional terrain features, which further affects the accuracy of Weibull parameter estimation, making it difficult to meet the needs of accurate wind resource assessment and restricting the scientific and economic development of wind power projects. Summary of the Invention

[0005] The purpose of this invention is to provide a wind resource assessment method and system based on optimized energy pattern factors, so as to solve the problem that current assessment technologies are difficult to meet the requirements for accurate wind resource assessment.

[0006] To address the aforementioned technical problems, in a first aspect, the present invention provides a wind resource assessment method based on optimized energy pattern factors, comprising the following steps:

[0007] S1: Obtain the raw wind speed data at a specified height in the target area, and standardize the collected raw wind speed data to obtain a standardized wind speed dataset at the target height.

[0008] S2: Calculate the cube mean of the wind speed samples in the standardized dataset and the mean wind speed of all wind speed samples in the standardized dataset. The ratio of the cube of the two factors yields the measured energy mode factor.

[0009] S3: Construct a nonlinear equation based on the difference between the energy mode factor of the Weibull distribution and the measured energy mode factor, and then use the trust region piecewise linear algorithm to iteratively find the root of the linear equation to obtain the shape parameters of the Weibull distribution. ;

[0010] S4: Based on shape parameters The mean wind speed of the standardized wind speed dataset And the gamma function, to calculate the scaling parameter of the Weibull distribution. ;

[0011] S5: Based on shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

[0012] Furthermore, the original wind speed data was sampled at an hourly frequency and covered the entire year; then, the following methods were used: The criteria remove abnormal raw wind speed data.

[0013] Furthermore, if the target area has flat terrain and a single type of surface cover, then a power-law formula is used. The raw wind speed data is standardized and corrected; if the target area has complex terrain or diverse land cover types, a logarithmic formula is used. The original wind speed data is standardized and corrected. During the standardization and correction process, the applicability of the power law formula or logarithmic formula is confirmed through cross-validation. If the cross-validation error exceeds the preset threshold, the parameters of the power law formula or logarithmic formula are adjusted until the error meets the requirements.

[0014] Furthermore, the nonlinear equation for the component based on the difference between the Weibull distribution energy mode factor and the measured energy mode factor is as follows:

[0015] ;

[0016] in: For gamma function, ; This is the measured energy pattern factor.

[0017] Furthermore, the iterative root-finding of linear equations using the trust region piecewise linear algorithm includes:

[0018] Initial value setting: The initial iteration value is set based on the statistical characteristics of wind speed fluctuations in the target area. ;

[0019] Iteration step size calculation: Calculate the function value and its Jacobian matrix corresponding to the nonlinear equation at the current iteration point; calculate the Cauchy step and Gauss-Newton step based on the Jacobian matrix respectively;

[0020] Final iteration step determination: Set the initial trust region radius, and according to the relationship between the magnitudes of the Cauchy step and the Gauss-Newton step and the current trust region radius, merge the two steps into the final iteration step size by adjusting the fusion coefficient; the final iteration step size does not exceed the current trust region radius.

[0021] Iteration termination judgment: Calculate the updated shape parameters and the corresponding nonlinear equation function values. If the function value is less than the preset error threshold and the parameter difference between two adjacent iterations is less than the preset accuracy threshold, the iteration terminates and the current shape parameter k is output; otherwise, the trust region radius is adjusted according to the degree of function value decrease, and the iteration step size calculation step is returned to continue iterating.

[0022] Furthermore, the Jacobian matrix is ​​calculated using numerical differentiation, and is obtained through the following formula:

[0023]

[0024] in, This is a disturbance.

[0025] Furthermore, scale parameters The calculation formula is:

[0026] .

[0027] Furthermore, the probability density function model of the Weibull distribution is as follows:

[0028] ;

[0029] The formula for calculating average wind power density is:

[0030] ;

[0031] in, The air density of the target area.

[0032] Furthermore, the five goodness-of-fit metrics include:

[0033] Root mean square error (RMSE) calculation ;

[0034] Mean Absolute Error (MAE) Calculation ;

[0035] Calculation of correlation coefficient R ;

[0036] Efficiency coefficient CoE calculation, ;

[0037] Calculation of maximum absolute error MaxAE ;

[0038] in, Let be the probability density function of the measured wind speed. The probability density function estimated for the model. This represents the mean of the measured probability density function.

[0039] In a first aspect, the present invention provides a wind resource assessment system based on optimized energy pattern factors, comprising:

[0040] The data acquisition and preprocessing module is used to acquire raw wind speed data at a specified height in the target area and to standardize the acquired raw wind speed data to obtain a standardized wind speed dataset at the target height.

[0041] The measured energy pattern factor calculation module is used to calculate the cube mean of wind speeds for all wind speed samples in the standardized dataset. The mean wind speed of all wind speed samples in the standardized dataset cubed The ratio of these values ​​is used to obtain the measured energy mode factor;

[0042] The shape parameter calculation module is used to construct a nonlinear equation based on the difference between the energy mode factor of the Weibull distribution and the measured energy mode factor. Then, a trust-region piecewise linear algorithm is used to iteratively solve the linear equation to obtain the shape parameters of the Weibull distribution. ;

[0043] The scale parameter calculation module is used to calculate the scale parameter based on the shape parameter. The mean wind speed of the standardized wind speed dataset That is, the gamma function, which is used to calculate the scaling parameter of the Weibull distribution. ;

[0044] Evaluation module for using shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

[0045] The beneficial effects of this invention are as follows: By integrating data preprocessing, parameter estimation, model verification, and final evaluation into a standardized and repeatable process, the consistency and reliability of the evaluation results are ensured; by introducing an energy pattern factor directly related to wind energy as the core objective and using a trust region piecewise linear algorithm to optimize parameter solving, the final Weibull parameters are ensured to most realistically reflect the energy characteristics of wind, rather than just the statistical distribution of wind speed; due to the improved accuracy of Weibull parameter estimation, the average wind power density calculated based on this distribution model is also closer to the true value, which can provide scientific support for wind farm site selection, wind turbine selection, and wind energy potential assessment, and promote the refined development of wind power projects. Attached Figure Description

[0046] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, use the same reference numerals to denote the same or similar parts. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0047] Figure 1 This is a flowchart of one embodiment of the present invention. Detailed Implementation

[0048] like Figure 1 The wind resource assessment method based on optimized energy pattern factors, as shown, includes the following steps:

[0049] S1: Obtain the raw wind speed data at a specified height in the target area, and standardize the collected raw wind speed data to obtain a standardized wind speed dataset at the target height.

[0050] S2: Calculate the cube mean of the wind speed samples in the standardized dataset and the mean wind speed of all wind speed samples in the standardized dataset. The ratio of the cube of the two factors yields the measured energy mode factor.

[0051] S3: Construct a nonlinear equation based on the difference between the energy mode factor of the Weibull distribution and the measured energy mode factor, and then use the trust region piecewise linear algorithm to iteratively find the root of the linear equation to obtain the shape parameters of the Weibull distribution. ;

[0052] S4: Based on shape parameters The mean wind speed of the standardized wind speed dataset And the gamma function, to calculate the scaling parameter of the Weibull distribution. ;

[0053] S5: Based on shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

[0054] This invention integrates data preprocessing, parameter estimation, model validation, and final evaluation into a standardized and repeatable process, ensuring the consistency and reliability of the evaluation results. By introducing an energy pattern factor directly related to wind energy as the core objective and employing a trust region piecewise linear algorithm to optimize parameter solving, it ensures that the final Weibull parameters most accurately reflect the energy characteristics of wind, rather than merely the statistical distribution of wind speed. Due to the improved accuracy of Weibull parameter estimation, the average wind power density calculated based on this distribution model is also closer to the true value, providing scientific support for wind farm site selection, wind turbine selection, and wind energy potential assessment, and promoting the refined development of wind power projects.

[0055] According to one embodiment of this application, the designated height for collecting raw wind speed data is preferably 50m, and the reference height for referencing the raw wind speed data can be preferably 10m. The sampling frequency of the raw wind speed data is hourly, and the raw wind speed data needs to cover the entire year to ensure statistical representativeness. The data source needs to meet meteorological observation-level accuracy standards. Then, the following is adopted... The criterion is to remove abnormal raw wind speed data, i.e., to calculate the average wind speed. with standard deviation Eliminate those that meet the requirements or The sample size is determined by the weather record. If extreme weather records exist in the target area, the validity of the extreme wind speed sample needs to be manually confirmed in conjunction with the meteorological records to avoid accidental rejection.

[0056] This embodiment employs the 3σ criterion and combines it with manual verification of meteorological records. This effectively eliminates abnormal data caused by equipment malfunctions or other reasons, while retaining valid wind speed samples under extreme weather conditions. This improves the robustness and accuracy of all subsequent analyses from the data source.

[0057] According to one embodiment of this application, if the terrain of the target area is flat (plain, desert, etc.) and the surface cover type is uniform (such as grassland, farmland, etc.), then a power-law formula is used. The original wind speed data is standardized and corrected; among them, For a specified height Wind speed, 10m reference height ( Wind speed at that location, The coefficient of friction is 0.12-0.25 in plain areas and 0.25-0.35 in suburban areas.

[0058] If the target area has complex terrain (mountains, hills, etc.) or diverse land cover types (such as forests, built-up areas, etc.), then the logarithmic formula is used. The raw wind speed data is standardized and corrected; among which... The surface roughness length is defined as follows: grassland 0.03-0.1m, farmland 0.1-0.2m, forest 0.5-2m. After selection, its applicability is confirmed through cross-validation (substituting measured data from a portion of the 10m section with the specified height into the formula to calculate the error). If the error exceeds 5%, adjustments are made. or Continue until the error meets the requirements;

[0059] During the standardization and correction of the original wind speed data, the applicability of the power law formula or logarithmic formula is confirmed through cross-validation. If the cross-validation error exceeds the preset threshold (e.g., 5%), the parameters of the power law formula or logarithmic formula are adjusted until the error meets the requirements.

[0060] This embodiment specifies the most suitable vertical extrapolation model for wind speed in both flat and complex terrains, ensuring that the selected wind speed correction model and its parameters are optimal for the specific site, thereby improving the accuracy of the standardized wind speed dataset and providing high-quality input for subsequent core calculations.

[0061] According to one embodiment of this application, the nonlinear equation of the component based on the difference between the Weibull distribution energy mode factor and the measured energy mode factor is:

[0062] ;

[0063] in: For gamma function, ; This is the measured energy pattern factor.

[0064] According to one embodiment of this application, iteratively finding the root of a linear equation using a trust region piecewise linear algorithm includes:

[0065] Initial value setting: The initial iteration value is set based on the statistical characteristics of wind speed fluctuations in the target area. The value range is 1.8-2.5. If the wind speed in the target area fluctuates greatly, the initial value should be 1.8-2.1. If the wind speed fluctuates little, the initial value should be 2.1-2.5.

[0066] Iteration step size calculation: at the current iteration point ( Calculate the function value corresponding to the nonlinear equation (for the number of iterations). and its Jacobian matrix Based on Jacobian matrix Calculate the Cauchy step and the Gauss-Newton step separately;

[0067] Final iteration step determined: setting the initial trust region radius According to Cauchy ( (The step size coefficient is determined through line search) and the Gauss-Newton step. ( The relationship between the magnitude of the inverse Jacobian matrix and the radius of the current trust region is used to fuse the two steps into the final iteration step size by adjusting the fusion coefficients. : , Ensure iterative stability; the final iteration step size does not exceed the current trust region radius;

[0068] Iteration termination criteria: Calculate the updated shape parameters and the corresponding nonlinear equation function values If the function value is less than the preset error threshold ( And the parameter difference between two adjacent iterations is less than the preset precision threshold. If the shape parameter is not found, the iteration terminates and the current shape parameter is output. Otherwise, adjust the trust region radius according to the degree of function value decrease (when the decrease ratio is > 0.75). When <0.25, Then return to the iteration step calculation step and continue iterating.

[0069] This embodiment cleverly combines the robustness of the steepest descent method (Cauchy step) in the initial convergence stage with the fast convergence of the Gauss-Newton method when approaching the true solution; compared to the fixed formula of traditional EPF-like methods, it achieves convergence through iterative optimization. It provides accurate solutions and only requires a single linear solution to update parameters, balancing efficiency and robustness. Compared to genetic algorithms and particle swarm optimization, it can obtain optimal parameters without multiple runs, reducing computational costs.

[0070] According to one embodiment of this application, the Jacobian matrix is ​​calculated using numerical differentiation, and is obtained through the following formula:

[0071]

[0072] in, For perturbation, take values Ensure the accuracy and stability of gradient calculation.

[0073] According to one embodiment of this application, the scale parameter The calculation formula is:

[0074] ;

[0075] Among them, numerical integration method is used to calculate (Integration step size 0.001, integration limit 100) to ensure accuracy.

[0076] According to one embodiment of this application, the Weibull distribution probability density function (PDF) model is as follows:

[0077] ;

[0078] The formula for calculating average wind power density is:

[0079] ;

[0080] in, The air density for the target area is taken as 1.225 kg / m³ under standard atmospheric pressure, and needs to be corrected according to altitude and temperature.

[0081] According to one embodiment of this application, the five goodness-of-fit indices include:

[0082] Root mean square error (RMSE) calculation ;

[0083] Mean Absolute Error (MAE) Calculation ;

[0084] Calculation of correlation coefficient R ;

[0085] Efficiency coefficient CoE calculation, ;

[0086] Calculation of maximum absolute error MaxAE ;

[0087] in, Let be the probability density function of the measured wind speed. The probability density function estimated for the model. This represents the mean of the measured probability density function.

[0088] when and The Weibull distribution probability density function (PDF) model meets the standard. After meeting the standard, the wind power density (WPD) is calculated, and then the wind resource potential level is quantified (e.g., WPD≥300W / m³ is a high-quality wind resource area) to complete the assessment.

[0089] This implementation employs five goodness-of-fit indices with different focuses (RMSE for overall error, MAE for average deviation, R for linear correlation, CoE for model explanatory power, and MaxAE for worst-case scenario), which can comprehensively and thoroughly evaluate the goodness of fit between the Weibull model and the measured data.

[0090] In a first aspect, the present invention provides a wind resource assessment system based on optimized energy pattern factors, comprising:

[0091] The data acquisition and preprocessing module is used to acquire raw wind speed data at a specified height in the target area and to standardize the acquired raw wind speed data to obtain a standardized wind speed dataset at the target height.

[0092] The measured energy pattern factor calculation module is used to calculate the cube mean of wind speeds for all wind speed samples in the standardized dataset. The mean wind speed of all wind speed samples in the standardized dataset cubed The ratio of these values ​​is used to obtain the measured energy mode factor;

[0093] The shape parameter calculation module is used to construct a nonlinear equation based on the difference between the energy mode factor of the Weibull distribution and the measured energy mode factor. Then, a trust-region piecewise linear algorithm is used to iteratively solve the linear equation to obtain the shape parameters of the Weibull distribution. ;

[0094] The scale parameter calculation module is used to calculate the scale parameter based on the shape parameter. The mean wind speed of the standardized wind speed dataset That is, the gamma function, which is used to calculate the scaling parameter of the Weibull distribution. ;

[0095] Evaluation module for using shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A wind resource assessment method based on optimized energy pattern factors, characterized in that, Includes the following steps: S1: Obtain the raw wind speed data at a specified height in the target area, and standardize the collected raw wind speed data to obtain a standardized wind speed dataset at the target height. S2: Calculate the cube mean of the wind speed of all wind speed samples in the standardized dataset and the mean wind speed of all wind speed samples in the standardized dataset. The ratio of the cube of the two factors yields the measured energy mode factor. S3: Construct a nonlinear equation based on the difference between the Weibull distribution energy mode factor and the measured energy mode factor, and then use the trust region piecewise linear algorithm to iteratively find the root of the linear equation to obtain the shape parameters of the Weibull distribution. ; S4: According to the shape parameters The mean wind speed of the standardized wind speed dataset And the gamma function, to calculate the scaling parameter of the Weibull distribution. ; S5: Based on the shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

2. The wind resource assessment method based on optimized energy pattern factors according to claim 1, characterized in that, The original wind speed data was sampled at an hourly rate and covered the entire year; then, it was used... The criteria remove abnormal raw wind speed data.

3. The wind resource assessment method based on optimized energy pattern factors according to claim 1 or 2, characterized in that, If the target area has flat terrain and a single type of surface cover, then the power law formula is used. The original wind speed data is standardized and corrected; if the target area has complex terrain or diverse land cover types, a logarithmic formula is used. The original wind speed data is standardized and corrected. During the standardization and correction process, the applicability of the power law formula or logarithmic formula is confirmed through cross-validation. If the cross-validation error exceeds a preset threshold, the parameters of the power law formula or logarithmic formula are adjusted until the error meets the requirements.

4. The wind resource assessment method based on optimized energy pattern factors according to claim 1, characterized in that, The nonlinear equation for the component of the difference between the Weibull distribution energy mode factor and the measured energy mode factor is: ； in: For gamma function, ; This is the measured energy pattern factor.

5. The wind resource assessment method based on optimized energy pattern factors according to claim 4, characterized in that, The step of iteratively finding the root of the linear equation using the trust region piecewise linear algorithm includes: Initial value setting: The initial iteration value is set based on the statistical characteristics of wind speed fluctuations in the target area. ; Iteration step size calculation: Calculate the function value and its Jacobian matrix corresponding to the nonlinear equation at the current iteration point; calculate the Cauchy step and Gauss-Newton step based on the Jacobian matrix respectively; Final iteration step determination: Set an initial trust region radius, and based on the relationship between the magnitudes of the Cauchy step and the Gauss-Newton step and the current trust region radius, merge the two steps into a final iteration step size by adjusting the fusion coefficient; the final iteration step size does not exceed the current trust region radius. Iteration termination judgment: Calculate the updated shape parameters and the corresponding nonlinear equation function values. If the function value is less than the preset error threshold and the parameter difference between two adjacent iterations is less than the preset accuracy threshold, the iteration terminates and the current shape parameter k is output; otherwise, the trust region radius is adjusted according to the degree of function value decrease, and the iteration step size calculation step is returned to continue iterating.

6. The wind resource assessment method based on optimized energy pattern factors according to claim 5, characterized in that, The Jacobian matrix is ​​calculated using numerical differentiation and obtained through the following formula: in, This is a disturbance.

7. The wind resource assessment method based on optimized energy pattern factors according to claim 1, characterized in that, The scale parameter The calculation formula is: 。 8. The wind resource assessment method based on optimized energy pattern factors according to claim 1, characterized in that, The Weibull distribution probability density function model is as follows: ； The formula for calculating the average wind power density is as follows: ； in, The air density of the target area.

9. The wind resource assessment method based on optimized energy pattern factors according to claim 8, characterized in that, The five goodness-of-fit indices include: Root mean square error (RMSE) calculation ; Mean Absolute Error (MAE) Calculation ; Calculation of correlation coefficient R ; Efficiency coefficient CoE calculation, ; Calculation of maximum absolute error MaxAE ; in, Let be the probability density function of the measured wind speed. The probability density function estimated for the model. This represents the mean of the measured probability density function.

10. A wind resource assessment system based on optimized energy pattern factors, characterized in that, include: The data acquisition and preprocessing module is used to acquire raw wind speed data at a specified height in the target area and to standardize the acquired raw wind speed data to obtain a standardized wind speed dataset at the target height. The measured energy pattern factor calculation module is used to calculate the cube mean of wind speeds for all wind speed samples in the standardized dataset. The mean wind speed of all wind speed samples in the standardized dataset cubed The ratio of these values ​​is used to obtain the measured energy mode factor; The shape parameter calculation module is used to construct a nonlinear equation based on the difference between the Weibull distribution energy mode factor and the measured energy mode factor. Then, a trust-region piecewise linear algorithm is used to iteratively solve the linear equation to obtain the shape parameters of the Weibull distribution. ; The scale parameter calculation module is used to calculate the scale parameter based on the shape parameter. The mean wind speed of the standardized wind speed dataset That is, the gamma function, which is used to calculate the scaling parameter of the Weibull distribution. ; Evaluation module, used to evaluate based on the shape parameters and scale parameters A probability density function model of the Weibull distribution is constructed, and five goodness-of-fit indices are used to verify the fitting accuracy between the model and the measured wind speed data. Then, the average wind power density of the target area is calculated based on the verified probability density function model to complete the wind resource assessment.

Images