Pulse wind speed spectrum evaluation method, device and equipment based on envelope line method and storage medium

Abstract

The invention provides a fluctuating wind speed spectrum evaluation method, device and equipment based on an envelope line method and a storage medium, and aims at more accurately capturing and representing turbulent flow wind speed spectrum characteristics in different seasons and under different direction conditions and improving the stability and safety of a wind turbine. Actually measured data are analyzed through Fourier transform, and the upper and lower envelope line technology is combined, so that the problem that a traditional single empirical spectrum cannot comprehensively reflect a complex turbulence structure of a real wind field is effectively solved. And by utilizing a machine learning algorithm, high-precision prediction of the upper envelope line, the mean value curve and the lower envelope line is realized, and a scientific basis is provided for design and maintenance of the wind turbine.

Description

Technical Field

[0001] The present invention relates to the field of wind power generation technology, and specifically to a pulsating wind speed spectrum evaluation method, device, equipment and storage medium based on an envelope method, which aims to improve the stability and safety of wind turbines in complex turbulent environments and optimize the design and maintenance strategies of wind turbines. Background Art

[0002] Characterizing the turbulent energy of eddies of different scales using the turbulent wind power density spectrum is an important basis for evaluating turbulence intensity. Accurate evaluation of the turbulent wind speed spectrum is crucial for the design and optimization of wind turbines and the operation optimization of wind farms. Based on field measurements, a series of empirical analysis formulas for the power spectra of fluctuating wind speeds in the downwind, crosswind and vertical wind directions with deterministic parameters were established, and verified under normal wind and typhoon conditions. However, natural wind has inherent uncertainty and randomness, and the traditional single empirical spectrum may not fully reflect the complex fluctuation structure in the real wind field. Therefore, based on a large amount of measured wind data, the uncertainty wind spectrum was studied. Fenerci and In theory, a non-deterministic wind spectrum model composed of spectral parameters of log-normal distribution was proposed. Liu et al. studied the log-normal distribution parameters of the measured wind spectrum at Jiangyin Bridge. Fang et al. proposed the generalized extreme value distribution of the uncertain wind spectrum at Xi'aomen Bridge. Liu et al. proposed a random wind field dimensionality reduction simulation method that considers the probability of truncated fluctuating wind spectrum to evaluate wind energy. However, most of the above studies are based on the establishment of parameter models based on existing empirical turbulence spectra. When applying these models, it is first necessary to use various models and calculation methods to estimate the spectral parameters, and then use the empirical spectrum model to describe the relationship between turbulent energy and frequency. When applied to the calculation of wind load and wind-induced vibration response of wind turbines and numerical simulation of the atmosphere, the process is relatively cumbersome. Therefore, in order to more accurately and simply evaluate the turbulent wind speed spectrum, improve the performance of wind turbines and the operating efficiency of wind farms, we urgently need a more concise, efficient and pulsating wind speed spectrum evaluation method that can fully reflect the characteristics of the real wind field. Summary of the invention

[0003] The purpose of the present invention is to solve the problems existing in the prior art and to provide a method, device, equipment and storage medium for evaluating a pulsating wind speed spectrum based on an envelope method.

[0004] The object of the present invention is achieved through the following technical scheme: a fluctuating wind speed spectrum evaluation method based on the envelope method, the method comprising the following steps:

[0005] (1) Obtain wind speed time series data in the monitoring area and perform preprocessing;

[0006] (2) Perform seasonal and directional Fourier transform on the preprocessed wind speed time series data, and calculate and obtain the power spectrum density at each frequency;

[0007] (3) The upper envelope is constructed with the maximum value range in the spectrum, and the lower envelope is constructed with the minimum value range in the spectrum;

[0008] (4) Using the constructed upper and lower envelope data of the pulsating wind speed spectrum, a machine learning model is trained to predict the upper and lower envelopes of future wind speed spectrum data, thereby obtaining the detailed characteristics of the pulsating wind speed spectrum.

[0009] Furthermore, the acquired wind speed time series data is preprocessed, including: using a preset outlier identification method to identify and delete outlier data in the wind speed data, and using a preset missing value processing method to interpolate missing data in the wind speed data to obtain preprocessed data.

[0010] Furthermore, the seasonal and directional Fourier transform of the preprocessed wind speed data includes: the seasons are divided into spring (November-January), summer (February-April), autumn (May-July) and winter (August-October), and the wind directions are divided into longitudinal, lateral and vertical directions; the three-way pulsating wind speed of the four seasons is decomposed by the Fourier transform method, the time domain signal is converted into a frequency domain signal, and the power spectral density at each frequency is calculated and obtained to reflect the spectral characteristics of the wind speed.

[0011] Furthermore, the upper envelope and the lower envelope are optimized through smoothing and extreme point screening to more accurately describe the dynamic changes of the wind speed spectrum and comprehensively characterize the range and distribution characteristics of the turbulent wind speed spectrum.

[0012] Furthermore, the use of a machine learning algorithm to train the upper and lower envelope data sets includes: the data sets are divided into a training set and a test set in a ratio of 3:1, wherein the training set is used for learning and the test set is used to evaluate the algorithm model; the model takes the frequency f as input and the logarithm of the power spectral density as output, and establishes a machine learning evaluation model for the wind speed spectrum envelope; during the training process, the machine learning algorithm parameters are tuned, and the corresponding loss function values ​​are calculated by trying different parameter values, and the parameter combination that minimizes the loss function is selected as the final model parameter setting.

[0013] Furthermore, the prediction results of the machine learning model are output in a visual form for use by wind turbine designers, operation and maintenance personnel, and scientific researchers to optimize wind turbine design and improve operational efficiency and safety.

[0014] Furthermore, the newly collected wind speed data were evaluated using the trained optimal machine learning model. A variety of machine learning algorithms were used for training. The predicted results based on the test set were compared with the measured results. The root mean square error (RMSE) and the coefficient of determination (R) were used to compare the wind speed data. 2 The size of determines the optimal model to evaluate the newly collected wind speed data.

[0015] To achieve the above object, the present invention also provides the following technical solution: a pulsating wind speed spectrum evaluation device based on the envelope method, comprising:

[0016] Data acquisition module, used to collect wind speed time series data in the monitoring area in real time or at regular intervals;

[0017] A data preprocessing module, used to perform data cleaning operations;

[0018] The spectrum analysis module is used to perform seasonal and directional Fourier transform on the cleaned wind speed time series data, calculate and obtain the power spectrum density at each frequency;

[0019] The envelope construction module is used to construct upper and lower envelopes according to the power spectrum density data; specifically, the upper envelope is constructed with the maximum value range in the spectrum, and the lower envelope is constructed with the minimum value range in the spectrum;

[0020] A machine learning module is used to train and apply a machine learning model using the constructed upper and lower envelope data of the pulsating wind speed spectrum to predict the upper and lower envelopes of the wind speed spectrum data, thereby obtaining detailed features of the pulsating wind speed spectrum;

[0021] The result output module is used to display the evaluation and prediction results.

[0022] To achieve the above-mentioned purpose, the present invention also provides the following technical solution: an electronic device, comprising a memory for storing a computer program; and a processor for executing the computer program to implement the pulsating wind speed spectrum evaluation method based on the envelope method.

[0023] To achieve the above-mentioned purpose, the present invention also provides the following technical solution: a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the pulsating wind speed spectrum evaluation method based on the envelope method.

[0024] Beneficial effects of the present invention: The present invention provides a pulsating wind speed spectrum evaluation method based on the envelope method and related devices and equipment. The method uses high-precision wind speed measurement equipment and detailed data preprocessing steps. The present invention ensures the accuracy and consistency of wind speed data and improves the accuracy of wind speed spectrum evaluation. In addition, combined with seasonal Fourier transform analysis and envelope construction, the range and distribution characteristics of the turbulent wind speed spectrum can be fully characterized, thereby improving the accuracy of wind speed spectrum evaluation. Finally, by using advanced machine learning algorithms, the complex relationship between historical wind speed spectrum data and envelope information is learned, thereby improving the ability to predict future wind speed spectra. The present invention provides strong technical support for fields such as wind power generation and meteorological observation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a flow chart of an embodiment of a fluctuating wind speed spectrum evaluation method based on an envelope method of the present application;

[0026] Figure 2 This is a schematic diagram of the module structure of a pulsating wind speed spectrum evaluation device based on the envelope method of the present application;

[0027] Figure 3 This is a schematic diagram of the equipment structure of the hardware operating environment involved in a pulsating wind speed spectrum evaluation method based on the envelope method in this application. DETAILED DESCRIPTION

[0028] In order to facilitate ordinary technicians in the field to understand and implement the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the implementation examples described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0029] An embodiment of the present invention is a method for evaluating the pulsating wind speed spectrum using the envelope method based on the measured data of the Doppler wind radar, such as Figure 1 As shown, the specific steps are as follows:

[0030] High-precision wind speed measurement equipment was used to continuously collect one year of observation data from Doppler wind radars deployed in the Shagohuang area of ​​Yumen, Gansu Province, northwest China. The sampling frequency was at least 1 Hz, including data in the x, y, and z directions. The wind speed data was then cleaned and outliers were identified and removed using a moving median filter with a window size of 600. Missing values ​​were then interpolated using linear interpolation.

[0031] The three-way wind speed data is processed according to the following formula to calculate the longitudinal, transverse and vertical wind speeds. Longitudinal: Horizontal: Vertical: Among them, u(t), v(t) and w(t) represent the longitudinal, lateral and vertical wind speeds respectively, and U is the 10-minute average horizontal wind speed. is the mean wind direction angle, u x (t),u y (t) and u z (t) represents the instantaneous wind speed in the x, y, and z directions respectively, Represents the average wind speed in the z direction.

[0032] The three-way pulsating wind speed of the four seasons is decomposed using the Fourier transform method with a time interval of 10 minutes. The time domain signal is converted into a frequency domain signal, and the power spectrum density at each frequency is calculated and obtained to reflect the spectrum characteristics of the wind speed. Based on the calculated power spectrum density data, the upper envelope and the lower envelope are constructed, where the upper envelope represents the maximum value range in the spectrum, and the lower envelope represents the minimum value range in the spectrum. The envelope is optimized by means of smoothing, extreme point screening, etc., so that it can more accurately describe the dynamic changes of the wind speed spectrum.

[0033] Multilayer perceptron, support vector machine and random forest machine learning algorithms are used to train the model using historical wind speed spectrum data and corresponding envelope information to predict future wind speed spectra. Specifically, the data set is first divided into a training set and a test set for machine learning in a ratio of 3:1, where the training set is used for learning and the test set is used for training and evaluating the algorithm model; then the data is preprocessed using a logarithmic method, with frequency f as input and the logarithm of the power spectrum density as output for training. Through continuous parameter adjustment and optimization, it is determined that the multilayer perceptron training process is repeated for 100 cycles, and the number of samples used each time the weight is updated is 10; the support vector machine algorithm uses the radial basis function as the kernel function, and the hyperparameters for controlling the regularization strength and kernel function width are 100 and 0.1 respectively; the random forest algorithm contains 200 trees and the number of random number seeds is 42; then the RMSE and R of the three machine learning algorithms are compared. 2 , the optimal machine learning evaluation model for turbulent wind speed spectrum data in different seasons and directions was determined;

[0034] The newly collected wind speed data was preprocessed using the trained machine learning model, i.e., the one with the smallest RMSE and the largest R2, and then calculated and classified into 12 types of data, including four seasons of spring (November-January), summer (February-April), autumn (May-July), and winter (August-October) and three wind directions of longitudinal, lateral, and vertical. Subsequently, the optimal model under the 12 working conditions was used to evaluate the new data, predict the upper envelope, lower envelope, and mean curve, and thus obtain the detailed characteristics of the pulsating wind speed spectrum.

[0035] Outputting the evaluation and prediction results in a visual form, such as charts or reports, is not limited to using widely recognized tools such as Matplotlib or Tableau, but also combines interactive visualization libraries such as Plotly and D3.js to provide a more dynamic and intuitive analysis experience. These technologies enable complex wind speed spectrum characteristics, envelope predictions, and mean curves to be presented in the form of multi-dimensional charts, dynamic graphics, and interactive interfaces. Users can explore the details of the data in depth through operations such as zooming, dragging, and hovering. In addition, automatic data report generation technology has been introduced, combined with tools such as LaTeX or R Markdown, which can generate professional reports containing detailed data analysis, prediction results, and recommended measures with one click. These reports not only contain beautiful charts, but also detailed data explanations and conclusions, which greatly improve the efficiency of information transmission and the convenience of user decision-making.

[0036] An embodiment of the present invention provides a pulsating wind speed spectrum device based on an envelope method, Figure 2 is a device module diagram according to an exemplary embodiment, including:

[0037] The data acquisition module is used to collect wind speed data in real time or at regular intervals. As the front end of the pulsating wind speed spectrum device based on the envelope method, this module undertakes the key task of obtaining accurate and comprehensive wind speed data. The module is designed with an efficient data acquisition mechanism, which can collect data from various wind speed sensors in real time or according to preset time intervals. It supports multiple communication protocols to ensure seamless connection with anemometers, anemometers and other equipment of different models and brands. At the same time, the module also has data caching and breakpoint resumption functions, which can ensure the integrity and continuity of data even when the signal is unstable or the equipment fails. The collected wind speed data will undergo preliminary processing, such as timestamp marking, format unification, etc., to lay a solid foundation for subsequent data preprocessing and analysis.

[0038] The data preprocessing module is used to perform operations such as data cleaning, filtering and calibration. This module receives the original wind speed data from the data acquisition module and first cleans it to remove invalid, abnormal or repeated data points to ensure data quality. Then, the module normalizes or standardizes the data to eliminate dimensional differences and improve the accuracy of subsequent analysis. In addition, the module can also perform operations such as data smoothing and filtering to reduce noise interference and highlight the main characteristics of wind speed changes. After this series of refined preprocessing processes, the wind speed data is optimized, providing a reliable foundation for subsequent wind speed spectrum analysis, envelope construction and wind speed prediction.

[0039] The spectrum analysis module is used to perform Fourier transform and calculate power spectrum density. This module focuses on in-depth spectrum analysis of pre-processed wind speed data. This module uses advanced spectrum estimation techniques, such as fast Fourier transform (FFT), to convert time-domain wind speed signals into frequency-domain representations, revealing the different frequency components and their energy distribution contained in the wind speed. Through visualization methods such as spectrum graphs and power spectrum density, users can intuitively observe the characteristics of the wind speed spectrum, such as the main frequency band, energy peak, etc., which is crucial for understanding the dynamic behavior of wind speed, identifying potential wind speed change patterns, and subsequent envelope construction and wind speed prediction.

[0040] The envelope construction module is used to construct upper and lower envelopes based on the power spectrum density data. This module is responsible for accurately constructing the upper and lower envelopes of the wind speed spectrum based on the power spectrum density data output by the spectrum analysis module. This module uses advanced mathematical algorithms to capture the extreme values ​​of fluctuations in the wind speed spectrum and then form envelopes. These envelopes not only intuitively display the dynamic range of the wind speed spectrum, but also provide key boundary information for subsequent wind speed feature extraction, machine learning model training, and wind speed prediction, which helps to more accurately grasp the laws and trends of wind speed changes.

[0041] The machine learning module is used to train and apply machine learning models to predict wind speed spectra. This module uses advanced algorithms to perform deep learning and pattern recognition on preprocessed wind speed data and envelope features. This module can automatically extract key features of wind speed changes, establish a wind speed prediction model, and improve prediction accuracy through continuous learning and optimization. It supports a variety of machine learning algorithms, such as support vector machines, neural networks, random forests, etc., to meet the needs of wind speed prediction in different scenarios. Through the machine learning module, the device can achieve real-time prediction of wind speed, long-term trend analysis, and early warning of abnormal wind speed, providing a scientific basis for the operation management and structural safety assessment of wind farms.

[0042] The result output module is used to display the evaluation and prediction results. This module can generate a variety of charts and curves to clearly show the real-time dynamics of the wind speed spectrum, historical trends, and future wind speed conditions predicted by machine learning. At the same time, the intuitive display of the upper and lower envelopes helps users quickly grasp the extreme range of wind speed fluctuations. In order to meet the needs of different users, the module provides highly customized options, such as data filtering, time span setting, spectrum detail magnification, etc., to enhance the user's interactive experience. In addition, the module can automatically generate a detailed wind speed spectrum analysis report, covering an overview of wind speed characteristics, in-depth analysis of spectrum distribution, prediction accuracy assessment and uncertainty discussion, etc., providing a scientific basis for wind farm operations, structural safety assessments, and climate research. Users can also easily export report data for subsequent analysis or sharing with relevant parties.

[0043] An embodiment of the present invention provides an electronic device, Figure 3 This is a structural diagram of an electronic device according to an exemplary embodiment. The content in the diagram cannot be regarded as any limitation on the scope of use of the present application.

[0044] The electronic device may specifically include: at least one processor, at least one memory, a power supply, a communication interface, an input / output interface, and a communication bus. The memory is used to store a computer program, and the computer program is loaded and executed by the processor. In addition, the electronic device in this embodiment may specifically be an electronic computer.

[0045] In this embodiment, the power supply is used to provide working voltage for each hardware device on the electronic device; the communication interface can create a data transmission channel between the electronic device and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of this application, and it is not specifically limited here; the input and output interface is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and it is not specifically limited here. In addition, the memory, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating systems, computer programs, etc., and the storage method can be temporary storage or permanent storage.

[0046] The operating system is used to manage and control various hardware devices and computer programs on the electronic device, and it can be Windows Server, Netware, Unix, Linux, etc.

[0047] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program.

[0048] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0049] The computer-readable storage medium may be included in the pulsating wind speed spectrum evaluation device based on the envelope method; or may exist independently without being assembled into the pulsating wind speed spectrum evaluation device based on the envelope method.

[0050] The above-described embodiment is only a preferred solution of the present invention, but it is not intended to limit the present invention. A person skilled in the relevant technical field may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A fluctuating wind speed spectrum evaluation method based on the envelope method, characterized in that: The method comprises the following steps: (1) Obtain wind speed time series data in the monitoring area and perform preprocessing; (2) Perform seasonal and directional Fourier transform on the preprocessed wind speed time series data, and calculate and obtain the power spectrum density at each frequency; (3) The upper envelope is constructed with the maximum value range in the spectrum, and the lower envelope is constructed with the minimum value range in the spectrum; (4) Using the constructed upper and lower envelope data of the pulsating wind speed spectrum, a machine learning model is trained to predict the upper and lower envelopes of future wind speed spectrum data, thereby obtaining the detailed characteristics of the pulsating wind speed spectrum.

2. The method for evaluating the fluctuating wind speed spectrum based on the envelope method according to claim 1, characterized in that: The acquired wind speed time series data is preprocessed, including: using a preset outlier identification method to identify and delete outlier data in the wind speed data, and using a preset missing value processing method to interpolate missing data in the wind speed data to obtain preprocessed data.

3. The method for evaluating the fluctuating wind speed spectrum based on the envelope method according to claim 1, characterized in that: The seasonal and directional Fourier transform of the preprocessed wind speed data includes: the seasons are divided into spring (November-January), summer (February-April), autumn (May-July) and winter (August-October), and the wind directions are divided into longitudinal, transverse and vertical directions; the three-way pulsating wind speed of the four seasons is decomposed by the Fourier transform method, the time domain signal is converted into a frequency domain signal, and the power spectral density at each frequency is calculated and obtained to reflect the spectral characteristics of the wind speed.

4. A fluctuating wind speed spectrum evaluation method based on the envelope method according to claim 1, characterized in that: The upper envelope and lower envelope are optimized through smoothing and extreme point screening to more accurately describe the dynamic changes of the wind speed spectrum and comprehensively characterize the range and distribution characteristics of the turbulent wind speed spectrum.

5. The method for evaluating the fluctuating wind speed spectrum based on the envelope method according to claim 1, characterized in that: The method of using a machine learning algorithm to train the upper and lower envelope data sets includes: dividing the data sets into a training set and a test set in a ratio of 3:1, wherein the training set is used for learning and the test set is used for evaluating the algorithm model; the model takes the frequency f as input and the logarithm of the power spectrum density as output, and establishes a machine learning evaluation model for the wind speed spectrum envelope; during the training process, the machine learning algorithm parameters are tuned, and the corresponding loss function values ​​are calculated by trying different parameter values, and the parameter combination that minimizes the loss function is selected as the final model parameter setting.

6. The method for evaluating the fluctuating wind speed spectrum based on the envelope method according to claim 1, characterized in that: The prediction results of the machine learning model are output in a visual form for use by wind turbine designers, operation and maintenance personnel, and scientific researchers to optimize wind turbine design and improve operational efficiency and safety.

7. The method for evaluating the fluctuating wind speed spectrum based on the envelope method according to claim 1, characterized in that: The newly collected wind speed data were evaluated using the trained optimal machine learning model. A variety of machine learning algorithms were used for training. The predicted results based on the test set were compared with the measured results. The root mean square error (RMSE) and the coefficient of determination (R) were used to compare the wind speed data. 2 The size of determines the optimal model to evaluate the newly collected wind speed data.

8. A device for implementing the fluctuating wind speed spectrum evaluation method based on the envelope method as claimed in claim 1, characterized in that: include: Data acquisition module, used to collect wind speed time series data in the monitoring area in real time or at regular intervals; A data preprocessing module, used to perform data cleaning operations; The spectrum analysis module is used to perform seasonal and directional Fourier transform on the cleaned wind speed time series data, calculate and obtain the power spectrum density at each frequency; The envelope construction module is used to construct upper and lower envelopes according to the power spectrum density data; specifically, the upper envelope is constructed with the maximum value range in the spectrum, and the lower envelope is constructed with the minimum value range in the spectrum; A machine learning module is used to train and apply a machine learning model using the constructed upper and lower envelope data of the pulsating wind speed spectrum to predict the upper and lower envelopes of the wind speed spectrum data, thereby obtaining detailed features of the pulsating wind speed spectrum; The result output module is used to display the evaluation and prediction results.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the method for evaluating the pulsating wind speed spectrum based on the envelope method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, implements the pulsating wind speed spectrum evaluation method based on the envelope method as described in any one of claims 1 to 7.

Images