Method for acquiring wind speed and wind direction of wind turbine generator

Abstract

The invention relates to the field of wind power generation, and discloses a wind turbine generator wind speed and wind direction acquisition method which comprises the following steps: 1) acquiring wind speed and wind direction data of a wind turbine generator operation area in real time through a sensor; 2) eliminating high-frequency disturbance based on a turbulence model; 3) correcting wind speed and wind direction data based on a wind shear model, and compensating a wind speed gradient; 4) constructing a wind field optimization model, and dynamically optimizing wind speed and wind direction data by using a global optimization method; 5) fusing the macroscopic trend and the local details based on multi-scale decomposition to generate final wind speed and wind direction data; and 6) outputting high-precision corrected wind speed and wind direction data. By introducing turbulence correction and wind shear models, main measurement errors caused by turbulence disturbance and height difference are eliminated for complex wind field characteristics in a wind turbine generator operation area. The layering correction method effectively reduces the randomness and deviation of the wind speed and direction data, and enables the output data to have high reliability and accuracy in various wind field environments.

Description

Technical Field

[0001] The invention relates to the technical field of wind power generation, and in particular to a method for acquiring wind speed and direction of a wind turbine set. Background Art

[0002] The efficient operation of wind turbines is inseparable from the accurate acquisition of wind speed and direction data. At present, the measurement technology of wind speed and direction has made certain progress. The traditional method mainly relies on mechanical anemometers and wind vanes. With the increase in measurement needs, new equipment such as lidar and ultrasonic sensors are gradually introduced. Through these devices, real-time monitoring of wind speed and wind direction can be achieved, and the measurement accuracy and environmental adaptability can be improved to a certain extent. In addition, some methods try to deal with turbulent interference by introducing mathematical models, or compensate for wind shear effects by simple gradient correction, so as to further improve the credibility of the data. These technical means provide basic support for the operation and control of wind turbines. However, in complex wind field environments, these technologies still have room for improvement when dealing with the dynamic characteristics and complex changes of wind speed and direction.

[0003] Existing technologies still face some limitations when solving key problems in wind speed and direction measurement. First, the processing capacity of existing methods for high-frequency disturbances of turbulence is limited, especially in scenes with severe turbulent disturbances, the measurement data is easily affected by random fluctuations and lacks stability. Secondly, the processing of wind shear effects mostly uses fixed correction factors, which is difficult to adapt to the differences in wind speed gradients at different heights and operating areas of wind turbines. In addition, the processing of wind speed and direction data is difficult to balance the continuity of the global wind field and the capture of local dynamic characteristics. Existing technologies tend to be more inclined to local corrections, resulting in insufficient overall consistency of the data. Finally, although the optimization model can improve the measurement results in some scenarios, the dynamic adaptability of the existing model is weak and fails to fully cope with the real-time changes in the wind field environment. Therefore, when dealing with complex wind field environments, existing technologies still need further breakthroughs to better meet the multiple requirements of high precision, dynamic adaptability and global consistency. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a method for obtaining wind speed and direction of a wind turbine set, which solves the problems of measurement errors and insufficient dynamic adaptability caused by turbulent disturbances, wind shear effects and complex wind field environments during the acquisition of wind speed and direction data in the operating area of ​​the turbine set.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for obtaining wind speed and direction of a wind turbine generator system, comprising the following steps:

[0006] 1) Obtain wind speed and direction data in the wind turbine operating area in real time through sensors;

[0007] 2) Based on the turbulence model, the collected wind speed and direction data are corrected to eliminate the high-frequency disturbance caused by turbulence;

[0008] 3) Correct the wind speed and direction data based on the wind shear model to compensate for the wind speed gradient caused by height differences;

[0009] 4) Construct a wind field optimization model and dynamically optimize wind speed and direction data using a global optimization method;

[0010] 5) Based on the multi-scale decomposition method, the wind speed and direction data are integrated with the macro trend and local details to generate the final corrected wind speed and direction data;

[0011] 6) Output corrected high-precision wind speed and direction data.

[0012] Preferably, the sensor includes a laser radar and an ultrasonic sensor, and the wind speed and direction data are represented in the form of a three-dimensional velocity vector, including three directional components of the wind speed.

[0013] Preferably, in step 2), the turbulence correction includes the following process:

[0014] Construct a dynamic model of turbulence, and estimate the random disturbance of turbulence to wind speed and direction data by describing the characteristics of turbulent energy transfer and dissipation;

[0015] The turbulence closure model is used to calculate the turbulence disturbance components and eliminate these disturbances to correct the wind speed and direction data.

[0016] Preferably, the turbulence dynamics model uses the Reynolds average method to analyze turbulent disturbances, and the turbulence closed model eliminates the influence of turbulence on wind speed and direction data by calculating the relationship between turbulent kinetic energy generation and dissipation.

[0017] Preferably, in step 3), the wind shear model is used to describe the gradient change of wind speed at different heights, and the wind shear model dynamically adjusts the wind speed correction factor according to the wind speed difference in the area around the wind turbine to compensate for the wind speed deviation caused by the height difference.

[0018] Preferably, in step 4), the objectives of constructing the wind farm optimization model include the following:

[0019] Data fitting constraints ensure the consistency between wind field data and measured data;

[0020] Smoothness constraints ensure the global continuity of wind farm data;

[0021] Turbulence energy constraints limit the impact of turbulence disturbances on wind speed and direction data.

[0022] Preferably, in step 4), the wind field optimization model is solved by a global optimization method, and the obtained corrected wind speed and direction data satisfy the optimization results of data fitting constraints and smoothness constraints, and are dynamically adjusted in combination with boundary conditions.

[0023] Preferably, in step 5), the multi-scale decomposition and fusion are iteratively calculated by dynamically adjusting the fusion coefficients of the macro trend component and the local detail component, wherein:

[0024] The macro trend component is used to describe the overall distribution characteristics of the wind field;

[0025] The local detail component is used to capture the dynamic changing characteristics of turbulence and wind shear;

[0026] The fusion process adjusts the multi-scale decomposition results by dynamically updating the fusion coefficients to complete the multi-scale reconstruction of the wind field distribution.

[0027] Preferably, the fusion coefficient of the macro trend component is calculated based on the smoothness constraint of the spatial distribution of the global wind speed and direction in the wind field optimization model, and the fusion coefficient of the local detail component is calculated based on the sensitivity constraint of the turbulent disturbance and wind shear characteristics in the wind field optimization model, and the dynamic adjustment of the multi-scale decomposition results is completed through iterative calculation.

[0028] Preferably, in step 5), the multi-scale decomposition and dynamic fusion process is combined with the iterative calculation process of the wind field optimization model to adjust the fusion parameters of the macro trend component and the local detail component in real time, thereby performing multi-scale optimization of the dynamic distribution of wind speed and direction data.

[0029] The present invention provides a method for obtaining wind speed and direction of a wind turbine. It has the following beneficial effects:

[0030] 1. The present invention introduces turbulence correction and wind shear model to eliminate the main measurement errors caused by turbulent disturbance and height difference in view of the complex wind field characteristics in the wind turbine operation area. This layered correction method effectively reduces the randomness and deviation of wind speed and direction data, making the output data have high reliability and accuracy in a variety of wind field environments.

[0031] 2. The present invention adopts a multi-scale decomposition and fusion method to divide the wind speed and direction data into macro trend components and local detail components, and reconstructs the wind field data by dynamically adjusting the fusion coefficient. The macro trend ensures the continuity of the global wind field, and the local details capture the dynamic characteristics of the wind field, such as turbulence and wind shear. The fused data has both the integrity of the global perspective and the flexibility to adapt to the complex needs of local changes.

[0032] 3. By constructing a wind farm optimization model and combining it with a global optimization method, the present invention can dynamically adjust the wind speed and direction data so that it satisfies the data fitting constraints, global smoothness constraints, and turbulent energy constraints at the same time. This optimization method ensures the rationality of the wind speed and direction data in terms of spatial distribution, and at the same time has a strong dynamic adaptability in a variable wind farm environment, providing reliable data support for the efficient operation of wind turbines. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The figure is a flow chart of the method steps of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Please refer to the attached Figure 1 The embodiment of the present invention provides a method for obtaining wind speed and direction of a wind turbine, comprising the following steps:

[0036] 1) Obtain wind speed and direction data in the wind turbine operating area in real time through sensors;

[0037] 2) Based on the turbulence model, the collected wind speed and direction data are corrected to eliminate the high-frequency disturbance caused by turbulence;

[0038] 3) Correct the wind speed and direction data based on the wind shear model to compensate for the wind speed gradient caused by height differences;

[0039] 4) Construct a wind field optimization model and dynamically optimize wind speed and direction data using a global optimization method;

[0040] 5) Based on the multi-scale decomposition method, the wind speed and direction data are integrated with the macro trend and local details to generate the final corrected wind speed and direction data;

[0041] 6) Output corrected high-precision wind speed and direction data.

[0042] Specifically, first, wind speed and direction data are collected in real time through sensors, which is the basis of the entire method. The collected data includes three directional components of wind speed, expressed in the form of three-dimensional vectors. This design can fully reflect the dynamic characteristics of wind speed and direction in the operating area of ​​the wind turbine. The sensor uses lidar and ultrasonic sensors, and the combination of the two improves the data acquisition accuracy and spatial coverage. The collected raw data is affected by turbulent disturbances and wind shear effects. These data themselves are not suitable for direct use in subsequent analysis, so they need to be further corrected and processed.

[0043] After completing the turbulence and wind shear corrections, the data enters the wind field optimization stage. The construction of the wind field optimization model is based on the global optimization method, and the goals include data fitting, smoothness, and turbulence energy constraints. The optimization process not only ensures the consistency of the data with the actual measurement, but also improves the continuity of the global wind field distribution through smoothness constraints. This step not only makes local adjustments to the single-point data, but also corrects the distribution pattern of the wind field as a whole, providing a guarantee for the overall consistency and rationality of the data.

[0044] After that, the data enters the multi-scale decomposition and fusion stage. Through the multi-scale decomposition method, the wind speed and direction data are split into two components: macro trends and local details. The macro trend component describes the global distribution characteristics of the wind field, while the local detail component captures the dynamic changes of turbulent disturbances and wind shear. The two parts of the data are recombined by dynamically adjusting the fusion coefficient. In this process, the adjustment of the fusion coefficient is dynamically calculated based on the results of the wind field optimization model to ensure that the decomposed data can balance the global trend and local dynamic changes. This step increases the resolution of the data to multiple levels, making it more adaptable to complex wind field environments.

[0045] Finally, the corrected high-precision wind speed and direction data are output. These data have high precision, global continuity, and consistency of local dynamic changes, and can be directly used in application scenarios such as wind turbine operation monitoring and control strategy adjustment. The entire method, from data collection to processing and then to final output, forms an interlocking technical closed loop, and each step serves to improve data quality and lays the foundation for the next step of processing.

[0046] In the step 1), the sensor includes a laser radar and an ultrasonic sensor, and the wind speed and direction data are represented in the form of a three-dimensional velocity vector, including three directional components of the wind speed.

[0047] Specifically, two different types of devices, laser radar and ultrasonic sensor, are used for data collection in this embodiment. Laser radar is used to measure wind speed in long-distance areas, while ultrasonic sensors focus on close-range high-precision measurements. This combination of sensors can cover changes in wind speed and direction over a wide area while taking into account the ability to capture spatial details. The collected data is expressed as a three-dimensional velocity vector, specifically including the wind speed components in the x, y, and z directions:

[0048]

[0049] Among them, t represents time, and x, y, and z represent the coordinates of the measurement point in space.

[0050] As an option, the LiDAR can be installed on the downwind side of the wind turbine, usually at a certain distance from the turbine on the ground or near the cabin, to measure wind speed at multiple heights through vertical scanning. The ultrasonic sensor can be directly installed on the top of the cabin to collect wind speed and direction information in the local area around the turbine. This installation method can reduce signal interference between devices and ensure measurement stability.

[0051] Specifically, in some embodiments, the measurement range of the laser radar can cover the entire height range from the ground to the top of the blade, and its height resolution can be set to between 5 meters and 10 meters. For example, for a wind turbine with a height of 150 meters, the laser radar can collect wind speed and direction data in layers of no less than 15 layers. Each layer of data contains wind speed vector information on a horizontal plane, which is used to represent the horizontal and vertical components of the wind speed.

[0052] In another possible implementation, the measurement data of the ultrasonic sensor is mainly used to supplement the insufficient response of the lidar under high-frequency changes. The sampling frequency of the ultrasonic sensor can be as high as 10Hz, which can capture the rapid wind speed changes caused by turbulence. The measurement accuracy of this high-frequency data in the area around the wind turbine nacelle can reach within 0.1m / s.

[0053] Generally, the data from the LiDAR and ultrasonic sensors need to be spatially synchronized. The long-range data from the LiDAR is usually output in the form of fixed-height slices, while the data from the ultrasonic sensor is based on the 3D spatial coordinates of the cabin. The data from both are fused through a coordinate transformation method to obtain a unified 3D wind speed representation.

[0054] In some embodiments, in order to improve the real-time performance of data, sensors can also be arranged in a distributed manner. For example, laser radar and ultrasonic sensors are connected to the unit control center through optical fibers, and each sensor module completes the preliminary analysis and calibration of data through an independent processing unit. Then, the analyzed data is transmitted to the back-end centralized processing system through a communication interface.

[0055] In order to further improve the stability and anti-interference of the data, the sampling data of the sensor in this embodiment is also subjected to preliminary filtering processing. Specifically:

[0056] The data filtering adopts a Kalman filtering method, which dynamically adjusts the filter gain according to the covariance of the measured value and the estimated value to reduce the influence of the measurement noise on the data.

[0057] In another possible implementation, the multiple frames of data output by the sensor can be smoothed by a sliding window averaging method. The time span of the sliding window can be set to 1-5 seconds to balance the real-time performance and stability of the data.

[0058] In addition, in this embodiment, the collected wind speed and direction data will be stored in a time series format and sent to the subsequent modules in the form of three-dimensional vectors for turbulence correction, wind shear compensation and wind field optimization. The interface design between the acquisition module and the correction module ensures efficient transmission of the data stream and reduces the error accumulation caused by data delay.

[0059] In the step 2), the turbulence correction includes the following process:

[0060] Construct a dynamic model of turbulence, and estimate the random disturbance of turbulence to wind speed and direction data by describing the characteristics of turbulent energy transfer and dissipation;

[0061] Using the turbulence closed model, the turbulence disturbance components are calculated and these disturbances are eliminated to correct the wind speed and direction data;

[0062] The turbulence dynamics model uses the Reynolds average method to analyze turbulence disturbances, and the turbulence closed model eliminates the influence of turbulence on wind speed and direction data by calculating the relationship between turbulence kinetic energy generation and dissipation.

[0063] Specifically, in the operating area of ​​wind turbines, turbulence effects are usually one of the main factors leading to unstable wind speed and direction measurement data. The presence of turbulence will cause random disturbances in the wind speed and direction data, interfering with the acquisition of actual wind field characteristics. In general, the generation of turbulence is closely related to the energy transfer process, including various complex factors such as wind turbine blade wake and terrain. In the above step 1), the collected wind speed and direction data contain high-frequency components of turbulent disturbances, which need to be eliminated through turbulence correction. This step corrects the errors caused by turbulence in the collected data by constructing a turbulence dynamics model and combining it with the turbulence closure method, so as to provide more reliable input data for subsequent wind field optimization and multi-scale decomposition.

[0064] In this embodiment, the turbulence correction first decomposes the wind speed and direction data based on the Reynolds average method. Specifically, the measurement data can be divided into an average component and a turbulent disturbance component:

[0065]

[0066] Among them, u is the actually measured wind speed vector, represents the average component of the wind field, u ′ is the turbulence disturbance component. The average component reflects the overall trend of the wind field, while the turbulence disturbance component shows high-frequency changes.

[0067] As an option, in this embodiment, a turbulence model is used to model and correct the turbulent disturbance. The expression of turbulent kinetic energy k is:

[0068]

[0069] Among them, u ′2 、v ′2 、w ′2 are the components of turbulent disturbance in the x, y, and z directions respectively. The turbulent kinetic energy k represents the energy of turbulent disturbance.

[0070] In general, the dissipation rate of turbulent energy ε represents the rate at which turbulent energy is transferred to small scales and eventually converted into thermal energy. Its expression is:

[0071]

[0072] Among them, C μ is a constant parameter in the turbulence model, and l represents the characteristic length scale of turbulence.

[0073] Specifically, in this embodiment, turbulence correction is achieved by the following steps:

[0074] First, the turbulent kinetic energy k and the dissipation rate ε are calculated based on the measured data. The turbulent kinetic energy is calculated from the fluctuating component of the measured wind speed, while the dissipation rate is calculated using the turbulent closed equation.

[0075] Second, construct the Reynolds stress tensor of turbulence Its expression is:

[0076]

[0077] Among them, ν t represents the turbulent viscosity coefficient, S ij is the strain rate tensor of the average velocity, δ ij is the Kronecker symbol.

[0078] Finally, the high-frequency components in the wind speed and direction data are eliminated by using the calculated turbulence disturbance components, and the corrected wind speed vector is:

[0079] u 修正 =u 测量 -u′

[0080] In one possible implementation, the turbulent viscosity coefficient ν t Calculated by the following relationship:

[0081]

[0082] This expression shows that the turbulent viscosity coefficient is determined by the turbulent kinetic energy and dissipation rate, and is directly related to the strength of turbulence.

[0083] In step 3), the wind shear model is used to describe the gradient change of wind speed at different heights. The wind shear model dynamically adjusts the wind speed correction factor according to the wind speed difference in the area around the wind turbine to compensate for the wind speed deviation caused by the height difference.

[0084] Specifically, in this embodiment, the wind shear correction uses a wind shear model to describe the wind speed gradient changes at different heights, and dynamically adjusts the correction factor to adapt to complex wind field characteristics. Specifically, the basic expression of the wind shear model is:

[0085]

[0086] Where V(z) is the wind speed at height z; V ref is the reference height z ref α is the wind shear index, which is used to reflect the nonlinear characteristics of wind speed changing with height.

[0087] In general, the reference height z ref The hub height of the wind turbine can be selected. As an option, the reference height can also be dynamically determined based on the measurement conditions of the wind farm area, such as selecting the middle layer height of the lidar measurement range.

[0088] In this embodiment, the wind shear index α is not a fixed value, but is dynamically adjusted according to the actual characteristics of the wind field. Specifically, the calculation of the wind shear index is related to the turbulence parameter, and its expression is:

[0089]

[0090] Where k is the turbulent kinetic energy, k ref is the reference turbulent kinetic energy. This formula shows that when the turbulent kinetic energy is large, the change amplitude of the wind shear index will gradually decrease, thus better adapting to the height distribution characteristics in a strong turbulent environment.

[0091] As a possible implementation method, in the wind shear correction step, the wind speed data at different heights are first calculated to obtain the height correction factor. Specifically, the correction factor can be calculated by the following formula:

[0092]

[0093] Among them, Δα represents the dynamic change of wind shear index. Then, the corrected wind speed can be expressed as:

[0094] V 修正 (z)=V(z)·F 修正 (z)

[0095] In some embodiments, in order to improve the adaptability of wind shear correction, multi-layer data of the lidar can be combined to achieve more accurate height compensation by calculating the wind shear index and correction factor in layers.

[0096] In some special scenarios, such as offshore wind farms, the spatial characteristics of wind shear may be significantly affected by humidity and temperature difference. This embodiment can expand the wind shear model by introducing additional correction factors (such as temperature difference coefficient and humidity correction coefficient). Specifically, the temperature difference correction coefficient and humidity correction coefficient can be calculated by the following formula:

[0097]

[0098]

[0099] Among them, ΔT represents the temperature difference change, T ref is the reference temperature; ΔH represents the humidity change, H ref For reference humidity.

[0100] In step 4), the objectives of constructing a wind farm optimization model include the following:

[0101] Data fitting constraints ensure the consistency between wind field data and measured data;

[0102] Smoothness constraints ensure the global continuity of wind farm data;

[0103] Turbulence energy constraints limit the impact of turbulence disturbances on wind speed and direction data;

[0104] In step 4), the wind field optimization model is solved by a global optimization method, and the obtained modified wind speed and direction data meet the optimization results of data fitting constraints and smoothness constraints, and are dynamically adjusted in combination with boundary conditions.

[0105] Specifically, in this embodiment, the objective function design of wind farm optimization is centered on minimizing the overall error of wind speed and direction data, while taking into account the global smoothness constraint and turbulent disturbance energy limit. The mathematical expression of the objective function is:

[0106]

[0107] Where u represents the optimized wind speed and direction vector; u 测量 is the corrected measurement data; Ω is the spatial domain of the wind field; represents the gradient distribution of the wind field; λ and μ are the smoothness weight coefficient and turbulence energy weight coefficient.

[0108] Item 1 Ensure that the optimized wind speed and direction data have a high degree of fit with the corrected measurement data;

[0109] Item 2 Constrain the global smoothness of the wind field and reduce the spatial mutation of wind speed and direction data;

[0110] The third term μ∥u ′ ∥ 2 : Limit the energy of turbulent disturbance to ensure that the optimized wind speed and direction data are physically reasonable.

[0111] In this embodiment, the constraint conditions of the wind farm optimization model mainly include the boundary conditions and physical continuity constraints of the wind farm.

[0112] Boundary conditions:

[0113] During the optimization process, the wind speed and direction data need to meet the no-flux condition at the boundary of the wind field.

[0114] In general, the values ​​of weight coefficients λ and μ need to be adjusted according to the characteristics of the actual wind field. For example, for scenes with high turbulence intensity, the weight of μ can be appropriately increased to enhance the turbulence suppression effect.

[0115] In step 5), the multi-scale decomposition and fusion are iteratively calculated by dynamically adjusting the fusion coefficients of the macro trend component and the local detail component, wherein:

[0116] The macro trend component is used to describe the overall distribution characteristics of the wind field;

[0117] The local detail component is used to capture the dynamic changing characteristics of turbulence and wind shear;

[0118] The fusion process adjusts the multi-scale decomposition results by dynamically updating the fusion coefficients to complete the multi-scale reconstruction of the wind field distribution;

[0119] The fusion coefficient of the macro trend component is calculated based on the spatial distribution smoothness constraint of the global wind speed and direction in the wind field optimization model, and the fusion coefficient of the local detail component is calculated based on the sensitivity constraint of the turbulent disturbance and wind shear characteristics in the wind field optimization model. The dynamic adjustment of the multi-scale decomposition results is completed through iterative calculation;

[0120] In step 5), the multi-scale decomposition and dynamic fusion process is combined with the iterative calculation process of the wind field optimization model to adjust the fusion parameters of the macro trend component and the local detail component in real time, thereby performing multi-scale optimization of the dynamic distribution of wind speed and direction data.

[0121] Specifically, in this embodiment, the multi-scale decomposition method is to decompose the wind speed and direction data into a macro trend component and a local detail component. The macro trend component is used to describe the overall change trend of the wind field, and the local detail component focuses on capturing the local dynamic changes of turbulent disturbances and wind shear. The mathematical form of the decomposition can be expressed as:

[0122] u(x,y,z,t)=u 趋势 (x,y,z,t)+u 细节 (x,y,z,t)

[0123] Among them, u 趋势 is the macro trend component, u 细节 is the local detail component.

[0124] In the specific implementation, this embodiment uses wavelet transform for multi-scale decomposition. Wavelet transform can simultaneously obtain the global low-frequency components and local high-frequency components of wind speed and direction data by decomposing the time domain and frequency domain characteristics of the data. The specific decomposition formula is:

[0125]

[0126] u 细节 =uu 趋势

[0127] Among them, c k is the wavelet decomposition coefficient, φ k is the wavelet basis function.

[0128] As an option, the basis function may use Daubechies wavelet to describe the changing characteristics of wind speed and direction. In another possible implementation, fast Fourier transform (FFT) may also be used to decompose the high-frequency and low-frequency components of wind speed and direction data, where the low-frequency component represents the trend characteristics and the high-frequency component corresponds to the local disturbance.

[0129] After decomposing the macro trend component and the local detail component, in order to achieve effective fusion of the two, this embodiment introduces a dynamic weight adjustment mechanism. The mathematical expression of the fused wind speed and direction data is:

[0130] u 融合 (x,y,z,t)=w T ·u 趋势 (x,y,z,t)+w D ·u 细节 (x,y,z,t)

[0131] Among them, w T and w D Represent the weights of trend component and detail component respectively, satisfying w T +w D =1.

[0132] In general, the weight w T and w D The dynamic adjustment of is based on the global smoothness parameter and local dynamic sensitivity parameter of the wind farm. Specifically, the weight calculation formula is:

[0133] w T (t) = w T (1+γsin(ωt))

[0134] w D =1-w T

[0135] Among them, γ is the adjustment parameter and γ represents the global smoothness critical value.

[0136] Local dynamic sensitivity parameter S 局部 The calculation formula is:

[0137]

[0138] In this embodiment, multi-scale decomposition and fusion are implemented using an iterative optimization method. The steps of each iteration include:

[0139] Calculate the fused wind speed and direction data using the current trend component and detail component;

[0140] w T and w D ;

[0141] Determine whether the objective function meets the convergence condition.

[0142] The mathematical form of the objective function is:

[0143]

[0144] Among them, β is the smoothness weight and Ω represents the wind field spatial domain.

[0145] Generally, when the value of the objective function E is less than the preset convergence threshold, the iteration process is terminated and the fused wind speed and direction data are finally output.

[0146] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for obtaining wind speed and direction of a wind turbine, characterized in that: The following steps are involved: 1) Obtain wind speed and direction data in the wind turbine operating area in real time through sensors; 2) Based on the turbulence model, the collected wind speed and direction data are corrected to eliminate the high-frequency disturbance caused by turbulence; 3) Correct the wind speed and direction data based on the wind shear model to compensate for the wind speed gradient caused by height differences; 4) Construct a wind field optimization model and dynamically optimize wind speed and direction data using a global optimization method; 5) Based on the multi-scale decomposition method, the wind speed and direction data are integrated with the macro trend and local details to generate the final corrected wind speed and direction data; 6) Output corrected high-precision wind speed and direction data.

2. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In the step 1), the sensor includes a laser radar and an ultrasonic sensor, and the wind speed and direction data are represented in the form of a three-dimensional velocity vector, including three directional components of the wind speed.

3. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In the step 2), the turbulence correction includes the following process: Construct a dynamic model of turbulence, and estimate the random disturbance of turbulence to wind speed and direction data by describing the characteristics of turbulent energy transfer and dissipation; The turbulence closure model is used to calculate the turbulence disturbance components and eliminate these disturbances to correct the wind speed and direction data.

4. A method for obtaining wind speed and direction of a wind turbine according to claim 3, characterized in that: The turbulence dynamics model uses the Reynolds average method to analyze turbulence disturbances, and the turbulence closed model eliminates the influence of turbulence on wind speed and direction data by calculating the relationship between turbulence kinetic energy generation and dissipation.

5. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In step 3), the wind shear model is used to describe the gradient change of wind speed at different heights. The wind shear model dynamically adjusts the wind speed correction factor according to the wind speed difference in the area around the wind turbine to compensate for the wind speed deviation caused by the height difference.

6. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In step 4), the objectives of constructing a wind farm optimization model include the following: Data fitting constraints ensure the consistency between wind field data and measured data; Smoothness constraints ensure the global continuity of wind farm data; Turbulence energy constraints limit the impact of turbulence disturbances on wind speed and direction data.

7. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In step 4), the wind field optimization model is solved by a global optimization method, and the obtained modified wind speed and direction data meet the optimization results of data fitting constraints and smoothness constraints, and are dynamically adjusted in combination with boundary conditions.

8. A method for obtaining wind speed and direction of a wind turbine according to claim 1, characterized in that: In step 5), the multi-scale decomposition and fusion are iteratively calculated by dynamically adjusting the fusion coefficients of the macro trend component and the local detail component, wherein: The macro trend component is used to describe the overall distribution characteristics of the wind field; The local detail component is used to capture the dynamic changing characteristics of turbulence and wind shear; The fusion process adjusts the multi-scale decomposition results by dynamically updating the fusion coefficients to complete the multi-scale reconstruction of the wind field distribution.

9. A method for obtaining wind speed and direction of a wind turbine according to claim 8, characterized in that: The fusion coefficient of the macro trend component is calculated according to the smoothness constraint of the spatial distribution of the global wind speed and direction in the wind field optimization model, and the fusion coefficient of the local detail component is calculated according to the sensitivity constraint of the turbulent disturbance and wind shear characteristics in the wind field optimization model. The dynamic adjustment of the multi-scale decomposition results is completed through iterative calculation.

10. A method for acquiring wind speed and direction of a wind turbine according to claim 1, characterized in that: In step 5), the multi-scale decomposition and dynamic fusion process is combined with the iterative calculation process of the wind field optimization model to adjust the fusion parameters of the macro trend component and the local detail component in real time, thereby performing multi-scale optimization of the dynamic distribution of wind speed and direction data.

Images