Method, system, device and medium for calculating wind turbine wake field velocity distribution

Abstract

This disclosure pertains to the field of wind power technology and provides a method, system, equipment, and medium for calculating the velocity distribution of a wind turbine wake field. The method includes: acquiring wind speed, rotor speed, and airfoil aerodynamic data; constructing a prediction function for the wind turbine based on blade deformation data to output load distribution; inputting current blade deformation data to obtain the load distribution; calculating the blade spanwise circulation distribution based on relevant data; obtaining the intensity of wake vortices and shedding vortices based on the circulation distribution, thereby obtaining the wake control point position vector; and solving for the velocity distribution at any location within the downstream wake field of the wind turbine. This process aims to more accurately predict the relevant characteristics of the wind turbine wake field and solve the problem that existing wake field prediction models cannot consider the true shape of the wake. This method can consider the true shape of the wake to obtain an accurate wake field velocity distribution.

Description

Technical Field

[0001] This disclosure pertains to the field of wind power technology, and particularly relates to methods, systems, equipment, and media for calculating the velocity distribution of the wake field of wind turbines. Background Technology

[0002] When a wind turbine is running, it generates a wake field behind it. The presence of this wake field directly affects the surrounding environment and the operation of other wind turbines in the wind farm. Predicting the wake distribution and related characteristics of wind turbines can help optimize the layout and design of wind turbines, and improve the power generation efficiency and stability of the wind farm.

[0003] Patent CN114091377A discloses a method for calculating the wake velocity of a dynamic dual-Gaussian wind turbine based on spatial variation. The method involves determining the turbine rotor diameter and thrust coefficient, calculating the initial wake radius, collecting the incoming wind speed, determining the initial wake expansion coefficient, power exponent, and radial distance from the wake velocity minimum point relative to the hub centerline, and calculating the effective rotor diameter. The wake region is divided into a wake region adjacent to the rotor, a near-wake region, and a far-wake region, and the corresponding wake expansion coefficient and power exponent are determined. Combined with the initial wake radius, the wake radius distribution is calculated. A single-sided Gaussian profile distribution function of the wake is established on both sides of the hub centerline, and the dual-Gaussian distribution profile of the wake is calculated. The average momentum-flow equation and thrust calculation formula are combined to calculate the maximum normalized velocity attenuation, and then the velocity attenuation in the wake region is calculated to determine the velocity distribution in the wake region. This technology improves the accuracy of calculating the wake velocity distribution throughout the entire wake region.

[0004] Chinese patent application CN116415421A discloses a method and apparatus for modeling the flow field in the wake region of a wind turbine. The method includes: determining the radius of the wind turbine wake region based on the basic parameters required for flow field modeling and the assumption of linear wake expansion; determining the location of the maximum velocity deficit in the wake region based on the actual evolution law of the wake; dividing the wind turbine wake region into regions based on the wake radius, the location of the maximum velocity deficit, and the wake influence of the wind turbine blades; performing regional modeling of the flow field in the divided wind turbine wake region to establish a flow field model to be solved in the wind turbine wake region; and solving the flow field model to be solved in the wind turbine wake region to obtain the flow field model. This method and apparatus for flow field modeling considers the actual generation and evolution law of the wind turbine wake, accurately models the flow field in the wind turbine wake region, and improves the accuracy of wake calculation.

[0005] Chinese patent application CN109255184B discloses a method and system for determining the wind speed distribution of the entire wake field of a wind turbine. The method includes: determining the near-wake wind speed distribution of the wind turbine based on a vortex wake model; determining the far-wake wind speed distribution of the wind turbine based on an engineering wake model; determining the wind speeds at 2D and 3D positions based on the near-wake and far-wake wind speed distributions, respectively; correcting the wake radius based on the wind speeds at the 2D and 3D positions and a correction formula; further correcting the near-wake and far-wake wind speed distributions based on the corrected wake radius to obtain corrected near-wake and far-wake wind speed distributions; and determining the wind speed distribution of the entire wake field of the wind turbine based on the corrected near-wake and far-wake wind speed distributions. This method and system for determining the wind speed distribution of the entire wake field of a wind turbine can fit the wind speed distribution of the entire wake field, not only meeting engineering requirements in terms of accuracy but also significantly reducing the computational load.

[0006] However, in existing technologies, the prediction of the wake field is based on the empirical wake model, which only focuses on the macroscopic velocity characteristics of the wake field and cannot take into account the true shape of the wake. Summary of the Invention

[0007] To address the aforementioned issues, this disclosure provides a method, system, equipment, and medium for calculating the velocity distribution of wind turbine wake fields. By employing an improved technique for determining the induced velocity in wind turbine wake development prediction, it is possible to resolve the problem that existing wake field prediction models cannot consider the true shape of the wake.

[0008] This invention is achieved through the following technical solution:

[0009] Calculation method for velocity distribution in the wake field of a wind turbine:

[0010] Acquire average incoming wind speed data, wind turbine speed, and airfoil family aerodynamic data;

[0011] For a given type of wind turbine in a wind farm, construct a blade load prediction function that outputs the blade load distribution based on blade deformation data; input the blade deformation data of the wind turbine blade at the current moment into the blade load prediction function, and output the blade load distribution at the current moment.

[0012] The spanwise circulation distribution of the wind turbine blades at the current moment is calculated based on the average incoming wind speed data, rotor speed, airfoil family aerodynamic data, and blade load distribution data at the current moment.

[0013] The intensity of the wake vortex and shedding vortex of the wind turbine blade is obtained based on the current spanwise circulation distribution of the blade, and the position vector of the wake control point is obtained based on the intensity of the wake vortex and shedding vortex.

[0014] The velocity distribution at any spatial location within the downstream wake field of the wind turbine is solved based on the wake control point position vector.

[0015] Furthermore,

[0016] The construction of a blade load prediction function based on blade deformation data and outputting blade load distribution includes:

[0017] Blade deformation data and corresponding blade load distribution data are generated using a finite element structural model and used as a machine learning training dataset.

[0018] A neural network is trained using a machine learning training dataset to obtain a blade load prediction function; the blade load prediction function is:

[0019]

[0020] In the formula, Let be the lift force at the j-th airfoil section of the blade; f is the blade load prediction function obtained from training. , , These are blade flapping, oscillation, and torsional deformation, respectively.

[0021] Furthermore,

[0022] The blade deformation data includes:

[0023] The flapping of the blades, the oscillation of the blades, and the torsional deformation of the blades.

[0024] Furthermore,

[0025] The formula for calculating the spanwise circulation distribution of the blade is:

[0026]

[0027] In the formula, It is air density; It is the length of the airfoil section; and These are the normal and tangential direction vectors at the airfoil section, respectively; Relative inflow velocity at the airfoil section; is the lift at the airfoil section; j is the airfoil section number; Based on the synthesis of incoming wind speed and impeller speed.

[0028] Furthermore,

[0029] The formulas for calculating the intensity of the blade's wake vortex and shedding vortex are as follows:

[0030]

[0031]

[0032] In the formula, and These represent the intensity of the wake vortex and the shedding vortex, respectively; the superscript i indicates the time; and the subscript j indicates the airfoil section number.

[0033] Furthermore,

[0034] The formula for calculating the position vector of the wake control point is:

[0035]

[0036] In the formula, The position vector of the control points of the wake; For free flow velocity; The average induced velocity at the wind turbine plane; The induced velocity decay coefficient, Lifetime angle with trail change;

[0037] Among them, the induced velocity decay coefficient The induced velocity decay coefficient at the wake initiation position varies with the lifetime angle. Set to 1, the induced velocity decay coefficient at the wake termination position. Taking a value of 0.5, the induced velocity decay coefficient is calculated by lifetime angle interpolation at the midpoint of the wake. .

[0038] Furthermore,

[0039] The velocity distribution at any spatial location within the downstream wake field of a wind turbine can be calculated using Biot-Savart's law. The expression is as follows:

[0040]

[0041] In the formula, , Let be the position vector of the control point of the wake from the vortex segment to the spatial point. l denoted as vortex segment length.

[0042] A system for calculating the velocity distribution of the wake field of a wind turbine, characterized in that it includes:

[0043] The data acquisition module is used to acquire average incoming wind speed data, wind turbine rotation speed, and airfoil family aerodynamic data;

[0044] The blade load distribution output module is used to construct a blade load prediction function based on blade deformation data for a given type of wind turbine in a wind farm; the blade deformation data of the wind turbine blade at the current moment is input into the blade load prediction function, and the blade load distribution at the current moment is output.

[0045] The blade spanwise circulation distribution calculation module is used to calculate the current moment's wind turbine blade spanwise circulation distribution based on the average incoming wind speed data, rotor speed, airfoil family aerodynamic data, and the current moment's blade load distribution data.

[0046] The wake spatial distribution construction module is used to obtain the wake vortex and shedding vortex intensity of the wind turbine blade based on the current moment's spanwise circulation distribution, and to obtain the wake control point position vector based on the wake vortex and shedding vortex intensity.

[0047] The wake field distribution construction module is used to solve the velocity distribution at any spatial location in the wake field downstream of the wind turbine based on the wake control point position vector.

[0048] Compared with the prior art, this disclosure has the following advantages:

[0049] This invention establishes a blade load prediction function based on blade deformation data to output blade load distribution, which can accurately infer the blade load distribution based on real-time monitored blade geometric deformation; blade load directly affects the formation and development of wind turbine wake;

[0050] Based on accurate blade load distribution, combined with other relevant data such as average incoming wind speed, rotor speed, and airfoil family aerodynamic data, the spanwise circulation distribution of wind turbine blades can be estimated more accurately. Furthermore, the intensity of the blade wake vortex and shedding vortex can be obtained from the circulation distribution. The intensity of the wake vortex and shedding vortex determines the dynamic development and spatial distribution of the rotor wake, thus enabling accurate calculation of the characteristics of the rotor wake. Based on the accurately obtained wake control point position vector, the velocity distribution at any spatial location within the downstream wake field of the wind turbine can be further calculated, achieving real-time prediction of the wind turbine wake geometry and wake field velocity distribution.

[0051] Such real-time prediction can comprehensively and accurately describe the shape of the wake and the characteristics of the wake field; it solves the problem that existing technologies cannot take into account the true shape of the wake.

[0052] Other features and advantages of this disclosure will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the disclosure. The objects and other advantages of this disclosure may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 A schematic diagram of the method of the present invention is shown. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0056] Figure 1 A schematic diagram of the method according to the present invention is shown. Specific implementation details of the present invention include:

[0057] Step 1: Establish a machine learning training dataset of "blade deformation data - blade load distribution" for a given type of wind turbine in the wind field, and train the blade load prediction function.

[0058] 1.1) In step one, the machine learning training dataset of "blade deformation data - blade load distribution" is generated through a finite element structural model or a multibody dynamics model of a wind turbine blade. The specific steps are as follows:

[0059] Blade geometry modeling: Using professional CAD software, accurately draw the three-dimensional geometry of a given wind turbine blade, including the blade's airfoil, length, width, thickness, and internal structural details.

[0060] Finite element mesh generation: Import the geometric model of the blade into the finite element analysis software.

[0061] Choose an appropriate mesh type (such as tetrahedral or hexahedral mesh) and mesh the blade according to its structural characteristics and required analytical accuracy. Refine the mesh in critical areas (such as the blade root and areas with significant airfoil variations) to improve computational accuracy.

[0062] Material property definition: Assigning corresponding material properties to each part of the blade, such as elastic modulus, Poisson's ratio, density, etc. These properties should be determined based on the actual material used in the blade.

[0063] Boundary conditions and load settings: Define the boundary conditions for the blades, usually by fixing the blade root to simulate their installation on a wind turbine.

[0064] Various types and sizes of loads are applied, including aerodynamic loads (calculated based on different wind speeds, wind directions, and turbulence intensities), centrifugal force (considering the rotational speed of the blades), and gravity.

[0065] Finite element analysis: Run finite element analysis to obtain deformation data (such as displacement and strain) and load distribution data (such as bending moment, torque, and shear force) of the blade under different load conditions.

[0066] Extract the required blade deformation data and load distribution data from the finite element analysis results. The blade deformation data and load distribution data are in a one-to-one correspondence.

[0067] The data is labeled to clarify the load conditions and deformation results corresponding to each data sample.

[0068] The labeled data is the machine learning training dataset.

[0069] 1.2) In step one, the specific steps for training the blade load prediction function are as follows:

[0070] A neural network is trained using a machine learning training dataset of blade deformation data and blade load distribution to construct a blade load prediction function. A backpropagation (bp) neural network can be used as the neural network.

[0071] The blade load prediction function has the following functional form:

[0072]

[0073] In the formula, Let f be the lift force at the j-th airfoil section of the blade; f is the black-box function obtained through training. , , These are blade flapping, oscillation, and torsional deformation, respectively.

[0074] The input to the blade load prediction function is blade deformation data, including blade flapping, oscillation and torsional deformation, and the output of the blade load prediction function is the blade spanwise lift distribution.

[0075] Step 2: Obtain the geometric deformation of the blade at the current moment based on the blade deformation data monitoring device, and inversely deduce the blade load distribution at the current moment through the blade load prediction function.

[0076] Step 3: Estimate the spanwise circulation distribution of the wind turbine blades at the current moment based on the blade load distribution at the current moment, combined with the average incoming wind speed data, rotor speed, and airfoil family aerodynamic data monitored by the SCADA system.

[0077] The blade spanwise circulation distribution is calculated using the following formula:

[0078]

[0079] In the formula, It is air density; It is the length of the airfoil section; and These are the normal and tangential direction vectors at the airfoil section, respectively; Relative inflow velocity at the airfoil section; This represents the lift at the airfoil section, and j is the airfoil segment number. The blade is divided into several segments along its length, for example, 20 segments would be numbered 20.

[0080] In the formula The acquisition requires the synthesis of incoming wind speed and rotation speed (conventional vector synthesis), and the wind speed and rotor speed are provided by the SCADA system.

[0081] Step 4: Obtain the wake vortex and shedding vortex intensity of each blade based on the circulation distribution of each blade at the current moment, and solve the dynamic development and spatial distribution of the wind turbine wake based on this.

[0082] 4.1) In step four, the specific formulas for calculating the intensity of the wake vortex and shedding vortex based on the blade spanwise circulation distribution are as follows:

[0083] The wake vortex intensity is the circulation of the (j+1)th vortex segment on the blade minus the circulation of the jth vortex segment on the blade.

[0084] The shedding vortex intensity is the circulation of the vortex segment on the blade at time i+1 minus the circulation of the vortex segment on the blade at time i.

[0085] Its mathematical expression is:

[0086]

[0087]

[0088] In the formula, and These represent the intensity of the wake vortex and the shedding vortex, respectively; the superscript i indicates the time; and the subscript j indicates the airfoil section number.

[0089] 4.2) Step four, solving for the dynamic development and spatial distribution of the wind turbine wake, specifically includes:

[0090] Traditional CFD numerical simulations and eddy wake models can predict the development of wind turbine wakes, but they are slow to solve and require huge computational resources, making them unsuitable for wake control in wind farms.

[0091] Therefore, in order to achieve a rapid solution for the wind turbine wake distribution to meet the control requirements of the wind farm, the solution for the wake geometry is accelerated by setting a wake-induced velocity attenuation coefficient k instead of the Biot-Savart law in the traditional numerical solution scheme:

[0092]

[0093] In the formula The position vector of the control points of the wake; For free flow velocity; The average induced velocity at the wind turbine plane; This is the induced velocity decay coefficient, which varies with the wake lifetime angle. change.

[0094] Among them, the induced velocity decay coefficient The induced velocity decay coefficient varies with the lifetime angle. It is set to 1 at the start of the wake, 0.5 at the end, and calculated via lifetime angle interpolation at the middle of the wake. This method allows for faster determination of the induced velocity, avoids complex calculations, significantly reduces computational load, and thus substantially lowers the computational resources required for prediction.

[0095] The interpolation calculation is a basic algorithm, which will not be described in detail here.

[0096] Step 5: Solve for the velocity distribution at any spatial location in the wake field downstream of the wind turbine based on the wake control point position vector.

[0097] The velocity distribution at any spatial location within the downstream wake field of the wind turbine is estimated using the Biot-Savart law:

[0098]

[0099] In the formula, , This is the vector from the vortex segment to a point in space. l denoted as vortex segment length.

[0100] Based on the method of the present invention, this disclosure also provides a prediction system corresponding to the above method, which includes:

[0101] The system comprises the following modules: a data acquisition module for acquiring average incoming wind speed, rotor speed, and airfoil family aerodynamic data; a blade load distribution output module for constructing a blade load prediction function based on blade deformation data for a given type of wind turbine in a wind field; and a module that outputs the blade load distribution at the current moment by inputting the blade deformation data of the wind turbine blades into the blade load prediction function. A blade spanwise circulation distribution calculation module is used to calculate the current moment's wind turbine blade spanwise circulation distribution based on average incoming wind speed, rotor speed, airfoil family aerodynamic data, and the current moment's blade load distribution data. A wake space distribution construction module is used to obtain the wake vortex and shedding vortex intensity of the blades based on the current moment's wind turbine blade spanwise circulation distribution, and to obtain the wake control point position vector. A wake field distribution construction module is used to solve for the velocity distribution at any spatial location within the wake field downstream of the wind turbine based on the wake control point position vector.

[0102] Based on the same inventive concept as disclosed above, embodiments of this disclosure also provide a prediction device corresponding to the above method, which includes at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the above method.

[0103] It should be noted that the electrical connection between the above-mentioned units does not necessarily mean the connection between lines. The indirect connection method can be applied to the embodiments of this disclosure as long as it achieves the purpose of this disclosure.

[0104] Based on the same inventive concept, this disclosure also provides a computer storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the above-described method.

[0105] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for calculating the velocity distribution of the wake field of a wind turbine, characterized in that, include: Acquire average incoming wind speed data, wind turbine speed, and airfoil family aerodynamic data; For a given type of wind turbine in a wind farm, construct a blade load prediction function that outputs the blade load distribution based on blade deformation data; input the blade deformation data of the wind turbine blade at the current moment into the blade load prediction function, and output the blade load distribution at the current moment. The spanwise circulation distribution of the wind turbine blades at the current moment is calculated based on the average incoming wind speed data, rotor speed, airfoil family aerodynamic data, and blade load distribution data at the current moment. The intensity of the wake vortex and shedding vortex of the wind turbine blade is obtained based on the current spanwise circulation distribution of the blade, and the position vector of the wake control point is obtained based on the intensity of the wake vortex and shedding vortex. The velocity distribution at any spatial location within the downstream wake field of the wind turbine is solved based on the wake control point position vector.

2. The method for calculating the velocity distribution of the wind turbine wake field according to claim 1, characterized in that, The construction of a blade load prediction function based on blade deformation data and outputting blade load distribution includes: Blade deformation data and corresponding blade load distribution data are generated using a finite element structural model and used as a machine learning training dataset. A neural network is trained using a machine learning training dataset to obtain a blade load prediction function; the blade load prediction function is: In the formula, Let be the lift force at the j-th airfoil section of the blade; f is the blade load prediction function obtained from training. , , These are blade flapping, oscillation, and torsional deformation, respectively.

3. The method for calculating the velocity distribution of the wind turbine wake field according to claim 1, characterized in that, The blade deformation data includes: The flapping of the blades, the oscillation of the blades, and the torsional deformation of the blades.

4. The method for calculating the velocity distribution of the wind turbine wake field according to claim 1, characterized in that, The formula for calculating the spanwise circulation distribution of the blade is: In the formula, It is air density; It is the length of the airfoil section; and These are the normal and tangential direction vectors at the airfoil section, respectively; Relative inflow velocity at the airfoil section; is the lift at the airfoil section; j is the airfoil section number; Based on the synthesis of incoming wind speed and impeller speed.

5. The method for calculating the velocity distribution of the wind turbine wake field according to claim 4, characterized in that, The formulas for calculating the intensity of the blade's wake vortex and shedding vortex are as follows: In the formula, and These represent the intensity of the wake vortex and the shedding vortex, respectively; the superscript i indicates the time; and the subscript j indicates the airfoil section number.

6. The method for calculating the velocity distribution of the wind turbine wake field according to claim 1, characterized in that, The formula for calculating the position vector of the wake control point is: In the formula, The position vector of the control points of the wake; For free flow velocity; The average induced velocity at the wind turbine plane; The induced velocity decay coefficient, Lifetime angle with trail change; Among them, the induced velocity decay coefficient The induced velocity decay coefficient at the wake initiation position varies with the lifetime angle. Set to 1, the induced velocity decay coefficient at the wake termination position. Taking a value of 0.5, the induced velocity decay coefficient is calculated by lifetime angle interpolation at the midpoint of the wake. .

7. The method for calculating the velocity distribution of the wind turbine wake field according to claim 1, characterized in that, The velocity distribution at any spatial location within the downstream wake field of a wind turbine can be calculated using Biot-Savart's law. The expression is as follows: In the formula, , Let be the position vector of the control point of the wake from the vortex segment to the spatial point. l The length of the vortex segment.

8. A calculation system for the velocity distribution of the wake field of a wind turbine, characterized in that, include: The data acquisition module is used to acquire average incoming wind speed data, wind turbine rotation speed, and airfoil family aerodynamic data; The blade load distribution output module is used to construct a blade load prediction function based on blade deformation data for a given type of wind turbine in a wind farm; the blade deformation data of the wind turbine blade at the current moment is input into the blade load prediction function, and the blade load distribution at the current moment is output. The blade spanwise circulation distribution calculation module is used to calculate the current moment's wind turbine blade spanwise circulation distribution based on the average incoming wind speed data, rotor speed, airfoil family aerodynamic data, and the current moment's blade load distribution data. The wake spatial distribution construction module is used to obtain the wake vortex and shedding vortex intensity of the wind turbine blade based on the current moment's spanwise circulation distribution, and to obtain the wake control point position vector based on the wake vortex and shedding vortex intensity. The wake field distribution construction module is used to solve the velocity distribution at any spatial location in the wake field downstream of the wind turbine based on the wake control point position vector.

9. A calculation device for the velocity distribution of the wake field of a wind turbine, comprising: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method for constructing the wind turbine wake field distribution as described in any one of claims 1 to 7.

10. A computer storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 7.

Images