A wind turbine power prediction method and device based on a wake simulation model

Abstract

The application discloses a wind turbine power prediction method and device based on a wake simulation model, and the method comprises the following steps: determining the position coordinates of all observation points on a target wind turbine according to a preset position coefficient in a constructed wake simulation model; taking each observation point as a first target observation point, determining the observation point closest to each first target observation point in a predetermined search range by using a KNN search algorithm to obtain each second target observation point, calculating the target distance between each first target observation point and the corresponding second target observation point, and calculating the wind speed attenuation factor of each observation point; performing effective wind speed prediction on the upper limit of the height of a target wind farm and the wind speed attenuation factor, performing weighted calculation on the effective wind speed of each observation point, performing power calculation based on the effective wind speed of the target wind turbine and the acquired yaw angle of the target wind turbine, and obtaining the predicted power of the target wind turbine. The embodiment can shorten the simulation time and improve the wind turbine power prediction efficiency.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, and in particular to a method and apparatus for predicting wind turbine power based on a wake simulation model. Background Technology

[0002] In recent years, wind power generation has experienced rapid development as a clean energy source. Existing technologies typically use wake simulation models to predict wind turbine power, such as the FLORIDyn wake model, which supports three-dimensional dynamic simulation. To incorporate dynamic calculations, the FLORIDyn model creates observation points (OPs) on the turbine surface in each simulation step. The role of these observation points is to describe the local Gaussian FLORIS wake characteristics at their location. OPs inherit the turbine's state at the time of creation, such as yaw angle, turbulence intensity, and thrust coefficient, which are used to calculate wind speed in the Gaussian FLORIS model. Over time, each observation point moves downstream, representing a large amount of air moving in the wind. Their movement paths are determined by the boundaries of the FLORIS wake region. Their movement speed is equal to the effective wind speed they represent.

[0003] However, the simulation computation time of the FLORIDyn model increases superlinearly with the number of wind turbines. Simultaneously, as the scale of the wind farm increases, the time required for wake propagation also increases, leading to a sharp increase in simulation time. With a large number of wind turbines, the computational cost becomes excessive, making it unsuitable for optimal control. This results in existing wake models having high computational costs, long simulation times, and low efficiency in wind turbine power prediction when the number of wind turbines is large. Summary of the Invention

[0004] This invention provides a method and apparatus for predicting wind turbine power based on a wake simulation model, which can effectively predict the target wind turbine power, shorten the simulation time, and improve the wind turbine power prediction efficiency.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a wind turbine power prediction method based on a wake simulation model, comprising:

[0006] In the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to the preset position coefficients;

[0007] Using each observation point as the first target observation point, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. The target distance between each first target observation point and its corresponding second target observation point is calculated, and based on each target distance, the wind speed attenuation factor corresponding to each first target observation point is calculated to obtain the wind speed attenuation factor of each observation point.

[0008] Based on the location coordinates of each observation point, the effective wind speed is predicted by the upper limit of the target wind farm height and the wind speed attenuation factor. The effective wind speed of each observation point is then weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine.

[0009] As a preferred approach, in the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to preset position coefficients, specifically as follows:

[0010] The Gaussian FLORIS wake model is used as the underlying model, and the FLORIDyn model is used as the dynamic implementation of the Gaussian FLORIS wake model. Based on the Gaussian FLORIS wake model and the FLORIDyn model, a wake simulation model is constructed.

[0011] In the wake simulation model, several observation points are created on the surface of the target wind turbine; these observation points are used to represent the local wake characteristics at their location.

[0012] The position coefficients at the time of creation of each observation point are used as preset position coefficients. Based on these preset position coefficients, the position coordinates of each observation point on the target wind turbine are determined using the following formula:

[0013]

[0014] Where y and z represent the position coordinates of the observation point on the y-axis and z-axis, respectively, and k width,y and k width,z These represent the preset position coefficients of the observation point on the y-axis and z-axis, respectively; width y and width z These represent the widths of the wake region at the observation point along the y-axis and z-axis, respectively.

[0015] As a preferred approach, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thereby obtaining the second target observation point corresponding to each first target observation point. Specifically:

[0016] A binary tree is constructed using observation points within a predetermined search range as nodes of the kd-tree;

[0017] Use the current first target observation point as the query point of the kd-tree;

[0018] Within a predetermined search range, when using a kd-tree for searching, the traversal starts from the root node of the binary tree, and the distance between the query point and the hyperrectangular region represented by the current node is compared. If the query point is located within the region represented by the left subtree of the current node, the left subtree is traversed downwards; otherwise, the right subtree is traversed downwards.

[0019] During the traversal, the data points of several nodes closest to the query point are recorded, as well as the distance between the data points and the query point. After traversing the binary tree, the search result is returned; where the search result is the data point of the node closest to the query point.

[0020] The search result is used as the second target observation point corresponding to the current first target observation point.

[0021] As a preferred option, the pre-determined search range is as follows:

[0022] The initial search range is defined by the area encompassing all observation points on the target wind turbine.

[0023] The initial search range is pruned based on a pruning distance threshold to obtain a predetermined search range; the pruning distance threshold is determined based on the diameter of the target wind turbine.

[0024] As a preferred approach, the initial search range is pruned based on a threshold of pruning distance to obtain a predetermined search range, specifically as follows:

[0025] Determine the initial search range of the search space for the observation points of the front-row wind turbines in the target wind farm, as well as the threshold for the pruning distance;

[0026] Within the initial search range of the target wind turbine in the search space, the region where the distance between the first target observation point and other observation points is less than or equal to the pruning distance threshold is retained to obtain the pruning retention region, which is then used as the pre-determined search range; wherein, the expression for the pre-determined search range is:

[0027]

[0028] in, For a pre-defined search range, j and i represent different wind turbines, Ω i Let OP represent the set of observation points for the i-th wind turbine, and x represent the observation points. min (j) represents the minimum value of the plane of the j-th wind turbine in the x-axis direction, ymin (j) represents the minimum value of the plane of the j-th wind turbine along the y-axis, z min (j) represents the minimum value of the plane of the j-th wind turbine along the z-axis, x max (j) represents the maximum value of the plane of the j-th wind turbine in the x-axis direction, y max (j) represents the maximum value of the plane of the j-th wind turbine in the y-axis direction, z max (j) represents the maximum value of the plane of the j-th wind turbine in the z-axis direction, and d represents the diameter of the target wind turbine.

[0029] As a preferred approach, based on the distance to each target, the wind speed attenuation factor corresponding to each first target observation point is calculated to obtain the wind speed attenuation factor for each observation point, specifically as follows:

[0030] Determine whether the target distance corresponding to the current first target observation point is less than a preset distance threshold; wherein, the preset distance threshold is determined based on the diameter of the target wind turbine;

[0031] If not, the wind speed attenuation factor corresponding to the current first target observation point is 0;

[0032] If so, calculate the ratio of the wind speed attenuation to the wind speed at infinity at the target wind farm to obtain the wind speed attenuation factor corresponding to the current first target observation point.

[0033] The wind speed attenuation factor corresponding to each first target observation point is used as the wind speed attenuation factor for each observation point.

[0034] As a preferred approach, based on the location coordinates of each observation point, the effective wind speed is predicted by considering the upper limit of the target wind farm's height and the wind speed attenuation factor, resulting in the effective wind speed at each observation point, specifically:

[0035] The height of each observation point is obtained based on its position coordinates.

[0036] The effective wind speed at each observation point is predicted by considering the height of each observation point, the upper limit of the target wind farm's height, and the wind speed attenuation factor. The formula is as follows:

[0037]

[0038] Among them, u eff,OP The effective wind speed at the current observation point, u free,OP This represents the free wind speed at the current observation point. For the wind speed at infinity at the target wind farm, z OP The current observation point's altitude is represented by α, which is the altitude attenuation factor, and z0 represents the upper limit of the target wind farm's altitude. own n represents the wake attenuation factor at the current observation point. T Indicates the number of wind turbines, ri This represents the wind speed attenuation factor corresponding to all current observation points belonging to the i-th wind turbine.

[0039] To address the same technical problem, this invention also provides a wind turbine power prediction device based on a wake simulation model, comprising: an observation point location acquisition module, a distance wind speed attenuation calculation module, and a wind turbine power prediction module.

[0040] The observation point location acquisition module is used to determine the position coordinates of all observation points on the target wind turbine in the constructed wake simulation model according to the preset position coefficients.

[0041] The distance-wind-speed attenuation calculation module is used to take each observation point as the first target observation point, and within a predetermined search range, use the KNN search algorithm to determine the observation point closest to each first target observation point, obtain the second target observation point corresponding to each first target observation point, calculate the target distance between each first target observation point and the corresponding second target observation point, and calculate the wind speed attenuation factor corresponding to each first target observation point based on each target distance, thus obtaining the wind speed attenuation factor of each observation point;

[0042] The wind turbine power prediction module is used to predict the effective wind speed based on the location coordinates of each observation point, the upper limit of the target wind farm height and the wind speed attenuation factor, to obtain the effective wind speed of each observation point. The effective wind speed of each observation point is then weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine.

[0043] To address the same technical problem, embodiments of the present invention also provide a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a wind turbine power prediction method based on a wake simulation model.

[0044] To address the same technical problem, embodiments of the present invention also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements a wind turbine power prediction method based on a wake simulation model.

[0045] In implementing this embodiment of the invention, in the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to a preset position coefficient. Each observation point is used as a first target observation point. Within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. The target distance between each first target observation point and its corresponding second target observation point is calculated. Based on each target distance, the wind speed attenuation factor corresponding to each first target observation point is calculated, thus obtaining the wind speed attenuation factor for each observation point. Based on the position coordinates of each observation point, the effective wind speed is predicted using the upper limit of the target wind farm's height and the wind speed attenuation factor, thus obtaining the effective wind speed for each observation point. The effective wind speeds of each observation point are then weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine. By calculating the attenuation factor and filtering data using the distance between observation points, the computational load is effectively reduced, eliminating the need to calculate data between distant observation points. This solves the problem of high computational cost and long simulation time in existing wake models when there are many wind turbines, thus reducing the computational load and shortening the simulation time for wind turbine power prediction. This achieves effective prediction of target wind turbine power, shortens simulation time, and improves wind turbine power prediction efficiency.

[0046] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0047] Figure 1 : A flowchart illustrating an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0048] Figure 2 : A simplified flowchart of an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0049] Figure 3 : A schematic diagram of the wake model of an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0050] Figure 4 : A schematic diagram of the state inheritance and movement of the observation point in the wake model of an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0051] Figure 5 : A schematic diagram of the search space before optimization for an embodiment of the wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0052] Figure 6 : An optimized search space diagram for an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0053] Figure 7 : A flowchart illustrating the calculation of the wind speed attenuation factor in an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention;

[0054] Figure 8 : A schematic diagram of an embodiment of a wind turbine power prediction device based on a wake simulation model provided by the present invention;

[0055] Figure 9 : A schematic diagram of the unit structure of an embodiment of a wind turbine power prediction device based on a wake simulation model provided by the present invention;

[0056] Figure 10 : A schematic diagram of the structure of an electronic device for an embodiment of a wind turbine power prediction method based on a wake simulation model provided by the present invention. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Example 1

[0059] Please refer to Figure 1 This is a flowchart illustrating a wind turbine power prediction method based on a wake simulation model, provided by an embodiment of the present invention. The wind turbine power prediction method includes steps 101 to 103, each step of which is detailed below:

[0060] Step 101: In the constructed wake simulation model, determine the position coordinates of all observation points on the target wind turbine according to the preset position coefficients.

[0061] In this embodiment, a simplified flowchart of the wind turbine power prediction method is shown below. Figure 2As shown, in the pre-constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to preset position coefficients. Any observation point is taken as the first target observation point. Within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to the first target observation point as the second target observation point, and the target distance between the first and second target observation points is determined. The wind speed attenuation factor between the first and second target observation points is calculated based on the target distance. The effective wind speed of each observation point is calculated, using each observation point as the first target observation point. The effective wind speeds of all observation points are weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power of the target wind turbine is calculated. This solves the problem of high computational cost in wake models when there are a large number of wind turbines.

[0062] Optionally, step 101 specifically includes steps 1011 to 1013, each step being as follows:

[0063] Step 1011: Using the Gaussian FLORIS wake model as the underlying model and the FLORIDyn model as the dynamic implementation of the Gaussian FLORIS wake model, construct a wake simulation model based on the Gaussian FLORIS wake model and the FLORIDyn model.

[0064] In this embodiment, the underlying model of the constructed wake simulation model is the Gaussian FLORIS wake model, and a schematic diagram of the wake model is shown below. Figure 3 As shown. Based on this wake model, the formula for calculating wind speed in the wake region is as follows:

[0065]

[0066] in, This represents the decrease in wind speed. The wind speed at infinity in the wind field. C represents the actual wind speed at that point in the wake region. T σ is the thrust coefficient, γ is the yaw angle of the wind turbine, and σ is the thrust coefficient. y ,σ z These represent the widths of the wake regions along the y-axis and z-axis, respectively (the z-axis is perpendicular to...). Figure 3 (xy plane in the image), where d is the fan diameter, z h This represents the height of the fan along the z-axis.

[0067] In the above wind speed calculation formula, the thrust coefficient C T The definition is as follows:

[0068]

[0069] Where T is the total force acting on the surface of the fan, and ρ is the air density. For incoming wind speed.

[0070] In the above wind speed calculation formula, the width σ of the wake region in the y-axis and z-axis directions y ,σ z The calculation formula is as follows:

[0071]

[0072] Where, k y ,k z y and z are the wake growth rates, respectively, and x0 is the length of the potential core in the wake region. The formula for calculating x0 is as follows:

[0073]

[0074] Where I is the turbulence intensity, α* = 2.32, β* = 0.154.

[0075] The formula for calculating δ is as follows:

[0076]

[0077] Where δ is the wake offset, representing the distance the wake is offset radially.

[0078] The formula for calculating θ is as follows:

[0079]

[0080] Where θ is the deflection angle of the wake relative to the ambient wind.

[0081] Step 1012: In the wake simulation model, create several observation points on the surface of the target wind turbine; whereby the observation points are used to represent the local wake characteristics at their location;

[0082] In this embodiment, to introduce dynamic calculations, the FLORIDyn model creates observation points (OPs) on the wind turbine surface in each simulation step. The purpose of these observation points is to describe the local Gaussian-FLORIS wake characteristics at their location. Each observation point inherits the wind turbine's state at the time of its creation, such as yaw angle, turbulence intensity, and thrust coefficient, which are used to calculate wind speed in the Gaussian-FLORIS model. Over time, each observation point moves downstream, representing a large amount of air moving in the wind. The path of movement is determined by the boundary of the FLORIS wake region, and the speed of movement is equal to the effective wind speed it represents.

[0083] Step 1013: Use the position coefficients at the time of creation of each observation point as preset position coefficients, and determine the position coordinates of each observation point on the target wind turbine based on the preset position coefficients, using the following formula:

[0084]

[0085] Where y and z represent the position coordinates of the observation point on the y-axis and z-axis, respectively, and k width,y and k width,z These represent the preset position coefficients of the observation point on the y-axis and z-axis, respectively. These position coefficients are determined when the observation point is created and remain unchanged as the observation point moves; width y and width z These represent the widths of the wake region at the observation point along the y-axis and z-axis, respectively.

[0086] In this embodiment, the inheritance and movement of the observation point state in the wake model are illustrated as follows: Figure 4 As shown in Figure ①, the black observation point is created on the surface of the wind turbine, inheriting the state before the wind turbine yaws. In Figure ②, the observation point moves a distance uΔt along the positive x-axis (i.e., the wind speed direction), where u is the effective wind speed of the observation point and Δt is the simulation step size. The positions of the observation point on the y-axis and z-axis (perpendicular to the xy plane) are as shown in Equation (7). In Figure ③, when the wind turbine yaws, the generated observation point inherits the new state, distinguished by white observation points from the previously generated ones. The state inherited by the previously generated black observation points remains unchanged. In Figure ④, the observation points that inherit two different states move downstream according to their inherited states.

[0087] Step 102: Using each observation point as the first target observation point, within a predetermined search range, use the KNN search algorithm to determine the observation point closest to each first target observation point, obtain the second target observation point corresponding to each first target observation point, calculate the target distance between each first target observation point and the corresponding second target observation point, and calculate the wind speed attenuation factor corresponding to each first target observation point based on each target distance, thus obtaining the wind speed attenuation factor of each observation point.

[0088] In this embodiment, any observation point is taken as the first target observation point. Within a predetermined search range, the KNN (KNNsearch, K-Nearest Neighbors) search algorithm is used to determine the observation point closest to the first target observation point as the second target observation point, and the target distance between the first and second observation points is determined. It should be understood that "first" and "second" in the first and second target observation points are for descriptive convenience only and do not indicate any order or limitation.

[0089] Optionally, step 102 specifically includes steps 1021 to 1022, each of which is as follows:

[0090] Step 1021: Using each observation point as the first target observation point, within a predetermined search range, use the KNN search algorithm to determine the observation point closest to each first target observation point, obtain the second target observation point corresponding to each first target observation point, and calculate the target distance between each first target observation point and the corresponding second target observation point.

[0091] Optional, predefined search scope, specifically:

[0092] The initial search range is defined by the area encompassing all observation points on the target wind turbine.

[0093] The initial search range is pruned based on a pruning distance threshold to obtain a predetermined search range; the pruning distance threshold is determined based on the diameter of the target wind turbine.

[0094] In this embodiment, the area encompassing all observation points on the target wind turbine is used as the initial search range. The initial search range is then pruned to obtain a predetermined search range. Specifically, the pruning of the initial search range involves first determining the initial search range for the observation points of the front-row wind turbines and a threshold for the pruning distance. Within the initial search range, regions with distances less than or equal to the pruning distance threshold are retained, and these regions are used as the pruned search range. The pruning distance threshold can be equal to the target distance threshold, i.e., one-quarter of the wind turbine diameter is taken as the threshold.

[0095] Optionally, the initial search range can be pruned based on a threshold of pruning distance to obtain a predetermined search range, specifically:

[0096] Determine the initial search range of the search space for the observation points of the front-row wind turbines in the target wind farm, as well as the threshold for the pruning distance;

[0097] Within the initial search range of the target wind turbine in the search space, the region where the distance between the first target observation point and other observation points is less than or equal to the pruning distance threshold is retained to obtain the pruning retention region, which is then used as the pre-determined search range; wherein, the expression for the pre-determined search range is:

[0098]

[0099] in, For a pre-defined search range, j and i represent different wind turbines, Ω i Let OP represent the set of observation points for the i-th wind turbine, and x represent the observation points. min(j) represents the minimum value of the plane of the j-th wind turbine in the x-axis direction, y min (j) represents the minimum value of the plane of the j-th wind turbine along the y-axis, z min (j) represents the minimum value of the plane of the j-th wind turbine along the z-axis, x max (j) represents the maximum value of the plane of the j-th wind turbine in the x-axis direction, y max (j) represents the maximum value of the plane of the j-th wind turbine in the y-axis direction, z max (j) represents the maximum value of the plane of the j-th wind turbine in the z-axis direction, and d represents the diameter of the target wind turbine.

[0100] In this embodiment, the search space is Ω. i The search space is large, and the time cost of the search is high. Furthermore, the found OP2 might be affected by ||POS. OP,1 -POS OP,2 The value is greater than 0.25d, but it has no effect on the effective wind speed of OP1, resulting in a waste of computation. Therefore, the search process can be optimized in the following way.

[0101] First, according to formula (14) ||POS OP,1 -POS OP,2 Pruning the search range by ||≤0.25d and then performing the search again will significantly reduce the time cost of the search, meaning the original search space will be reduced from:

[0102]

[0103] It becomes:

[0104]

[0105] However, a large number of pruning operations will still take up a lot of computation time. As can be seen from formula (17), the effective wind speed of the fan is only equal to the effective wind speed at the observation point on the fan surface (denoted as Ω). rotor,i This relates to [the previous statement]. The following search space can be obtained:

[0106]

[0107] When calculating the impact of the i-th wind turbine on the effective wind speed of the j-th wind turbine, the search space for each search is... This can reduce the computation time required for pruning operations. However... The calculations remain cumbersome. and The calculation is simple and can be used. replace

[0108] structure as follows:

[0109]

[0110] in:

[0111]

[0112] Due to Ω rotor,j The observation points in the image are relatively densely packed in space, therefore The search space has been significantly reduced. However, due to the potential existence of ||POS OP,1 -POS OP,2 In the case where ||>0.25d, in order to maintain complete consistency between the simulation results before and after optimization, it is still necessary to first determine the distance between the two points and then calculate r as in formula (14). i .

[0113] Optionally, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. Specifically:

[0114] A binary tree is constructed using observation points within a predetermined search range as nodes of the kd-tree;

[0115] Use the current first target observation point as the query point of the kd-tree;

[0116] Within a predetermined search range, when using a kd-tree for searching, the traversal starts from the root node of the binary tree, and the distance between the query point and the hyperrectangular region represented by the current node is compared. If the query point is located within the region represented by the left subtree of the current node, the left subtree is traversed downwards; otherwise, the right subtree is traversed downwards.

[0117] During the traversal, the data points of several nodes closest to the query point are recorded, as well as the distance between the data points and the query point. After traversing the binary tree, the search result is returned; where the search result is the data point of the node closest to the query point.

[0118] The search result is used as the second target observation point corresponding to the current first target observation point.

[0119] In this embodiment, a kd-tree is used to optimize the KNN search algorithm based on an optimized, pre-determined search range. A kd-tree is a data structure used to efficiently solve the k-nearest neighbor search problem. It is a binary tree where each node represents a hyperrectangular region containing all data points of that node. During a search, the kd-tree first traverses from the root node, comparing the distance between the query point and the hyperrectangular region represented by the current node. If the query point is within the region represented by the left subtree of the current node, the traversal continues down the left subtree; otherwise, the traversal continues down the right subtree. During the traversal, the kd-tree records the k nearest data points and their distances to the query point. After traversing the entire tree, the kd-tree returns these k nearest data points as the search results.

[0120] For ease of understanding, such as Figure 5 and Figure 6 As shown, the changes in the search space before and after optimization are compared using simulations of two wind turbines. Figure 5 and Figure 6 The table below shows the results of the optimization process. Before optimization, the search space contained 6,000 observation points when calculating the effective wind speed of the upstream wind turbine on the downstream wind turbine. After optimization, the number of observation points in the search space was reduced to 70.

[0121] Table 1 Comparison of Optimization Results

[0122]

[0123] As shown in Table 1, the simulation efficiency of the optimized multi-turbine wind farm is improved by more than 80%, and all simulation data results obtained before and after optimization are completely consistent.

[0124] Step 1022: Based on the distance to each target, calculate the wind speed attenuation factor corresponding to each first target observation point, and obtain the wind speed attenuation factor of each observation point.

[0125] In this embodiment, the wind speed attenuation factor between the first target observation point and the second target observation point is calculated based on the target distance, that is, the wind speed attenuation factor of each observation point.

[0126] Optionally, step 1022 specifically includes:

[0127] Determine whether the target distance corresponding to the current first target observation point is less than a preset distance threshold; wherein, the preset distance threshold is determined based on the diameter of the target wind turbine;

[0128] If not, the wind speed attenuation factor corresponding to the current first target observation point is 0;

[0129] If so, calculate the ratio of the wind speed attenuation to the wind speed at infinity at the target wind farm to obtain the wind speed attenuation factor corresponding to the current first target observation point.

[0130] The wind speed attenuation factor corresponding to each first target observation point is used as the wind speed attenuation factor for each observation point.

[0131] In this embodiment, the flowchart for calculating the wind speed attenuation factor is as follows: Figure 7 As shown, the wind speed attenuation factor between the first target observation point and the second target observation point is calculated based on the target distance, specifically including the following steps S410-S430:

[0132] S410: Compare the target distance with a preset distance threshold; wherein the preset distance threshold is determined based on the diameter of the target wind turbine, for example, using one-quarter of the wind turbine diameter as the distance threshold.

[0133] S420: When the target distance is greater than or equal to the preset distance threshold, the wind speed attenuation factor is 0; that is, when the distance between two target wind turbines is far, the wake effect of the two is small and can be ignored, thereby reducing the amount of calculation.

[0134] S430: When the target distance is less than a preset distance threshold, the wind speed attenuation factor is the ratio of the wind speed attenuation to the wind speed at infinity in the wind field.

[0135] As an example of this embodiment, all observation points belonging to the i-th wind turbine (denoted as Ω) i The KNN search algorithm is used to find the observation point that corresponds to the current effective wind speed (denoted as OP1, with location POS). OP,1 =(x OP,1 ,y OP,1 ,z OP,1 The nearest observation point (denoted as OP2, located at POS) OP,2 =(x OP,2 ,y OP,2 ,z OP,2 )). Calculate r based on the distance between two observation points. i r i Let ri represent the wind speed attenuation factor of the observation point closest to the i-th wind turbine among all observation points belonging to that observation point (i.e., the observation point currently calculating the effective wind speed). The expression for ri is:

[0136]

[0137] Where d is the fan diameter, and when the distance between OP1 and OP2 is greater than 0.25d, the influence of OP2 on OP1 is ignored.

[0138] Step 103: Based on the location coordinates of each observation point, predict the effective wind speed by taking the upper limit of the target wind farm's height and the wind speed attenuation factor to obtain the effective wind speed at each observation point. Then, perform a weighted calculation on the effective wind speed at each observation point to obtain the effective wind speed of the target wind turbine. Finally, calculate the power based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine to obtain the predicted power of the target wind turbine.

[0139] In this embodiment,

[0140] Optionally, step 104 specifically includes steps 1031 to 1033, each of which is as follows:

[0141] Step 1031: Based on the location coordinates of each observation point, predict the effective wind speed by using the upper limit of the target wind farm's height and the wind speed attenuation factor, and obtain the effective wind speed at each observation point.

[0142] In this embodiment, since each observation point is used as a first target observation point, the effective wind speed of each observation point can be calculated based on the position coordinates of the first target observation point (observation point), the upper limit of the height of the target wind farm, and the wind speed attenuation factor; wherein, the target wind farm is the wind farm where the target wind turbine is located.

[0143] Optional, step 1031, specifically:

[0144] The height of each observation point is obtained based on its position coordinates.

[0145] The effective wind speed at each observation point is predicted by considering the height of each observation point, the upper limit of the target wind farm's height, and the wind speed attenuation factor. The formula is as follows:

[0146]

[0147] Among them, u eff,OP The effective wind speed at the current observation point, u free,OP This represents the free wind speed at the current observation point. For the wind speed at infinity at the target wind farm, z OP This represents the current observation point's altitude, α is the altitude attenuation factor, and z0 represents the upper limit of the target wind farm's altitude. OP / z0<1,(z OP / z0) α This indicates that the effective wind speed at the observation point decreases as the height of the observation point decreases, r own This represents the wake attenuation factor at the current observation point, i.e., the wake attenuation factor of that observation point itself, n. T Indicates the number of wind turbines, r iThis represents the wind speed attenuation factor corresponding to all current observation points belonging to the i-th wind turbine, that is, the wind speed attenuation factor of the observation point closest to the i-th wind turbine (the observation point currently calculating the effective wind speed).

[0148] In this embodiment, within the framework of the aforementioned FLORIDyn model, the state change of the upstream wind turbine will not immediately propagate to the downstream. Instead, the observation point inheriting the new state needs to gradually move to the downstream region. When the upstream observation point propagates to the downstream wind turbine, it will affect the wind speed at the observation point of the downstream wind turbine. The effective wind speed of an observation point is calculated as shown in equation (15), where r represents the wind speed attenuation factor, and the calculation formula is as follows (in equation (15), rown and ri are both calculated by the following formula):

[0149]

[0150] Formula (16) is the expression on the left side of the equal sign in formula (1) above.

[0151] Step 1032: Calculate the effective wind speed at each observation point by weighting the wind speed to obtain the effective wind speed of the target wind turbine.

[0152] In this embodiment, the effective wind speed at all observation points is calculated using a weighted average to obtain the effective wind speed of the target fan. Specifically, after obtaining the effective wind speed at each observation point, the effective wind speed at the fan is calculated by weighted summation of the effective wind speeds at the observation points on the fan surface.

[0153]

[0154] Where, n OP w represents the number of observation points on the surface of the fan. i This represents the weight. The magnitude of the weight is related to the position of the operating position (OP). For specific rules, please refer to the weight selection rules of the standard FLORIDyn model.

[0155] Step 1033: Calculate the power based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine to obtain the predicted power of the target wind turbine.

[0156] In this embodiment, the power of the target wind turbine is calculated based on its effective wind speed and the obtained yaw angle. Specifically, the power of the wind turbine can be calculated after obtaining its effective wind speed.

[0157]

[0158] Among them, C P η is the power coefficient, η is the efficiency, and p is a coefficient representing the magnitude of the effect of yaw angle on the power of the wind turbine.

[0159] In implementing this embodiment of the invention, in the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to a preset position coefficient. Each observation point is used as a first target observation point. Within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. The target distance between each first target observation point and its corresponding second target observation point is calculated. Based on each target distance, the wind speed attenuation factor corresponding to each first target observation point is calculated, thus obtaining the wind speed attenuation factor for each observation point. Based on the position coordinates of each observation point, the effective wind speed is predicted using the upper limit of the target wind farm's height and the wind speed attenuation factor, thus obtaining the effective wind speed for each observation point. The effective wind speeds of each observation point are then weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine. By calculating the attenuation factor and filtering data using the distance between observation points, the computational load is effectively reduced, eliminating the need to calculate data between distant observation points. This solves the problem of high computational cost and long simulation time in existing wake models when there are many wind turbines, thus reducing the computational load and shortening the simulation time for wind turbine power prediction. This achieves effective prediction of target wind turbine power, shortens simulation time, and improves wind turbine power prediction efficiency.

[0160] Example 2

[0161] Accordingly, see Figure 8 , Figure 8 This is a schematic diagram of a second embodiment of a wind turbine power prediction device based on a wake simulation model provided by the present invention. Figure 8 As shown, the wind turbine power prediction device based on the wake simulation model includes an observation point location acquisition module 801, a distance wind speed attenuation calculation module 802, and a wind turbine power prediction module 803.

[0162] Among them, the observation point location acquisition module 801 is used to determine the position coordinates of all observation points on the target wind turbine in the constructed wake simulation model according to the preset position coefficients;

[0163] In this embodiment, a schematic diagram of the unit structure corresponding to the wind turbine power prediction device based on the wake simulation model is shown below. Figure 9 As shown, the observation point location acquisition module 801 includes a location acquisition unit 710; the location acquisition unit 710 is used to determine the location coordinates of all observation points on the target wind turbine according to preset location coefficients in a pre-constructed wake simulation model.

[0164] The distance-wind-speed attenuation calculation module 802 is used to take each observation point as the first target observation point, and within a predetermined search range, use the KNN search algorithm to determine the observation point closest to each first target observation point, obtain the second target observation point corresponding to each first target observation point, calculate the target distance between each first target observation point and the corresponding second target observation point, and calculate the wind speed attenuation factor corresponding to each first target observation point based on each target distance, so as to obtain the wind speed attenuation factor of each observation point.

[0165] In this embodiment, the distance-wind speed attenuation calculation module 802 includes a distance generation unit 720 and an attenuation calculation unit 730. The distance generation unit 720 is used to take any observation point as a first target observation point, and within a predetermined search range, use the KNN search algorithm to determine the observation point closest to the first target observation point as a second target observation point, and determine the distance between the first target observation point and the second target observation point. The attenuation calculation unit 730 is used to calculate the wind speed attenuation factor between the first observation point and the second observation point based on the target distance.

[0166] The wind turbine power prediction module 803 is used to predict the effective wind speed based on the location coordinates of each observation point, the upper limit of the target wind farm height and the wind speed attenuation factor, to obtain the effective wind speed of each observation point, and to calculate the effective wind speed of the target wind turbine by weighting the effective wind speed of each observation point. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the module calculates the power to obtain the predicted power of the target wind turbine.

[0167] In this embodiment, the wind turbine power prediction module 803 includes a wind speed calculation unit 740, a wind speed weighting unit 750, and a power calculation unit 760. The wind speed calculation unit 740 is used to calculate the effective wind speed at each observation point, taking each observation point as a first target observation point, based on the position coordinates of the observation point, the upper limit of the target wind farm's height, and the wind speed attenuation factor; wherein, the target wind farm is the wind farm where the target wind turbine is located. The wind speed weighting unit 750 is used to perform weighted calculations on the effective wind speeds of all observation points to obtain the effective wind speed of the target wind turbine. The power calculation unit 760 is used to calculate the power of the target wind turbine based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine.

[0168] Optionally, in the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to preset position coefficients, specifically as follows:

[0169] The Gaussian FLORIS wake model is used as the underlying model, and the FLORIDyn model is used as the dynamic implementation of the Gaussian FLORIS wake model. Based on the Gaussian FLORIS wake model and the FLORIDyn model, a wake simulation model is constructed.

[0170] In the wake simulation model, several observation points are created on the surface of the target wind turbine; these observation points are used to represent the local wake characteristics at their location.

[0171] The position coefficients at the time of creation of each observation point are used as preset position coefficients. Based on these preset position coefficients, the position coordinates of each observation point on the target wind turbine are determined using the following formula:

[0172]

[0173] Where y and z represent the position coordinates of the observation point on the y-axis and z-axis, respectively, and k width,y and k width,z These represent the preset position coefficients of the observation point on the y-axis and z-axis, respectively; width y and width z These represent the widths of the wake region at the observation point along the y-axis and z-axis, respectively.

[0174] Optionally, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. Specifically:

[0175] A binary tree is constructed using observation points within a predetermined search range as nodes of the kd-tree;

[0176] Use the current first target observation point as the query point of the kd-tree;

[0177] Within a predetermined search range, when using a kd-tree for searching, the traversal starts from the root node of the binary tree, and the distance between the query point and the hyperrectangular region represented by the current node is compared. If the query point is located within the region represented by the left subtree of the current node, the left subtree is traversed downwards; otherwise, the right subtree is traversed downwards.

[0178] During the traversal, the data points of several nodes closest to the query point are recorded, as well as the distance between the data points and the query point. After traversing the binary tree, the search result is returned; where the search result is the data point of the node closest to the query point.

[0179] The search result is used as the second target observation point corresponding to the current first target observation point.

[0180] Optional, predefined search scope, specifically:

[0181] The initial search range is defined by the area encompassing all observation points on the target wind turbine.

[0182] The initial search range is pruned based on a pruning distance threshold to obtain a predetermined search range; the pruning distance threshold is determined based on the diameter of the target wind turbine.

[0183] Optionally, the initial search range can be pruned based on a threshold of pruning distance to obtain a predetermined search range, specifically:

[0184] Determine the initial search range of the search space for the observation points of the front-row wind turbines in the target wind farm, as well as the threshold for the pruning distance;

[0185] Within the initial search range of the target wind turbine in the search space, the region where the distance between the first target observation point and other observation points is less than or equal to the pruning distance threshold is retained to obtain the pruning retention region, which is then used as the pre-determined search range; wherein, the expression for the pre-determined search range is:

[0186]

[0187] in, For a pre-defined search range, j and i represent different wind turbines, Ω i Let OP represent the set of observation points for the i-th wind turbine, and x represent the observation points. min (j) represents the minimum value of the plane of the j-th wind turbine in the x-axis direction, y min (j) represents the minimum value of the plane of the j-th wind turbine along the y-axis, z min (j) represents the minimum value of the plane of the j-th wind turbine along the z-axis, x max (j) represents the maximum value of the plane of the j-th wind turbine in the x-axis direction, y max (j) represents the maximum value of the plane of the j-th wind turbine in the y-axis direction, z max (j) represents the maximum value of the plane of the j-th wind turbine in the z-axis direction, and d represents the diameter of the target wind turbine.

[0188] Optionally, based on the distance to each target, the wind speed attenuation factor corresponding to each first target observation point is calculated to obtain the wind speed attenuation factor for each observation point, specifically as follows:

[0189] Determine whether the target distance corresponding to the current first target observation point is less than a preset distance threshold; wherein, the preset distance threshold is determined based on the diameter of the target wind turbine;

[0190] If not, the wind speed attenuation factor corresponding to the current first target observation point is 0;

[0191] If so, calculate the ratio of the wind speed attenuation to the wind speed at infinity at the target wind farm to obtain the wind speed attenuation factor corresponding to the current first target observation point.

[0192] The wind speed attenuation factor corresponding to each first target observation point is used as the wind speed attenuation factor for each observation point.

[0193] Optionally, as a preferred option, based on the location coordinates of each observation point, the effective wind speed is predicted by the upper limit of the target wind farm's height and the wind speed attenuation factor, resulting in the effective wind speed at each observation point, specifically:

[0194] The height of each observation point is obtained based on its position coordinates.

[0195] The effective wind speed at each observation point is predicted by considering the height of each observation point, the upper limit of the target wind farm's height, and the wind speed attenuation factor. The formula is as follows:

[0196]

[0197] Among them, u eff,OP The effective wind speed at the current observation point, u free,OP This represents the free wind speed at the current observation point. For the wind speed at infinity at the target wind farm, z OP The current observation point's altitude is represented by α, which is the altitude attenuation factor, and z0 represents the upper limit of the target wind farm's altitude. own n represents the wake attenuation factor at the current observation point. T Indicates the number of wind turbines, r i This represents the wind speed attenuation factor corresponding to all current observation points belonging to the i-th wind turbine.

[0198] In the above specific embodiments, the wind turbine power prediction device based on a wake simulation model provided by the present invention determines the position coordinates of all observation points on the target wind turbine according to a preset position coefficient in a pre-constructed wake simulation model; takes any observation point as the first target observation point, and within a predetermined search range, uses the KNN search algorithm to determine the observation point closest to the first target observation point as the second target observation point, and determines the target distance between the first target observation point and the second target observation point; calculates the wind speed attenuation factor between the first observation point and the second observation point based on the target distance; takes each observation point as the first target observation point, and calculates the effective wind speed of each observation point based on the position coordinates of the target observation point, the upper limit of the height of the target wind farm, and the wind speed attenuation factor; wherein, the target wind farm is the wind farm where the target wind turbine is located; performs a weighted calculation on the effective wind speed of all observation points to obtain the effective wind speed of the target wind turbine; and calculates the power of the target wind turbine based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine.

[0199] Thus, the device provided by the present invention calculates the attenuation factor and performs filtering by utilizing the distance between observation points, thereby eliminating the need to calculate data between observation points that are far apart, effectively reducing the amount of computation, and solving the problem of high computational cost and long simulation time in the wake model when there are many wind turbines in the prior art, thereby reducing the amount of computation for wind turbine power prediction and shortening the simulation time.

[0200] Figure 10 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 10 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call logic instructions in the memory 830 to execute a wind turbine power prediction method based on a wake simulation model, the method including:

[0201] In the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to the preset position coefficients;

[0202] Using each observation point as the first target observation point, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each first target observation point, thus obtaining the second target observation point corresponding to each first target observation point. The target distance between each first target observation point and its corresponding second target observation point is calculated, and based on each target distance, the wind speed attenuation factor corresponding to each first target observation point is calculated to obtain the wind speed attenuation factor of each observation point.

[0203] Based on the location coordinates of each observation point, the effective wind speed is predicted by the upper limit of the target wind farm height and the wind speed attenuation factor. The effective wind speed of each observation point is then weighted to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine.

[0204] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in any of the above method embodiments.

[0205] The aforementioned wind turbine power prediction device based on a wake simulation model can implement the wind turbine power prediction method based on a wake simulation model described in the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application's embodiments can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0206] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention in detail. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A wind turbine power prediction method based on a wake simulation model, characterized in that, include: In the constructed wake simulation model, the position coordinates of all observation points on the target wind turbine are determined according to preset position coefficients. Specifically, the Gaussian FLORIS wake model is used as the underlying model, and the FLORIDyn model is used as the dynamic implementation of the Gaussian FLORIS wake model. Based on the Gaussian FLORIS wake model and the FLORIDyn model, the wake simulation model is constructed. In the wake simulation model, several observation points are created on the surface of the target wind turbine. These observation points represent the local wake characteristics at their respective locations. The position coefficients at the time of creation of each observation point are used as preset position coefficients, and the position coordinates of each observation point on the target wind turbine are determined based on these preset position coefficients, using the following formula: in, and These respectively represent the observation points at... shaft and The position coordinates of the axis and These respectively represent the observation points at... shaft and The preset position coefficient of the shaft; and These respectively represent the wake regions at the observation points. shaft and Width along the axial direction; Using each observation point as a first target observation point, within a predetermined search range, the KNN search algorithm is used to determine the observation point closest to each of the first target observation points, thus obtaining the second target observation point corresponding to each of the first target observation points. Specifically, the following steps are taken: using the observation points within the predetermined search range as nodes of a kd-tree, a binary tree is constructed; the current first target observation point is used as the query point of the kd-tree; when searching within the predetermined search range using the kd-tree, the search begins from the root node of the binary tree, and the distance between the query point and the hyperrectangular region represented by the current node is compared; if the query point is located within the region represented by the left subtree of the current node, the left subtree is traversed downwards; otherwise, the right subtree is traversed downwards; during the traversal, the data points of several nodes closest to the query point, as well as the distance between the data points and the query point, are recorded. After traversing the binary tree, ... Return the search results; wherein the search results are the data points of the nodes closest to the query point; use the search results as the second target observation points corresponding to the current first target observation point, and calculate the target distance between each first target observation point and each corresponding second target observation point, and calculate the wind speed attenuation factor corresponding to each first target observation point based on each target distance, to obtain the wind speed attenuation factor of each observation point, specifically: determine whether the target distance corresponding to the current first target observation point is less than a preset distance threshold; wherein the preset distance threshold is determined according to the turbine diameter of the target wind turbine; if not, the wind speed attenuation factor corresponding to the current first target observation point is 0; if yes, calculate the ratio of the wind speed attenuation to the wind speed at infinity of the target wind farm, to obtain the wind speed attenuation factor corresponding to the current first target observation point; use the wind speed attenuation factor corresponding to each first target observation point as the wind speed attenuation factor of each observation point; The height of each observation point is obtained based on its position coordinates. The effective wind speed at each observation point is predicted by combining the height of each observation point, the upper limit of the height of the target wind farm, and the wind speed attenuation factor. The formula is as follows: in, The effective wind speed at the current observation point. This represents the free wind speed at the current observation point. Let be the wind speed at infinity at the target wind farm. This indicates the height of the current observation point. As a highly attenuating factor, This indicates the upper limit of the height of the target wind farm. This represents the wake attenuation factor at the current observation point. Indicates the number of wind turbines. This represents the wind speed attenuation factor corresponding to all current observation points belonging to the i-th wind turbine. The effective wind speeds of each observation point are weighted and calculated to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine.

2. The wind turbine power prediction method based on wake simulation model as described in claim 1, characterized in that, The predetermined search range is specifically as follows: The initial search range is defined by the range of all the observation points on the target wind turbine. The initial search range is pruned based on a pruning distance threshold to obtain the predetermined search range; the pruning distance threshold is determined based on the diameter of the target wind turbine.

3. The wind turbine power prediction method based on wake simulation model as described in claim 2, characterized in that, The initial search range is pruned based on a threshold value determined by the pruning distance to obtain the predetermined search range, specifically as follows: Determine the initial search range of the search space for the observation points of the front wind turbines of the target wind farm and the threshold value of the pruning distance; Within the initial search range of the target wind turbine in the search space, a region is retained where the distance between the first target observation point and other observation points is less than or equal to the pruning distance threshold, thus obtaining a pruning retention region, which is used as the predetermined search range; wherein, the expression for the predetermined search range is: in, For the predetermined search range, Indicates different types of fans. Indicates the first Collection of observation points for typhoon generators. Indicates the observation point. For the first Typhoon plane Minimum value in the axial direction, For the first Typhoon plane Minimum value in the axial direction, For the first Typhoon plane Minimum value in the axial direction, For the first Typhoon plane Maximum value in the axial direction, For the j-th wind turbine plane in Maximum value in the axial direction, For the j-th wind turbine plane in Maximum value in the axial direction, The diameter of the target fan is given.

4. A wind turbine power prediction device based on a wake simulation model, characterized in that, include: The module includes an observation point location acquisition module, a distance-wind-speed attenuation calculation module, and a wind turbine power prediction module. The observation point location acquisition module is used to determine the position coordinates of all observation points on the target wind turbine in the constructed wake simulation model according to preset position coefficients. Specifically, it uses the Gaussian FLORIS wake model as the underlying model and the FLORIDyn model as the dynamic implementation of the Gaussian FLORIS wake model. Based on the Gaussian FLORIS wake model and the FLORIDyn model, the wake simulation model is constructed. In the wake simulation model, several observation points are created on the surface of the target wind turbine. The observation points represent the local wake characteristics at their location. The position coefficients at the time of creation of each observation point are used as preset position coefficients, and the position coordinates of each observation point on the target wind turbine are determined based on the preset position coefficients, using the following formula: in, and These respectively represent the observation points at... shaft and The position coordinates of the axis and These respectively represent the observation points at... shaft and The preset position coefficient of the shaft; and These respectively represent the wake regions at the observation points. shaft and Width along the axial direction; The distance-wind-speed attenuation calculation module is used to determine the observation point closest to each of the first target observation points within a predetermined search range using the KNN search algorithm, thereby obtaining the second target observation point corresponding to each of the first target observation points. Specifically, it constructs a binary tree using the observation points within the predetermined search range as nodes of a kd-tree; it uses the current first target observation point as the query point of the kd-tree; when searching within the predetermined search range using the kd-tree, it traverses from the root node of the binary tree and compares the distance between the query point and the hyperrectangular region represented by the current node; if the query point is located within the region represented by the left subtree of the current node, it traverses the left subtree downwards; otherwise, it traverses the right subtree downwards; during the traversal, it records the data points of several nodes closest to the query point, as well as the distance between the data points and the query point. When the traversal is complete... After the binary tree is established, the search results are returned; wherein, the search results are the data points of the nodes closest to the query point; the search results are used as the second target observation points corresponding to the current first target observation point, and the target distances between each first target observation point and the corresponding second target observation points are calculated, and based on each target distance, the wind speed attenuation factor corresponding to each first target observation point is calculated to obtain the wind speed attenuation factor of each observation point, specifically: determining whether the target distance corresponding to the current first target observation point is less than a preset distance threshold; wherein, the preset distance threshold is determined according to the diameter of the target wind turbine; if not, the wind speed attenuation factor corresponding to the current first target observation point is 0; if yes, the ratio of the wind speed attenuation to the wind speed at infinity at the target wind farm is calculated to obtain the wind speed attenuation factor corresponding to the current first target observation point; the wind speed attenuation factor corresponding to each first target observation point is used as the wind speed attenuation factor of each observation point; The wind turbine power prediction module is used to obtain the height of each observation point based on its location coordinates; and to predict the effective wind speed of each observation point by combining the height of each observation point, the upper limit of the target wind farm's height, and the wind speed attenuation factor, as shown in the formula: in, The effective wind speed at the current observation point. This represents the free wind speed at the current observation point. Let be the wind speed at infinity at the target wind farm. This indicates the height of the current observation point. As a highly attenuating factor, This indicates the upper limit of the height of the target wind farm. This represents the wake attenuation factor at the current observation point. Indicates the number of wind turbines. This represents the wind speed attenuation factor corresponding to all current observation points belonging to the i-th wind turbine. The effective wind speeds of each observation point are weighted and calculated to obtain the effective wind speed of the target wind turbine. Based on the effective wind speed of the target wind turbine and the obtained yaw angle of the target wind turbine, the power is calculated to obtain the predicted power of the target wind turbine.

5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the wind turbine power prediction method based on the wake simulation model as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the wind turbine power prediction method based on a wake simulation model as described in any one of claims 1 to 3.

Images