Wind turbine power prediction method based on wake flow model
Through the fan power prediction method based on wake model, the problem of difficulty in considering the complex environment and wake influence of wind farms in the prior art is solved, and higher prediction accuracy and adaptability are achieved.
Patent Information
- Application Number
- CN202510095624.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-21
- Publication Date
- 2025-05-30
AI Technical Summary
The existing fan power prediction technology is difficult to consider the complex local physical environment changes of the wind farm, and it is impossible to accurately calculate the complex impact of wake flow, resulting in low prediction accuracy and relying too much on historical data, making it difficult to reflect real-time state.
Through the fan power prediction method based on wake model, wind farm data is collected, wind turbine clustering is carried out, coordinate systems are established, propagation wind speed is calculated, mutual influence of wake flow, analysis of prediction deviations, and prediction of wind speed and power.
The comprehensiveness of wind speed and wind direction data collection in different areas of the wind farm has been improved, the wind turbine data has been refined, the wake impact has been accurately quantified, and the accuracy of fan power prediction has been improved.
Smart Images
Figure CN120068703A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of new energy power generation, and more specifically, to a method for predicting wind turbine power based on a wake model. Background Art
[0002] With the continuous growth of global energy demand and the concern for sustainable development, wind power generation is a clean and sustainable way to obtain energy; during the construction and operation of a wind farm, the power of wind turbines directly affects the economic benefits and energy utilization rate of the wind farm, and the wake effect between wind turbines will cause the wind speed of downstream wind turbines to decrease and the turbulence intensity to increase, thus affecting the power generation efficiency of the entire wind farm.
[0003] To achieve the efficient, stable and reliable operation of wind power generation, wind turbine power prediction technology has emerged; currently, based on historical meteorological data and the operation data of wind turbines, statistical methods are used to predict the power of wind turbines. Through a data acquisition module, various environmental and wind turbine operation data from the wind farm are collected, and through a data analysis module, corresponding statistical models are used to process and analyze the data, and the power prediction results are output.
[0004] However, in actual use, there are still some disadvantages, such as it is difficult to consider the complex local physical environment changes in the wind farm and unable to accurately calculate the complex influence of the wake; it is overly dependent on historical data and difficult to accurately reflect the wind speed changes and power losses caused by the wake, resulting in low accuracy of wind turbine power prediction; it is difficult to flexibly adjust the prediction model according to the real-time state of the wind farm, and it has poor generality under different terrains, different wind turbine types and different atmospheric conditions. Summary of the Invention
[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method for predicting wind turbine power based on a wake model, and through the following solutions, the problems raised in the above background art are solved.
[0006] To achieve the above object, the present invention provides the following technical solutions:
[0007] A method for predicting wind turbine power based on a wake model, characterized by comprising:
[0008] S1: Collect data of the target wind farm: In response to the pre-set anemometer tower in the target wind farm, collect the first wind power data set of each wind turbine through an information collection device;
[0009] S2: Perform clustering operation on the wind turbines: Perform clustering analysis on the first wind power data set of each wind turbine to obtain an optimal set of wind turbine clusters;
[0010] S3: Establish the coordinate system of the clustering clusters: Construct an initial coordinate system for the wind measurement towers of the target wind farm, and perform transformation operations on the corresponding wind turbines in the optimal set of wind turbine clusters. The transformation operations are used to obtain the corresponding coordinate points of each wind turbine in the initial coordinate system;
[0011] S4: Calculate the propagation wind speed: Calculate based on the first wind power data sets of each wind turbine and the corresponding coordinate points to obtain the second wind power data sets corresponding to the first wind power data sets of each wind turbine;
[0012] S5: Evaluate the mutual influence of the wake: Obtain a multiple wake model, and evaluate the wake influence matrix in multiple directions through the multiple wake model with the second wind power data sets;
[0013] S6: Analyze the prediction deviation: Calculate the deviation between the predicted wind speed and the actual wind speed of each wind turbine to obtain the wind speed prediction deviation value;
[0014] S7: Predict the wind speed of the wind turbines: Based on the wake influence matrix in multiple directions and the prediction deviation correction value, predict the wind speed of each wind turbine to obtain the predicted wind speed of each wind turbine;
[0015] S8: Predict the ultra-short-term wind turbine power: Based on the actual power generation efficiency of the wind turbine and the wind speed prediction deviation value, perform prediction operations on the short-time interval corresponding to a single wind turbine to obtain the ultra-short-term wind turbine power.
[0016] Preferably, for the said S2, obtaining the optimal set of wind turbine clusters specifically includes:
[0017] The clustering condition is that the minimum distance between the mass points of multiple clusters is greater than 10 times the wind wheel diameter;
[0018] The height difference of the hub of the mass points within the cluster is less than or equal to 100 meters.
[0019] Preferably, for the said S3, constructing the initial coordinate system for the wind measurement towers of the target wind farm specifically includes:
[0020] B1: In the optimal set of wind turbine clusters, obtain the wind measurement tower with the closest calculated distance to each wind turbine;
[0021] B2: Denote the 0° direction of the wind vane of the wind measurement tower corresponding to B1 as the X+ axis direction, and rotate the X+ axis counterclockwise by 90° to be the Y+ direction, that is, the initial coordinate system;
[0022] B3: Based on the initial coordinate system, perform transformation operations on the geographical location coordinates of each wind turbine.
[0023] Preferably, for the said S4, obtaining the second wind power data sets corresponding to the first wind power data sets of each wind turbine specifically includes: The judgment rules for the wind speed inflow of each wind turbine are as follows:
[0024] C401: When X - XT ≤ 0, the wind turbine is located upstream of the incoming wind direction.
[0025] C402: When X - XT > 0, the wind turbine is located downstream of the incoming wind direction.
[0026] Preferably, in S5, the wake influence matrix is a matrix composed of wake influence factors calculated from each wind turbine as the target comparison wind turbine and the wake influence units.
[0027] The target comparison unit is the corresponding wind turbine for analyzing the wake influence.
[0028] The wake influence wind turbine is the wind turbine that generates wake influence on the target comparison wind turbine.
[0029] Preferably, in S5, evaluating the wake influence matrix in multiple directions specifically includes:
[0030] D1: Obtain a multiple wake model and determine the corresponding parameters of the multiple wake model.
[0031] D2: Calculate the wake influence factors corresponding to multiple cases.
[0032] Preferably, in S5, evaluating the wake influence matrix in multiple directions specifically includes:
[0033] Based on the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm and the distance CD between the center of the downstream turbine and the center of the wake effect nm , compare the interference relationship between the target comparison wind turbine and the wake influence unit. Denote the target comparison unit as n, the wake influence wind turbine as m, and the wake influence factor of wind turbine n on wind turbine m as w nm , where n ∈ N, m ∈ N, and N represents the positions of each wind turbine;
[0034] D201: When there is no interference relationship between the target comparison unit and the wake influence wind turbine, that is, when PD nm ≤ 0 or CD nm ≥ 2 × r 0 × α × PD nm at this time, the wake influence factor w1 nm = 0;
[0035] D202: When the shadows of the target comparison unit and the wake influence wind turbine completely overlap, that is, PD nm ≤ 0 and CD nm ≥ 2 × r 0 × α × PD nm at this time, based on the rotor radius r corresponding to the target comparison unit 0, thrust coefficient C corresponding to the target comparison unit T , and the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm , calculate the wake influence factor w2 nm , specifically expressed as:
[0036]
[0037] Among them, α represents the wind farm wake diffusion coefficient;
[0038] D203: The shadow of the research unit and the comparison fan partially overlap, that is, PD nm >0 and α×PD nm <CD nm <2×r 0 ×α×PD nm When comparing the rotor radius r of the unit based on the target 0 , and the distance CD between the center of the downstream turbine and the center of the wake effect nm , calculate the overlapping shadow area impact factor SA, which is specifically expressed as:
[0039] SA=[r n (PD nm )] 2 ×
[0040]
[0041] Among them, r n (PD nm ) represents the PD after the upstream wind turbine nm The wake radius at nm It is expressed as the distance from the intersection of the wake area and the target unit rotor rotation range to the center of the wake effect, z nm It is expressed as the distance between the two intersection points of the wake area and the target comparison unit rotor rotation range;
[0042] Based on the overlap shadow area influence factor SA and the thrust coefficient C corresponding to the target comparison unit T 、The rotor radius r corresponding to the target comparison unit 0 , and the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm , calculate the wake influence factor w 3 nm , specifically expressed as:
[0043]
[0044] Where α is the wind farm wake diffusion coefficient, SA 0 Expressed as the preset overlapping shadow area influence factor.
[0045] Preferably, for S6, obtaining the wind speed prediction deviation value specifically includes:
[0046] Calculating the relative deviation between the predicted wind speed of each wind turbine as the target comparison turbine and the measured wind speed of the turbine;
[0047] Calculating the unit average relative deviation value for the relative deviation values of each wind turbine as the target comparison turbine according to a preset time scale, and the preset time scale is a time period set according to user requirements;
[0048] Sorting and recording according to the wind turbine number and time interval to generate a relative prediction deviation sequence;
[0049] Performing a normal distribution on the relative prediction deviation sequence, and selecting the wind speed prediction deviation value corresponding to the maximum probability value.
[0050] Technical effects and advantages of the present invention:
[0051] 1. The present invention pre-sets the anemometer tower according to the topography, area size of the target wind farm and the expected distribution of wind turbines, improving the comprehensiveness of wind speed and wind direction data collection in different areas of the wind farm, and thus providing more accurate basic data for subsequent wind turbine power prediction.
[0052] 2. The present invention calculates and obtains the second wind power data set based on the first wind power data set of each wind turbine and the corresponding coordinate points, refining the wind turbine data, more comprehensively considering the propagation characteristics of the wind, and improving the accuracy of subsequent wake effect analysis and power prediction.
[0053] 3. The present invention obtains multiple wake models, calculates the wake influence factors respectively according to the interference situation between the target comparison turbine and the wake-affected wind turbines, and then obtains the wake influence matrix, more accurately quantifying the mutual influence of wakes between wind turbines, improving the accuracy of wake influence analysis, and helping to improve the accuracy of wind turbine power prediction. Description of the Drawings
[0054] Figure 1 It is a method step diagram of the present invention.
[0055] Figure 2 It is a wind speed and wind direction analysis diagram of the present invention.
[0056] Figure 3 It is an analysis structure diagram of the corresponding turbines and wind speed lines in the wake influence assessment of the present invention. Detailed Embodiments
[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0058] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular forms "a", "an", "the", "above-mentioned", "said", and "this" are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more of the listed items.
[0059] Hereinafter, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be construed as implying or suggesting relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first", "second", and "third" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.
[0060] As shown in the attached Figure 1 The wind turbine power prediction method based on the wake model includes S1: collecting target wind farm data, S2: clustering the wind turbines, S3: establishing a coordinate system for the clustering clusters, S4: calculating the propagation wind speed, S5: evaluating the mutual influence of the wakes, S6: analyzing the prediction deviation, S7: predicting the wind speed of the turbines, and S8: predicting the ultra-short-term wind turbine power.
[0061] S1: Collecting target wind farm data: In response to the pre-set anemometer towers in the target wind farm, the first wind power data set of each wind turbine is collected through the information collection device.
[0062] Specifically, through a variety of information collection devices, the information collection devices include but are not limited to anemometers, wind vanes, etc., and the first wind power data set of each wind turbine is collected based on the pre-set anemometer towers in the target wind farm. The pre-set anemometer towers are used to achieve full coverage of the actual layout of the target wind farm.
[0063] Further, according to the topography, area size, and expected wind turbine distribution of the target wind farm, pre-set the anemometer towers; when the target wind farm is an onshore wind farm, arrange multiple anemometer towers in the central area and the edge area of the target wind farm to obtain wind speed and wind direction data in different areas. The information acquisition device includes multiple sensor units with different functions, and the sensor units adopt ultrasonic technology; furthermore, the wind speed sensor adopts the ultrasonic principle, calculates the wind speed by measuring the time difference of ultrasonic waves propagating in the air, sets the measurement accuracy to ±0.1 m / s, and sets the acquisition frequency to 1 time per minute; the wind direction sensor is based on a wind vane and records the wind direction angle through an angle sensor, sets the angle resolution to 1°, and continuously records the wind direction change situation at a set unit time interval; the barometric pressure sensor is based on the piezoresistive effect principle, converts the barometric pressure change into a measurable signal, and the measurement accuracy reaches ±0.1 hPa.
[0064] It should be noted that the information acquisition device establishes independent data acquisition tasks for each wind turbine based on the number and position coordinates of each wind turbine. The first wind power data group obtains data from multiple anemometer towers with the shortest relative distance according to the effective radius of the data acquisition task. The relative distance is the anemometer tower within the pre-set range centered on the target wind turbine to be measured; to form the first wind power data group, the first wind power data group includes the air density of the target wind farm, the swept area of each wind turbine, the real-time wind speed value of each wind turbine, the real-time wind direction of each wind turbine, and the number and position coordinate information corresponding to the wind turbine.
[0065] In this embodiment, for a wind farm with 100 wind turbines, the data of the anemometer towers within a radius of 500 meters centered on the target wind turbine to be measured is used as key data.
[0066] S2: Perform clustering operation on the wind turbines: perform clustering analysis on the first wind power data group of each wind turbine to obtain the optimal set of wind turbine clusters.
[0067] Specifically, use the geographical coordinates and hub height of the wind turbines as labels, and perform clustering analysis on the first wind power data group of the wind turbines through a machine learning clustering algorithm. The clustering analysis operation is to calculate the different dimensional characteristics of the geographical coordinates and hub height by using a variety of distance measurement methods to generate multiple clusters; perform a preferred judgment condition on the generated multiple clusters to obtain the cluster with the greatest influence on the wind turbine wake, that is, the optimal set of wind turbine clusters.
[0068] In a possible implementation manner, obtaining the optimal set of wind turbine clusters includes:
[0069] A1: Determine the machine learning clustering algorithm; in this embodiment, the K-Means algorithm is used as the machine learning clustering algorithm for the clustering analysis operation;
[0070] A2: Determine the number of clusters generated by clustering according to the scale of the target wind farm; specifically, for a wind farm with an occupied area ranging from 9 square kilometers to 90 square kilometers and 50 to 300 wind turbines, initially set the number of clusters to 5, allocate the wind turbines to the nearest cluster center, and update the position of the cluster center through clustering iteration until the position of the cluster center is stable;
[0071] A3: Use a machine learning clustering algorithm to perform clustering based on the geographical coordinate positions and hub heights of the wind turbines as labels; regard the wind turbines as data points, construct the feature vector of the data points with the geographical coordinate position (x, y) and the hub height h, denoted as [x, y, h]; during the clustering iteration process, calculate the Euclidean distance between the feature vector of the wind turbine and each cluster center, and allocate the wind turbine to the corresponding cluster according to the minimum distance principle;
[0072] A4: Judge the multiple clusters generated by clustering based on the preferred judgment conditions, and select the optimal set of wind turbine clustering; it should be noted that the clustering condition is that the minimum distance between the mass points of multiple clusters is greater than 10 times the rotor diameter; the difference in hub height of the mass points within the cluster is less than or equal to 100 meters.
[0073] S3: Establish the coordinate system of the clustering cluster: construct the initial coordinate system for the wind measurement towers of the target wind farm, and perform transformation operations on the corresponding wind turbines in the optimal set of wind turbine clustering. The transformation operation is used to obtain the coordinate points corresponding to each wind turbine in the initial coordinate system.
[0074] Specifically, take the position of the nearest wind measurement tower corresponding to the optimal set of wind turbine clustering as the center coordinate point of the coordinate system, stipulate the axis direction according to the direction of the wind vane of the wind measurement tower, and then obtain the coordinate points corresponding to each wind turbine in the initial coordinate system through the transformation operation of the geographical location coordinates of each wind turbine.
[0075] In a possible implementation manner, performing the transformation operation on the corresponding wind turbines in the optimal set of wind turbine clustering includes:
[0076] B1: In the optimal set of wind turbine clustering, obtain the wind measurement tower with the shortest calculated distance for each wind turbine;
[0077] Specifically, take the wind turbine at the center of the optimal set of wind turbine clustering as the coordinate origin, and denote the coordinates of each wind turbine as (xg i , yg i ), and the coordinates of the wind measurement tower as (xt j , yt j ), calculate the distance DGT between each wind turbine and the corresponding wind measurement tower, specifically expressed as:
[0078]
[0079] Among them, xg i represents the abscissa of the i-th wind turbine relative to the central wind turbine in the optimal set of wind turbine clustering, and yg i represents the ordinate of the i-th wind turbine relative to the central wind turbine in the optimal set of wind turbine clustering. i represents the index of all wind turbines except the central wind turbine in the optimal set of wind turbine clustering. xt j represents the abscissa of the j-th wind measurement tower in the optimal set of wind turbine clustering, and yt j represents the ordinate of the j-th wind measurement tower in the optimal set of wind turbine clustering. i represents the index of the wind measurement tower except in the optimal set of wind turbine clustering;
[0080] B2: Denote the 0° direction of the wind vane of the wind measurement tower corresponding to B1 as the X+ axis direction, and rotate the X+ axis counterclockwise by 90° as the Y+ direction, that is, the initial coordinate system;
[0081] B3: Based on the initial coordinate system, perform a transformation operation on the geographical location coordinates of each wind turbine;
[0082] It should be noted that the geographical location coordinates of each wind turbine obtained through geographical survey are denoted as (X, Y). Calculate the relative coordinate difference of each wind turbine relative to the coordinate system center of the initial coordinate system, specifically expressed as: ΔX = X - X 0 and ΔY = Y - Y 0 where X 0 represents the abscissa corresponding to the coordinate system center of the initial coordinate system through geographical survey, and Y 0 represents the ordinate corresponding to the coordinate system center of the initial coordinate system through geographical survey; through coordinate translation and rotation for transformation operation, obtain the coordinate points (ΔX, ΔY) corresponding to each unit in the initial coordinate system.
[0083] S4: Calculate the propagation wind speed: Calculate based on the first wind power data set of each wind turbine and the corresponding coordinate points to obtain the second wind power data set corresponding to the first wind power data set of each wind turbine.
[0084] Specifically, determine the wind speed line equation according to the wind direction measurement angle of the wind vane of the wind measurement tower determined in S3, and calculate the distance from the corresponding coordinate point of each wind turbine to the wind speed line; calculate the coordinates of the projection points of each coordinate point on the wind tangent line, and judge whether each wind turbine is upstream or downstream of the wind speed incoming flow based on the position relationship between the coordinate point and the wind speed line; calculate the time for the wind to reach each unit; the second wind power data set includes the distance from the corresponding coordinate point of each wind turbine to the wind speed line, the coordinates corresponding to the wind tangent line projection point of each wind turbine, the judgment result of the wind speed incoming flow of each wind turbine, and the time for the wind to reach each wind turbine.
[0085] In a possible implementation manner, obtaining the second wind power data set corresponding to the first wind power data set of each wind turbine includes:
[0086] C1: Determine the wind speed line equation based on the wind vane of the anemometer tower corresponding to the coordinate system center of the initial coordinate system;
[0087] It should be noted that the wind direction measurement angle corresponding to the wind vane of the anemometer tower at the coordinate system center of the initial coordinate system is denoted as θ; according to the trigonometric function relationship, the slope k of the straight line passing through the origin of the coordinate axis in the direction indicated by the wind vane is obtained, and specifically expressed as:
[0088] k = tanθ;
[0089] Assume the wind speed line equation passing through the origin of the coordinate, and specifically expressed as:
[0090] A×ΔX + B×ΔY + C = 0,
[0091] where A, B, and C respectively represent the influence coefficients of the wind speed line equation, and ΔX and ΔY represent the abscissa and ordinate of each coordinate point;
[0092] Based on the condition Determine the actual wind speed line equation based on k and the slope of the straight line passing through the origin of the coordinate axis in the direction indicated by the wind vane;
[0093] In this embodiment, the measurement accuracy of the wind direction measurement angle θ of the anemometer tower wind vane can reach 1°. According to the slope and equation conditions, it is recorded that the influence coefficient B of the wind speed line equation is 1;
[0094] C2: Calculate the distance from the coordinate point corresponding to each wind turbine to the wind speed line;
[0095] It should be noted that based on the abscissa and ordinate ΔX and ΔY of the coordinate point corresponding to each wind turbine, calculate the distance d from each coordinate point to the wind speed line, and specifically expressed as:
[0096]
[0097] where A, B, and C respectively represent the influence coefficients of the wind speed line equation;
[0098] In this embodiment, the distances from the T1 to T7 units to the wind speed line calculated according to the target wind farm are respectively denoted as d1, d2, d3, d4, d5, d6, and d7;
[0099] C3: Calculate the wind tangent projection point corresponding to each wind turbine;
[0100] It should be noted that based on the geographical location (X, Y) of each wind turbine obtained by geographical mapping, calculate the coordinates of the projection point on the wind tangent as (XT, YT), and the solution of its coordinate point is:
[0101]
[0102] Among them, A, B, and C respectively represent the influence coefficients of the wind speed line equation;
[0103] C4: Determine the wind speed inflow of each wind turbine;
[0104] It should be noted that the judgment rules corresponding to the wind speed inflow of each wind turbine are as follows:
[0105] C401: When X - XT ≤ 0, then this wind turbine is located upstream of the wind speed inflow direction, and the farther it is from the wind tangent passing through the coordinate origin, the smaller the probability of being affected by the wake;
[0106] C402: When X - XT > 0, then this wind turbine is located downstream of the wind speed inflow direction, and the closer it is to the wind tangent passing through the coordinate origin, the smaller the probability of being affected by the wake;
[0107] C5: Calculate the time when the wind reaches each wind turbine based on the distance between the coordinate point and the wind speed line and the wind speed inflow;
[0108] In this embodiment, through the formula t = d / v, where v represents the actual wind speed of the wind speed inflow, the time when the wind reaches turbines T1 to T7 is calculated as t1 to t7, and t1 < t2 < t3 < t4 < t5 < t6 < t7.
[0109] S5: Evaluate the mutual influence of wakes: Obtain a multiple wake model, and evaluate the wake influence matrix in multiple directions by passing the second wind power data set through the multiple wake model.
[0110] Specifically, the multiple wake model is a pre - constructed learning model. By inputting the second wind power data set into the multiple wake model, the wake influence factor is obtained, and then the wake influence matrix is obtained. The wake influence matrix is a matrix composed of multiple wake influence factors calculated with each wind turbine as the target comparison wind turbine and the wake influence turbines. The target comparison turbine is the corresponding wind turbine for analyzing the wake influence, and the wake influence turbine is the wind turbine that generates wake influence on the target comparison turbine.
[0111] In a possible implementation manner, evaluating the wake influence matrix in multiple directions includes:
[0112] D1: Obtain the multiple wake model and determine the corresponding parameters of the multiple wake model;
[0113] Further, based on the fact that the wake of each wind turbine diffuses along both sides of the turbine rotor with a fixed diffusion coefficient, the wake diffusion coefficient of the wind farm is determined. The wake diffusion coefficient of the wind farm is the rate of wake diffusion and is determined according to the characteristics of the target wind farm and historical experience data; in this embodiment, the value of the wake diffusion coefficient of the wind farm is 0.05; based on the marked quantity of the main engine manufacturer of the wind turbine equipment under the local air density, the thrust coefficient C of each wind turbine is determined T ; based on the second wind power data set of each wind turbine, the radius r of the turbine rotor is obtained 0 ;
[0114] D2: Calculate the wake influence factors corresponding to multiple situations;
[0115] In a possible implementation manner, calculating the wake influence factor includes:
[0116] Based on the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm and the distance CD between the center of the downstream turbine and the center of the wake effect nm , compare the interference relationship between the target comparison wind turbine and the wake-influenced unit. Denote the target comparison unit as n, the wake-influenced wind turbine as m, and the wake influence factor of wind turbine n on wind turbine m as w nm , where n ∈ N, m ∈ N, and N represents the positions of each wind turbine;
[0117] In this embodiment, the T4 unit is denoted as the target comparison unit, and the influences of the T1, T2, T3, T5, T6, and T7 units on the T4 unit are judged; based on the propagation order of the wind speed time, it is determined that the T1 unit is not affected by the wake;
[0118] D201: When there is no interference relationship between the target comparison unit and the wake-influenced wind turbine, that is, when PD nm ≤0 or CD nm ≥2×r 0 ×α×PD nm , the wake influence factor w1 nm =0;
[0119] D202: When the shadows of the target comparison unit and the wake-influenced wind turbine completely overlap, that is, PD nm ≤0 and CD nm ≥2×r 0 ×α×PD nm , based on the rotor radius r 0 corresponding to the target comparison unit, the thrust coefficient C T corresponding to the target comparison unit, and the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm , calculate the wake influence factor w2 nm , which is specifically expressed as:
[0120]
[0121] Among them, α represents the wake diffusion coefficient of the wind farm;
[0122] D203: There is partial overlap between the shadow of the studied unit and the comparison wind turbine, that is, PD nm > 0 and α × PD nm < CD nm <2 × r 0 × α × PD nm When, based on the rotor radius r 0 corresponding to the target comparison unit, and the distance CD nm between the center of the downstream turbine and the center of the wake effect, calculate the overlapping shadow area influence factor SA, which is specifically expressed as:
[0123]
[0124] Among them, r n (PD nm ) represents the wake radius at PD nm behind the upstream wind turbine, L nm represents the distance from the intersection point of the wake area and the rotor rotation range of the target comparison unit to the center of the wake effect, z nm represents the distance between the two intersection points of the wake area and the rotor rotation range of the target comparison unit;
[0125] Based on the overlapping shadow area influence factor SA, the thrust coefficient C T corresponding to the target comparison unit, the rotor radius r 0 corresponding to the target comparison unit, and the projection distance PD nm of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit, calculate the wake influence factor w3 nm , which is specifically expressed as:
[0126]
[0127] Among them, α represents the wake diffusion coefficient of the wind farm, SA 0 represents the preset overlapping shadow area influence factor, and the preset overlapping shadow area influence factor is the basic quantity for comparing the overlapping shadow area influence factor. In this embodiment, it is set to 1.
[0128] S6: Analyze the prediction deviation: Calculate the deviation between the predicted wind speed and the actual wind speed of each wind turbine to obtain the wind speed prediction deviation value.
[0129] Specifically, denote the measured wind speed of the wind turbine with the shortest time to reach from the anemometer tower corresponding to the coordinate center of the initial coordinate system as v 0 , and calculate the predicted wind speed v j of each wind turbine as the target comparison turbine, specifically expressed as:
[0130]
[0131] where k ∈ K, K represents all the wind turbines on the target wind farm, and the value of w nm is determined by calculating the wake influence factor between each wind turbine as the target comparison turbine and the wake influence turbines.
[0132] Furthermore, calculate the relative deviation between the predicted wind speed of each wind turbine as the target comparison turbine and the measured wind speed of the turbine, that is, the relative deviation value of each wind turbine as the target comparison turbine = (the predicted wind speed of each wind turbine as the target comparison turbine - the measured wind speed of each wind turbine) / the measured wind speed of each wind turbine × 100%; calculate the unit average relative deviation value of each wind turbine as the target comparison turbine according to a preset time scale. The preset time scale is a time period set according to user requirements, and the time period includes but is not limited to hours, days, months, etc. Sort and record according to the wind turbine number and time interval to generate a relative prediction deviation sequence; perform a normal distribution on the relative prediction deviation sequence, and select the wind speed prediction deviation value corresponding to the maximum probability value, denoted as △φ.
[0133] S7: Predict the wind speed of the turbine: Based on the wake influence matrix in multiple directions and the prediction deviation correction value, predict the wind speed of each wind turbine to obtain the predicted wind speed of each wind turbine.
[0134] Specifically, add the initial predicted wind speed v j of each wind turbine as the target comparison turbine to the prediction deviation correction value △φ to obtain the predicted wind speed v' j of each wind turbine, that is, v' j = v j + △φ.
[0135] S8: Predict the ultra-short-term power of the wind turbine: Based on the actual power generation efficiency of the wind turbine and the wind speed prediction deviation value, perform a prediction operation on the short-term time interval corresponding to a single wind turbine to obtain the ultra-short-term power of the wind turbine.
[0136] Specifically, based on the actual power generation efficiency of each wind turbine and the wind speed prediction deviation value, calculate the ultra-short-term predicted power of the wind turbine at time T; sum the ultra-short-term predicted power of the wind turbine corresponding to time T within the prediction time interval to obtain the ultra-short-term prediction result of the wind turbine.
[0137] In a possible implementation, obtaining the ultra-short-term wind turbine power includes: calculating the ultra-short-term predicted power of the wind turbine based on the predicted wind speed and actual power generation efficiency of each wind turbine unit;
[0138] Based on the actual power generation efficiency P corresponding to the wind turbine at time T and the predicted wind speed v′ of each wind turbine unit j calculate the ultra-short-term predicted power C corresponding to a single wind turbine p Specifically expressed as:
[0139]
[0140] where ρ represents the air density of the target wind farm, and A represents the swept area of the wind turbine unit;
[0141] Perform a summation operation on the ultra-short-term predicted power corresponding to a single wind turbine within the prediction time interval, and the prediction time interval is a prediction time interval centered on time T divided according to a preset time scale.
[0142] Secondly: In the attached drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments are involved. For other structures, reference can be made to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other;
[0143] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A wind turbine power prediction method based on a wake model, characterized in that: include: S1: Collecting target wind farm data: In response to a pre-installed wind tower in the target wind farm, collecting a first wind power data group of each wind turbine generator set through an information collection device; S2: performing clustering operation on wind turbines: performing cluster analysis on the first wind power data group of each wind turbine to obtain an optimal set of wind turbine clusters; S3: Establishing the coordinate system of clustering clusters: constructing an initial coordinate system for the wind tower of the target wind farm, and performing a transformation operation on each unit corresponding to the optimal set of wind turbine clusters, the transformation operation is used to obtain the corresponding coordinate point of each unit in the initial coordinate system; S4: Calculate the propagation wind speed: perform calculation based on the first wind power data group of each wind turbine generator set and the corresponding coordinate point to obtain a second wind power data group corresponding to the first wind power data group of each wind turbine generator set; S5: Evaluate mutual influence of wakes: obtain a multiple wake model, and use the multiple wake model to evaluate the wake influence matrix of multiple directions of the second wind power data set; S6: Analyze the prediction deviation: Calculate the deviation between the predicted wind speed and the actual wind speed of each wind turbine to obtain the wind speed prediction deviation value; S7: Predict wind speed of each wind turbine: predict the wind speed of each wind turbine based on the wake influence matrix of multiple directions and the prediction deviation correction value to obtain the predicted wind speed of each wind turbine; S8: Predicting ultra-short-term wind turbine power: Based on the actual power generation efficiency of the wind turbine and the wind speed prediction deviation value, a prediction operation is performed on the short-term time interval corresponding to the single wind turbine to obtain the ultra-short-term wind turbine power.
2. The wind turbine power prediction method based on the wake model according to claim 1 is characterized in that: The step S2, obtaining the optimal set of wind turbine component clusters, specifically includes: The clustering condition is that the minimum distance between particles in multiple clusters is greater than 10 times the diameter of the wind wheel; The height difference between the particles and the hub within the cluster is less than or equal to 100 meters.
3. The wind turbine power prediction method based on the wake model according to claim 1 is characterized in that: S3, constructing an initial coordinate system for the wind tower of the target wind farm, specifically includes: B1: In the optimal set of wind turbine clusters, obtain the wind tower with the closest calculated distance to each wind turbine; B2: The 0° direction of the wind vane of the wind tower corresponding to B1 is recorded as the X+ axis direction, and the X+ axis is rotated 90° counterclockwise as the Y+ direction, which is the initial coordinate system; B3: Based on the initial coordinate system, the geographical location coordinates of each wind turbine are transformed.
4. The wind turbine power prediction method based on the wake model according to claim 1 is characterized in that: The step S4, obtaining the second wind power data group corresponding to the first wind power data group of each wind turbine generator set, specifically includes: the judgment rule corresponding to the wind speed flow of each wind turbine generator set is as follows: C401: When X-XT≤0, the wind turbine is located upstream in the wind flow direction; C402: When X-XT>0, the wind turbine is located downstream in the wind flow direction.
5. The wind turbine power prediction method based on the wake model according to claim 1 is characterized in that: The wake influence matrix S5 is a matrix composed of wake influence factors calculated by the target comparison wind turbine and the wake influence wind turbine, with each wind turbine set being used as the target comparison wind turbine; The target comparison unit is the corresponding wind turbine unit for analyzing the wake influence; The wake-affecting wind turbine is a wind turbine that produces a wake impact on the target comparison machine.
6. The wind turbine power prediction method based on the wake model according to claim 5 is characterized in that: S5, evaluating the wake impact matrix in multiple directions, specifically includes: D1: Obtain the multiple wake model and determine the corresponding parameters of the multiple wake model; D2: Calculate the wake impact factors corresponding to various situations.
7. The wind turbine power prediction method based on the wake model according to claim 5 is characterized in that: S5, evaluating the wake impact matrix in multiple directions, specifically includes: Based on the wind speed line translation to the projection distance PD on the wind speed line passing through the target comparison unit coordinate position nm The distance CD between the center of the downstream turbine and the center of the wake effect nm , the interference relationship between the target comparison fan and the wake affecting unit, the target comparison unit is n, the wake affecting fan is m, and the wake affecting factor of fan n on fan m is w nm , where n∈N, m∈N, N represents the location of each wind turbine; D201: When the target comparison unit has no interference relationship with the wake-affected wind turbine, that is, when PD nm ≤0 or CD nm ≥2×r0×α×PD nm When the wake influence factor w1 nm =0; D202: When the target comparison unit and the shadow of the wind turbine affected by the wake completely overlap, that is, PD nm ≤0 and CD nm ≥2×r0×α×PD nm When the rotor radius r0 corresponding to the target comparison unit and the thrust coefficient C corresponding to the target comparison unit are T , and the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm , calculate the wake influence factor w2 nm , specifically expressed as: Among them, α represents the wind farm wake diffusion coefficient; D203: The shadow of the research unit and the comparison fan partially overlap, that is, PD nm >0 and α×PD nm <CD nm <2×r0×α×PD nm When comparing the rotor radius r0 of the target unit and the distance CD between the center of the downstream turbine and the center of the wake effect, nm , calculate the overlapping shadow area impact factor SA, which is specifically expressed as: Among them, r n (PD nm ) represents the PD after the upstream wind turbine nm The wake radius at nm It is expressed as the distance from the intersection of the wake area and the target unit rotor rotation range to the center of the wake effect, z nm It is expressed as the distance between the two intersection points of the wake area and the target comparison unit rotor rotation range; Based on the overlap shadow area influence factor SA and the thrust coefficient C corresponding to the target comparison unit T , the rotor radius r0 corresponding to the target comparison unit, and the projection distance PD of the wind speed line translated to the wind speed line passing through the coordinate position of the target comparison unit nm , calculate the wake influence factor w3 nm , specifically expressed as: Among them, α represents the wake diffusion coefficient of the wind farm, and SA0 represents the preset overlapping shadow area influence factor.
8. The wind turbine power prediction method based on the wake model according to claim 1 is characterized in that: The step S6, obtaining the wind speed prediction deviation value, specifically includes: Calculate the relative deviation between the predicted wind speed of each wind turbine as the target comparison unit and the measured wind speed of the unit; The relative deviation value of each wind turbine set as the target comparison set is used to calculate the unit average relative deviation value according to a preset time scale, and the preset time scale is a time period set according to user needs; Sort the records by wind turbine group number and time interval to generate a relative prediction deviation sequence; The relative prediction deviation sequence is normally distributed, and the wind speed prediction deviation value corresponding to the maximum probability value is selected.