A wind speed prediction method and system for complex terrain wind farm

Abstract

The application provides a wind speed prediction method and system for a wind farm in complex terrain, and belongs to the field of wind farm prediction. The wind speed prediction method for the wind farm in complex terrain comprises the following steps: obtaining historical monitoring data, climate temperature data and wind farm geographic data in the wind farm; analyzing wind speed prediction influence data according to the climate temperature data and the wind farm geographic data; determining basic prediction wind speed data according to the historical monitoring data; and calculating comprehensive wind speed prediction data according to the wind speed prediction influence data and the basic prediction wind speed data. By predicting the wind speed of the wind farm, the operation parameters of the wind turbine can be adjusted in advance, the possibility of failure in the cooperation between the wind turbine and the wind speed can be reduced, and the power generation efficiency of the wind turbine can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wind field prediction, in particular to a wind speed prediction method and system for complex terrain of wind farm. BACKGROUND

[0002] The wind farm refers to a large-scale power generation place where wind energy is converted into electric energy by concentrating multiple wind turbines, i.e. wind machines. It usually includes core equipment such as wind machines and auxiliary systems such as monitoring equipment and wind towers. Generally, wind farms include two types, namely offshore wind farms and onshore wind farms. The onshore wind farm is affected by complex terrain. Therefore, it is necessary to predict the wind speed in advance in order to improve the power generation efficiency for the onshore wind farm. SUMMARY

[0003] The present application provides a wind speed prediction method and system for complex terrain of wind farm, which can predict the wind speed, adjust the wind turbine operation parameters in advance, so that the wind turbine runs in the best power generation efficiency interval under different wind speeds, and reduces the loss.

[0004] In a first aspect, the present application provides a wind speed prediction method for complex terrain of wind farm. The method comprises:

[0005] Obtaining historical monitoring data, climate temperature data and wind farm geographic data in the wind farm; the historical monitoring data includes historical wind speed data; the historical wind speed data includes multiple historical average wind speed data carrying historical time labels; the historical time labels include year, month, day and time period data; the wind farm geographic data includes wind tower position data and hilly terrain data;

[0006] Analyzing wind speed prediction influence data according to the climate temperature data and the wind farm geographic data; the wind speed prediction influence data is associated with the climate temperature data, the wind tower position data and the hilly terrain data;

[0007] Determining basic prediction wind speed data according to the historical monitoring data; the basic prediction wind speed data is associated with the historical average wind speed data;

[0008] Calculating comprehensive wind speed prediction data according to the wind speed prediction influence data and the basic prediction wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction influence data and the basic prediction wind speed data.

[0009] Further, the analyzing wind speed prediction influence data according to the climate temperature data and the wind farm geographic data; the wind speed prediction influence data is associated with the climate temperature data, the wind tower position data and the hilly terrain data comprises:

[0010] analyzing temperature influence data according to the climate temperature data; the climate temperature data comprises temperature data, seasonal data and climate data; the temperature influence data is associated with the temperature data, seasonal data and climate data;

[0011] analyzing terrain undulation data, roughness data, obstacle distribution data and special terrain data according to the hilly terrain data and the wind tower location data;

[0012] analyzing terrain influence data according to the terrain undulation data, roughness data, obstacle distribution data and special terrain data; the terrain influence data is associated with the terrain undulation data, roughness data and obstacle distribution data;

[0013] analyzing wind speed prediction influence data according to the temperature influence data and the terrain influence data; the wind speed prediction influence data is associated with the temperature influence data and the terrain influence data.

[0014] Further, the determining of the basic prediction wind speed data according to the historical monitoring data comprises:

[0015] obtaining task time labels of the wind speed prediction task; the task time labels comprise year, month, day and time period data;

[0016] screening a first reference group according to the prediction task time data; the first reference group comprises a plurality of first reference wind speed data; the first reference wind speed data is historical average wind speed data with part of historical time labels and part of task time labels being the same; the part of historical time labels and the part of task time labels are month, day and time period data;

[0017] screening a second reference group according to the prediction task time data; the second reference group comprises a plurality of second reference wind speed data; the second reference wind speed data is historical average wind speed data with part of historical time labels and part of task time labels being the same; the part of historical time labels and the part of task time labels are month and time period data;

[0018] calculating first wind speed basic quantity data based on the first reference group; calculating second wind speed basic quantity data based on the second reference group; determining the basic prediction wind speed data according to the first wind speed basic quantity data and the second wind speed basic quantity data; the basic prediction wind speed data is associated with the first wind speed basic quantity data and the second wind speed basic quantity data.

[0019] Further, the calculation method of the basic prediction wind speed data comprises:

[0020] if the first reference group comprises n first reference wind speed data, the i-th first reference wind speed data is

[0021] Supposing the second reference group comprises m second reference wind speed data, the jth second reference wind speed data is

[0022] FP ws = B ws × K1+ D ws × (1-K1)

[0023]

[0024] In the formula, FP ws is the basic prediction wind speed data, B ws is the first wind speed basic quantity data, D ws is the second wind speed basic quantity data; K1 is the pre-design calculation weight.

[0025] Further, the calculation method of the pre-design calculation weight comprises:

[0026] calculating a first variance of the first reference group based on all the first reference wind speed data contained in the first reference group;

[0027] calculating a second variance of the second reference group based on all the second reference wind speed data contained in the second reference group;

[0028]

[0029] In the formula, K1 is the pre-design calculation weight, S1 is the first variance, and S2 is the second variance; the first variance is negatively correlated with the pre-design calculation weight, and the second variance is positively correlated with the pre-design calculation weight.

[0030] The second aspect of the present application provides a wind speed prediction system for a wind farm in complex terrain. The system comprises:

[0031] an acquisition module configured to acquire historical monitoring data, climate temperature data, and wind farm geographic data in the wind farm; the historical monitoring data comprises historical wind speed data; the historical wind speed data comprises a plurality of historical average wind speed data carrying historical time labels; the historical time labels comprise year, month, day, and time period data; the wind farm geographic data comprises wind measurement tower position data and hilly terrain data;

[0032] an analysis module configured to analyze wind speed prediction influence data according to the climate temperature data and the wind farm geographic data; the wind speed prediction influence data is associated with the climate temperature data, the wind measurement tower position data, and the hilly terrain data;

[0033] a determination module configured to determine a basic prediction wind speed data according to the historical monitoring data; the basic prediction wind speed data is associated with the historical average wind speed data;

[0034] a calculating module configured to calculate comprehensive wind speed prediction data according to the wind speed prediction influence data and the basic prediction wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction influence data and the basic prediction wind speed data.

[0035] Further, the analyzing module is further configured to analyze wind speed prediction influence data according to the climate temperature data and the wind farm geographical data; the wind speed prediction influence data is associated with the climate temperature data, wind tower location data and hilly terrain data, and comprises:

[0036] analyze temperature influence data according to the climate temperature data; the climate temperature data comprises temperature data, seasonal data and climate data; the temperature influence data is associated with the temperature data, seasonal data and climate data;

[0037] analyze terrain undulation data, roughness data, obstacle distribution data and special terrain data according to the hilly terrain data and wind tower location data;

[0038] analyze terrain influence data according to the terrain undulation data, roughness data, obstacle distribution data and special terrain data; the terrain influence data is associated with the terrain undulation data, roughness data and obstacle distribution data;

[0039] analyze wind speed prediction influence data according to the temperature influence data and the terrain influence data; the wind speed prediction influence data is associated with the temperature influence data and the terrain influence data.

[0040] Further, the determining module is further configured to determine basic prediction wind speed data according to the historical monitoring data; the basic prediction wind speed data is associated with the historical average wind speed data, and comprises:

[0041] obtain a task time label of a wind speed prediction task; the task time label comprises year, month, day and time period data;

[0042] screen a first reference group according to the prediction task time data; the first reference group comprises a plurality of first reference wind speed data; the first reference wind speed data is historical average wind speed data with part of historical time labels and part of task time labels being the same; the part of historical time labels and the part of task time labels are month, day and time period data;

[0043] Filtering a second reference group according to the predicted task time data; the second reference group comprises a plurality of second reference wind speed data; the second reference wind speed data is historical average wind speed data with part of historical time labels and part of task time labels being the same; the part of historical time labels and the part of task time labels are both month and time period data;

[0044] Calculating first wind speed basis data based on the first reference group; calculating second wind speed basis data based on the second reference group; determining basis predicted wind speed data according to the first wind speed basis data and the second wind speed basis data; the basis predicted wind speed data is associated with the first wind speed basis data and the second wind speed basis data.

[0045] Further, the determining module is further configured to, the calculation method of the basis predicted wind speed data comprises:

[0046] Supposing that the first reference group comprises n first reference wind speed data, the i-th first reference wind speed data is

[0047] Supposing that the second reference group comprises m second reference wind speed data, the j-th second reference wind speed data is

[0048] FP ws =B ws ×K1+D ws ×(1-K1)

[0049]

[0050] In the formula, FP ws is the basis predicted wind speed data, B ws is the first wind speed basis data, D ws is the second wind speed basis data; K1 is a pre-designed calculation weight.

[0051] Further, the determining module is further configured to, the calculation method of the pre-designed calculation weight comprises:

[0052] Calculating a first variance of the first reference group based on all first reference wind speed data contained in the first reference group;

[0053] Calculating a second variance of the second reference group based on all second reference wind speed data contained in the second reference group;

[0054]

[0055] In the formula, K1 is the pre-designed calculation weight, S1 is the first variance, and S2 is the second variance; the first variance is negatively correlated with the pre-designed calculation weight, and the second variance is positively correlated with the pre-designed calculation weight.

[0056] By adopting the above technical solution, wind speed prediction for wind farms is achieved. Basic predicted wind speed data is determined based on historical monitoring data, while wind speed prediction impact data is determined based on climate temperature data and wind farm geographical data. Finally, the comprehensive wind speed prediction data is calculated by combining the basic predicted wind speed data and the wind speed prediction impact data. Based on this comprehensive wind speed prediction data, a pre-adjustment plan for wind turbines within the corresponding task time is formulated, and the operating parameters of the wind turbines are pre-adjusted to better match the wind speed for power generation, thereby improving the power generation efficiency of the wind farm and reducing losses.

[0057] It should be understood that the description in the Summary Section is not intended to limit the key or essential features of the embodiments of this application, nor is it intended to restrict the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0058] The above and other features, advantages, and aspects of the embodiments of this application will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0059] Figure 1 A flowchart of a wind speed prediction method for wind farms in complex terrain, as shown in an embodiment of this application, is illustrated.

[0060] Figure 2 A block diagram of a wind speed prediction system for wind farms in complex terrain, according to an embodiment of this application, is shown. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0062] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0063] This application provides a method and system for predicting wind speed in wind farms with complex terrain, which is capable of...

[0064] Firstly, this application provides a method for predicting wind speed in wind farms with complex terrain. (Reference) Figure 1 The specific steps included in the method are as follows.

[0065] Step S110: Obtain historical monitoring data, climate temperature data, and wind farm geographic data within the wind farm; the historical monitoring data includes historical wind speed data; the historical wind speed data includes multiple historical average wind speed data carrying historical time tags; the historical time tags include year, month, day, and time period data; the wind farm geographic data includes wind measurement tower location data and hilly terrain data.

[0066] In this scheme, the specific complex terrain refers to hilly terrain. In hilly terrain, there are differences in altitude, such as regional height differences and terrain undulation; special terrain features such as ridges and valleys. Such terrain features will affect the wind speed measurement of the wind farm by the wind measuring tower. In order to ensure the accuracy of wind speed prediction in wind farms in hilly terrain, it is necessary to consider data such as terrain features, meteorological seasons, and historical wind speeds, so as to make the adjustment of wind turbine operating parameters more accurate and improve the power generation efficiency of wind farms.

[0067] It is understandable that the wind speed prediction in this scheme refers to the average wind speed of the wind farm, not the wind speed of a single wind turbine. Conventional wind speed measurement methods can be obtained through anemometer towers. The average wind speed data (including 10-minute average wind speed, measured once per second; hourly average wind speed; and maximum wind speed, measured once every 3 seconds) detected by the anemometer tower's wind speed measuring device is then processed by regression analysis to obtain the average wind speed of the wind farm. The average wind speed of the wind farm reflects the overall level of wind energy resources in the area. It is an important indicator for assessing the power generation potential of a wind farm. The higher the average wind speed, the richer the available wind energy resources of the wind farm, and theoretically, the more electricity it can generate.

[0068] In this solution, the historical average wind speed data with historical time tags mentioned above generally includes time periods of one hour. The historical time tag is the year, month, date, and specific time period, such as the historical average wind speed data from 8 to 9 a.m. on April 22, 2012. Each time period is a wind speed measurement cycle. The historical average wind speed data here is the processed average wind speed, not the wind speed data directly collected by the wind tower.

[0069] Understandably, the location data of the side towers and the hilly terrain data here can be obtained by acquiring high-resolution remote sensing images through satellite remote sensing, and then specific analysis can be performed based on the remote sensing images to obtain the corresponding terrain and location data; the historical monitoring data here can be obtained through the data upload cloud platform corresponding to the wind farm, and the wind farm will upload the collected wind farm-related data to the corresponding cloud platform for storage.

[0070] Step S120: Analyze the wind speed prediction impact data based on the climate temperature data and the wind farm geographical data; the wind speed prediction impact data is associated with the climate temperature data, the wind measurement tower location data, and the hilly terrain data.

[0071] Specifically, temperature impact data is analyzed based on the climate and temperature data; the climate and temperature data includes temperature data, seasonal data, and climate data; the temperature impact data is associated with the temperature data, seasonal data, and climate data; terrain relief data, roughness data, obstacle distribution data, and special terrain data are analyzed based on the hilly terrain data and wind tower location data; terrain impact data is analyzed based on the terrain relief data, roughness data, obstacle distribution data, and special terrain data; the terrain impact data is associated with the terrain relief data, roughness data, and obstacle distribution data; wind speed prediction impact data is analyzed based on the temperature impact data and the terrain impact data; the wind speed prediction impact data is associated with the temperature impact data and the terrain impact data.

[0072] Specifically, the analysis of temperature impact data based on the climate and temperature data includes: determining the seasonal and climate labels of the wind farm at the current moment based on seasonal and climate data; obtaining the task time label for the wind speed prediction task, which includes year, month, day, and time period data; obtaining the first preset grid temperature data of the wind farm in the previous wind speed measurement cycle at the current moment; obtaining the historical preset grid temperature data of the wind farm in the historical wind speed measurement cycle, which has the same seasonal label, climate label, and task time label as the previous wind speed measurement cycle; and determining the cell temperature data of multiple different wind speed measurement cycles for a preset number of cells within a preset range around each meteorological tower based on the first preset grid temperature data and the historical preset grid temperature data. The following steps are taken: Using the cell temperature data of the same cell in the previous wind speed measurement cycle as the baseline temperature data, and using the cell temperature data of the same cell in other wind speed measurement cycles as the sample temperature data, a baseline temperature data and multiple sample temperature data are obtained; the sample average temperature data is calculated, which is the average of the multiple sample temperature data; the temperature difference data between the baseline temperature data and the sample average temperature data is calculated; the same processing method is applied to the cell temperature data of each cell to obtain multiple temperature difference data; the variance of the temperature difference data is calculated, which is the sample variance; the average of the temperature difference data is calculated, which is the sample average; and the temperature influence data is calculated based on the sample variance and the sample average.

[0073] Specifically, the calculation methods for temperature effect data include:

[0074]

[0075] In the formula, T im For data on the effect of temperature, S a This is the sample mean. S is the preset average value. v Here, the sample variance value is used. The smaller the difference between the sample mean and the preset mean, the closer the sample mean is to the preset mean, which means that the temperature in the previous wind speed measurement period is close to the historical average temperature. The smaller the sample variance value, the smaller the fluctuation of the temperature difference data and the more stable it is, which means that the temperature in the previous wind speed measurement period of each cell is close to the historical average temperature. The smaller the above two influencing factors are, the greater the influence of temperature on the data. The temperature influence data here represents the contribution of temperature to wind speed prediction.

[0076] Understandably, the preset grid temperature data here refers to dividing the wind farm into multiple identical grids and performing temperature tests on each grid to obtain the corresponding grid temperature data. The purpose of this is to capture the spatial heterogeneity of temperature within the wind farm. When dividing the grid, equally spaced regular grids are used, with square grids being preferred. The geographic coordinates of the grids can be determined using UTM projection (such as the WGS84 coordinate system, with the projection zone determined based on the longitude of the wind farm). The overall wind farm grid contains multiple cells. By determining the latitude and longitude center coordinates of each cell, the data is then imported into a GIS platform (such as QGIS) and overlaid with a wind farm topographic map and wind... The geographic information of each cell can be determined by the location of the wind turbine, the location of the wind measurement tower, and the distribution of the meteorological station. The temperature of each cell can be obtained by combining sensors, UAV temperature measurement, and satellite remote sensing data. Temperature data of the cell can be obtained by placing temperature sensors in the center or four corners of the cell, or by using remote sensing images. Specifically, when dividing the wind farm grid, a 1km resolution can be used. "1km resolution" usually refers to the smallest spatial unit used in spatial analysis or data acquisition, that is, the side length of the cell is 1km, that is, the actual ground area corresponding to each cell is 1km×1km.

[0077] The hilly terrain data here can be obtained through satellite remote sensing imagery. The terrain relief data is the elevation difference between the highest and lowest points in a specific area. By using satellite remote sensing imagery, the elevation data of the highest and lowest points in the wind farm can be determined, and then the difference between the two can be calculated to obtain the terrain relief data. Similarly, the elevation data of each wind measurement tower location can also be obtained through satellite remote sensing imagery.

[0078] By using satellite imagery classification, field surveys, LiDAR point clouds, or high-precision maps, the overall roughness data and obstacle distribution data of the wind farm can be determined. Roughness data reflects the resistance of the surface to airflow and is determined by the surface type. In this scheme, the roughness data is the roughness length; this roughness length is a characteristic parameter representing the degree of surface roughness, measured in meters (m). The smaller z0 is, the smoother the surface, and vice versa. Specific acquisition methods include using a standard roughness length table or satellite remote sensing imagery. The standard roughness length table is consulted based on the surface type (e.g., grassland, farmland, forest, urban area). (Example: grassland z0≈0.0) For a distance of 1-0.1m (for forests, z0≈1-5m), z0 is estimated using a model by combining satellite imagery (such as vegetation cover and land use type) or high-precision topographic data (such as LiDAR). Here, the height refers to the height of the wind speed measurement point. The frictional velocity here reflects the shear resistance of the surface to the wind and is a measure of the momentum exchange between the surface and the atmospheric boundary layer. It can be calculated by measuring high-frequency pulsating wind speed with an ultrasonic anemometer, calculating Reynolds stress, and then calculating the frictional velocity based on the Reynolds stress. It is understood that the specific implementation process and data acquisition methods given above are all conventional methods and can be obtained through conventional technologies, which will not be elaborated here.

[0079] The specific topographic data here includes the locations of ridges / valleys in hilly terrain. Ridges refer to long, narrow, elevated areas in hilly terrain, with a convex outline and steep slopes on both sides. When airflow passes over a ridge, the wind speed increases due to the terrain lifting and streamline contraction, especially at the ridge top, where the wind speed may be 20% to 40% higher than the surrounding flat areas, affecting the stability of wind speed prediction. Valleys, on the other hand, are low-lying areas between adjacent ridges, with a concave outline, usually covered by rivers or vegetation. If a valley is long and narrow and runs in the same direction as the prevailing wind, it may create a funnel effect, significantly increasing wind speed (e.g., when a north-south oriented valley corresponds to a north or south wind). If it is perpendicular to the wind direction, the wind speed may decrease (due to obstruction by the mountains on both sides), which also affects the stability of wind speed prediction.

[0080] Historical wind direction data is acquired, including wind direction data from the anemometer towers for each wind speed measurement cycle. Based on the wind direction data and obstacle distribution data, the elevation difference between each anemometer tower and the specific terrain along the same wind direction, the number of obstacles, and the volume of each obstacle are determined. When calculating the elevation difference, the difference is calculated between the elevation data of the anemometer tower's location and the elevation data of the specific terrain. The elevation of the anemometer tower's location is based on the surface elevation data of that location. If the specific terrain is a ridge, the elevation data of the highest point of the ridge is used; if the specific terrain is a valley, the elevation data of the lowest point of the valley is used. The number and volume of obstacles can be obtained by acquiring and processing images from satellite remote sensing imagery. "Same wind direction" here refers to wind blowing from the specific terrain towards the anemometer tower, and the straight-line distance between the anemometer tower and the specific terrain should not exceed a preset straight-line distance.

[0081] Specifically, the calculation method for terrain impact data includes: assuming n wind measurement towers, the roughness data of the a-th wind measurement tower is... The altitude of the a-th meteorological tower is h. a There are m special terrain features corresponding to the a-th wind measuring tower in the same wind direction, and the altitude of the b-th special terrain feature is h. ab There are q obstacles between the a-th wind measuring tower and the corresponding b-th special terrain feature in the same wind direction, and the obstacle volume of the p-th obstacle is V. abp ;

[0082]

[0083] F1 = f1×k1 + f2×k2 + f3×k3

[0084]

[0085]

[0086] In the formula, TID represents the terrain impact data, F1, f1, f2, and f3 are the first intermediate function, the second intermediate function, the third intermediate function, and the fourth intermediate function, respectively; k1, k2, and k3 are the preset first sub-weight coefficient, the preset second sub-weight coefficient, and the preset third sub-weight coefficient, respectively; and k1 + k2 + k3 = 1; specifically, k1, k2, and k3 can be taken as 0.3, 0.3, and 0.4, respectively.

[0087] After obtaining the temperature impact data and the terrain impact data, the wind speed prediction impact data is calculated by weighted summation of the temperature impact data and the terrain impact data; the weight of each is 0.5, and the sum of their corresponding weights is 1.

[0088] Step S130: Determine the basic predicted wind speed data based on the historical monitoring data; the basic predicted wind speed data is associated with the historical average wind speed data.

[0089] Specifically, the process involves: acquiring the task time tag for wind speed prediction; the task time tag including year, month, day, and time period data; filtering a first reference group based on the prediction task time data; the first reference group including multiple first reference wind speed data; the first reference wind speed data being historical average wind speed data with some historical time tags and some task time tags being the same; both the historical time tags and the task time tags being month, day, and time period data; filtering a second reference group based on the prediction task time data; the second reference group including multiple second reference wind speed data; the second reference wind speed data being historical average wind speed data with some historical time tags and some task time tags being the same; both the historical time tags and the task time tags being month and time period data; calculating first basic wind speed data based on the first reference group; calculating second basic wind speed data based on the second reference group; determining basic predicted wind speed data based on the first and second basic wind speed data; and the basic predicted wind speed data being associated with both the first and second basic wind speed data.

[0090] The calculation method for the basic predicted wind speed data specifically includes: assuming the first reference group includes n first reference wind speed data points, then the i-th first reference wind speed data point is... Let the second reference group include m second reference wind speed data points, then the j-th second reference wind speed data point is...

[0091] FP ws =B ws ×K1+D ws ×(1-K1)

[0092]

[0093] In the formula, FP ws Based on the predicted wind speed data, B ws For the first basic wind speed data, D ws This is the second basic wind speed data; K1 is the preset calculation weight.

[0094] Furthermore, the calculation method for the preset calculation weights specifically includes: calculating the first variance of the first reference group based on all the first reference wind speed data contained in the first reference group; and calculating the second variance of the second reference group based on all the second reference wind speed data contained in the second reference group.

[0095]

[0096] In the formula, K1 is the preset calculation weight, S1 is the first variance, and S2 is the second variance; the first variance is negatively correlated with the preset calculation weight, and the second variance is positively correlated with the preset calculation weight.

[0097] It is understandable that the task time label included in the wind speed prediction task here refers to the time period during which the wind farm needs to perform average wind speed prediction. For example, if the current time is 10:00 AM, the average wind speed of the wind farm needs to be predicted for the period from 11:00 AM to 12:00 PM. The first reference group refers to historical average wind speed data where the month, date, and time period are the same as the corresponding month, date, and time period of the task time label, i.e., historical average wind speed data for the same month, day, and time period of each year. The second reference group refers to historical average wind speed data where the month and time period are the same as the corresponding month and time period of the task time label, i.e., historical average wind speed data for the same time period of each day within the same month of each year. By calculating the average of all historical average wind speed data corresponding to the above two reference groups, and then weighted summing the two averages, the basic predicted wind speed data can be obtained. The first and second basic wind speed data are both positively correlated with the basic predicted wind speed data.

[0098] The calculation methods for the first and second variances here both adopt the conventional variance calculation method. The first variance is obtained by calculating the variance of all first reference wind speed data in the first reference group, and the second variance is obtained by calculating the variance of all second reference wind speed data in the second reference group. Variance reflects the stability of data. The smaller the variance, the more stable the data. Therefore, in this scheme, for the basic predicted wind speed data, the more stable the data, the greater the contribution of the average value of all first reference wind speed data corresponding to the first wind speed basic quantity data to the basic predicted wind speed data. This is reflected in the formula as a higher weight, and vice versa. Therefore, the larger the first variance, the lower the data stability in the first reference group, the lower the contribution of the first wind speed basic quantity data corresponding to the first reference group, the lower the weight, and the smaller K1. Conversely, the smaller the variance, the larger K1.

[0099] Step S140: Calculate comprehensive wind speed prediction data based on the wind speed prediction impact data and the basic predicted wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction impact data and the basic predicted wind speed data.

[0100] Specifically, the comprehensive wind speed forecast data = wind speed forecast impact data * basic forecast wind speed data; after calculating the comprehensive wind speed forecast data, the operating parameters of the wind turbine can be adjusted in advance based on the predicted wind speed data, thereby improving the power generation efficiency of the wind farm.

[0101] In this scheme, basic predicted wind speed data is obtained by calculating the average, variance, and summation of historical average wind speed data. Then, the influence data of wind speed prediction is analyzed by combining the hilly terrain, the location of the wind measurement tower, geographical factors, and meteorological seasonal factors. Finally, the basic predicted wind speed data and the wind speed prediction influence data are combined to obtain more realistic comprehensive wind speed prediction data. Based on this comprehensive wind speed prediction data, wind speed data for a certain period of time in the future can be obtained in advance. Then, based on this wind speed data, the operating parameters of each wind turbine can be adjusted accordingly, so that the wind turbine can better match the corresponding wind speed and generate more electricity, thereby reducing losses and improving the overall power generation efficiency of the wind farm.

[0102] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0103] Secondly, this application provides a wind speed prediction system for wind farms in complex terrain; such as Figure 2 As shown, the system includes an acquisition module 210, used to acquire historical monitoring data, climate temperature data, and wind farm geographical data within the wind farm; the historical monitoring data includes historical wind speed data; the historical wind speed data includes multiple historical average wind speed data carrying historical time tags; the historical time tags include year, month, day, and time period data; the wind farm geographical data includes wind measurement tower location data and hilly terrain data;

[0104] Analysis module 220 is used to analyze wind speed prediction impact data based on the climate temperature data and the wind farm geographical data; the wind speed prediction impact data is associated with the climate temperature data, the wind measurement tower location data, and the hilly terrain data;

[0105] The determination module 230 is used to determine basic predicted wind speed data based on the historical monitoring data; the basic predicted wind speed data is associated with the historical average wind speed data.

[0106] The calculation module 240 is used to calculate comprehensive wind speed prediction data based on the wind speed prediction impact data and the basic predicted wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction impact data and the basic predicted wind speed data.

[0107] Furthermore, the analysis module 220 is further configured to analyze wind speed prediction impact data based on the climate temperature data and the wind farm geographic data; the wind speed prediction impact data, which is associated with the climate temperature data, the meteorological tower location data, and the hilly terrain data, includes:

[0108] The temperature impact data is analyzed based on the climate and temperature data; the climate and temperature data includes temperature data, seasonal data, and climate data; the temperature impact data is correlated with the temperature data, seasonal data, and climate data.

[0109] Based on the hilly terrain data and wind tower location data, analyze the terrain relief data, roughness data, obstacle distribution data, and special terrain data;

[0110] The terrain impact data is analyzed based on the terrain relief data, roughness data, obstacle distribution data, and special terrain data; the terrain impact data is correlated with the terrain relief data, roughness data, and obstacle distribution data.

[0111] The wind speed prediction impact data is analyzed based on the temperature impact data and the terrain impact data; the wind speed prediction impact data is correlated with the temperature impact data and the terrain impact data.

[0112] Furthermore, the determining module 230 is further configured to determine the basic predicted wind speed data based on the historical monitoring data; the basic predicted wind speed data being associated with the historical average wind speed data includes:

[0113] Obtain the task time tag for the wind speed prediction task; the task time tag includes year, month, day, and time period data;

[0114] A first reference group is selected based on the predicted task time data; the first reference group includes multiple first reference wind speed data; the first reference wind speed data is historical average wind speed data with some historical time tags and some task time tags being the same; the some historical time tags and some task time tags are both month-day and time period data;

[0115] A second reference group is selected based on the predicted task time data; the second reference group includes multiple second reference wind speed data; the second reference wind speed data is historical average wind speed data with some historical time tags and some task time tags being the same; the some historical time tags and some task time tags are both monthly and time period data;

[0116] First wind speed baseline data is calculated based on a first reference group; second wind speed baseline data is calculated based on a second reference group; baseline predicted wind speed data is determined based on the first and second wind speed baseline data; the baseline predicted wind speed data is associated with the first and second wind speed baseline data.

[0117] Furthermore, the determining module 230 is further configured such that the calculation method for the basic predicted wind speed data includes:

[0118] Let the first reference group include n first reference wind speed data points, then the i-th first reference wind speed data point is...

[0119] Let the second reference group include m second reference wind speed data points, then the j-th second reference wind speed data point is...

[0120] FP ws =B ws ×K1+D ws ×(1-K1)

[0121]

[0122] In the formula, FP ws Based on the predicted wind speed data, B ws For the first basic wind speed data, D ws This is the second basic wind speed data; K1 is the preset calculation weight.

[0123] Furthermore, the determining module 230 is further configured such that the calculation method for the preset calculation weight includes:

[0124] The first variance of the first reference group is calculated based on all the first reference wind speed data contained in the first reference group;

[0125] The second variance of the second reference group is calculated based on all the second reference wind speed data included in the second reference group;

[0126]

[0127] In the formula, K1 is the preset calculation weight, S1 is the first variance, and S2 is the second variance; the first variance is negatively correlated with the preset calculation weight, and the second variance is positively correlated with the preset calculation weight.

[0128] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the described device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0129] Based on the disclosures of the first and second aspects of this application, the core inventiveness of this application is summarized as follows:

[0130] 1. The scheme disclosed in this application has a multi-source heterogeneous data fusion mechanism. It creatively proposes to combine three key data sources, namely historical monitoring data (including historical average wind speed with time labels), climate and temperature data (temperature, season, climate type) and wind farm geographical data (wind measurement tower location, hilly terrain, roughness, obstacle distribution), and construct a comprehensive prediction model through correlation analysis to solve the problem of low wind speed prediction accuracy under complex terrain.

[0131] 2. The scheme disclosed in this application creatively designs a specific method for quantifying the impact data of wind speed prediction. Specifically, it utilizes a method that includes temperature impact factors, topographic impact factors, and dynamic weighted summation. The temperature impact factor is calculated based on climate temperature data, calculating the temperature difference, sample variance, and mean, and formulating a temperature impact coefficient (T_impact=(|ΔT_avg|+σ). 2 ) -1 The topographic impact factor considers extracting the undulation, roughness, obstacle distribution and ridge / valley location of hilly terrain, and also considers combining the relationship between the location of the wind tower and the wind direction to calculate parameters such as elevation difference and obstacle volume. It also considers generating topographic impact data through a weighted function (including roughness, elevation difference and obstacle contribution), and dynamically weighted fusion to synthesize wind speed prediction impact data by combining temperature and topographic impact factors according to weight (e.g. 0.5:0.5).

[0132] 3. The proposed solution in this application creatively designs an adaptive dual-reference group strategy for basic forecast wind speed, which includes a first reference group and a second reference group. The first reference group contains historical wind speed data that perfectly matches the "month + day + time period" of the forecast task (high time accuracy), while the second reference group contains historical wind speed data that matches the "month + time period" of the forecast task (relaxed date constraints). Dynamic weight allocation is applied to the first and second reference groups, and the variance (σ1) of the two groups of data is calculated. 2 , σ2 2 Weights are allocated inversely to the variance: ω = σ² 2 / (σ1 2 +σ2 2 To ensure high stability, data sets are given higher weights. The basic wind speed formula is: V_base=ω·(ΣV 1i / n)+(1-ω)·(ΣV 2j / m)

[0133] 4. The proposed solution in this application creatively designs a coupled calculation model for comprehensive prediction, which directly multiplies the base predicted wind speed (V_base) with the wind speed prediction impact data (Impact): V_final = V_base × Impact, realizing the dynamic coupling of historical statistical patterns and real-time environmental factors, and improving the prediction robustness under complex terrain.

[0134] 5. The solution provided in this application creatively refines the treatment of hilly terrain in specific wind power scenarios. For example, it introduces parameters such as terrain undulation, roughness length, and obstacle volume to quantify the disturbance of terrain on airflow, and distinguishes between ridges (acceleration effect) and valleys (narrowing / blocking effect), and dynamically adjusts the terrain influence coefficient in combination with wind direction.

[0135] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for predicting wind speed in wind farms with complex terrain, characterized in that, include: Acquire historical monitoring data, climate and temperature data, and geographical data of the wind farm; the historical monitoring data includes historical wind speed data; the historical wind speed data includes multiple historical average wind speed data with historical time tags; the historical time tags include year, month, day, and time period data; the geographical data of the wind farm includes anemometer tower location data and hilly terrain data; The wind speed prediction impact data is analyzed based on the climate temperature data and the wind farm geographical data; the wind speed prediction impact data is correlated with the climate temperature data, the wind measurement tower location data, and the hilly terrain data; The basic predicted wind speed data is determined based on the historical monitoring data; the basic predicted wind speed data is correlated with the historical average wind speed data; The comprehensive wind speed prediction data is calculated based on the wind speed prediction impact data and the basic predicted wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction impact data and the basic predicted wind speed data. The wind speed prediction impact data, which is associated with the climate temperature data, wind tower location data, and hilly terrain data, includes: The temperature impact data is analyzed based on the climate temperature data; the climate temperature data includes temperature data, seasonal data, and climate data. Based on the hilly terrain data and wind tower location data, analyze the terrain relief data, roughness data, obstacle distribution data, and special terrain data; Analyze the terrain impact data based on the terrain relief data, roughness data, obstacle distribution data, and special terrain data; The wind speed prediction impact data is analyzed based on the temperature impact data and the terrain impact data; the wind speed prediction impact data is correlated with the temperature impact data and the terrain impact data. The analysis of temperature impact data based on the climate and temperature data specifically includes: determining the seasonal and climate labels of the wind farm at the current moment based on seasonal and climate data; obtaining the task time label for the wind speed prediction task, which includes year, month, day, and time period data; obtaining the first preset grid temperature data of the wind farm in the previous wind speed measurement cycle at the current moment; obtaining the historical preset grid temperature data of the wind farm in the historical wind speed measurement cycle, which has the same seasonal label, climate label, and task time label as the previous wind speed measurement cycle; and determining the cell temperature data of multiple different wind speed measurement cycles for a preset number of cells within a preset range around each meteorological tower based on the first preset grid temperature data and the historical preset grid temperature data; and using... The cell temperature data from the same cell in the previous wind speed measurement cycle is used as the baseline temperature data. The cell temperature data from the same cell in other wind speed measurement cycles are used as the sample temperature data, resulting in a baseline temperature data and multiple sample temperature data. The sample average temperature data is calculated, which is the average of the multiple sample temperature data. The temperature difference data between the baseline temperature data and the sample average temperature data is calculated. The same processing method is applied to the cell temperature data of each cell to obtain multiple temperature difference data. The variance of the temperature difference data is calculated, which is the sample variance value. The average of the temperature difference data is calculated, which is the sample average value. The temperature influence data is calculated based on the sample variance value and the sample average value. Specifically, the calculation methods for temperature effect data include: In the formula, Data on the influence of temperature, This is the sample mean. This is the preset average value. This represents the sample variance. The calculation method for terrain impact data includes assuming there are n wind measurement towers, and the roughness data of the a-th wind measurement tower is... The altitude of the a-th wind measuring tower is There are m special terrain features corresponding to the a-th wind measuring tower in the same wind direction, and the altitude of the b-th special terrain feature is... There are q obstacles between the a-th wind measuring tower and the corresponding b-th special terrain feature in the same wind direction, and the volume of the p-th obstacle is... ; ; ; ; ; In the formula, For topographic impact data, , , , These are the first intermediate function, the second intermediate function, the third intermediate function, and the fourth intermediate function, respectively. , , These are respectively the preset first sub-weight coefficient, the preset second sub-weight coefficient, and the preset third sub-weight coefficient; and, ; The step of determining the basic predicted wind speed data based on the historical monitoring data; the basic predicted wind speed data being associated with the historical average wind speed data includes: Obtain the task time tag for the wind speed prediction task; the task time tag includes year, month, day, and time period data; A first reference group is selected based on the predicted task time data; the first reference group includes multiple first reference wind speed data; the first reference wind speed data is historical average wind speed data with some historical time tags and some task time tags being the same; the some historical time tags and some task time tags are both month-day and time period data; A second reference group is selected based on the predicted task time data; the second reference group includes multiple second reference wind speed data; the second reference wind speed data is historical average wind speed data with some historical time tags and some task time tags being the same; the some historical time tags and some task time tags are both monthly and time period data; First wind speed baseline data is calculated based on a first reference group; second wind speed baseline data is calculated based on a second reference group; baseline predicted wind speed data is determined based on the first and second wind speed baseline data; the baseline predicted wind speed data is associated with the first and second wind speed baseline data. The calculation methods for the basic predicted wind speed data include: Let the first reference group include n first reference wind speed data points, then the i-th first reference wind speed data point is... ; Let the second reference group include m second reference wind speed data points, then the j-th second reference wind speed data point is: ; ; ; In the formula, Based on the predicted wind speed data, This is the basic data for the first wind speed. This is the second basic wind speed data; Preset calculation weights; The calculation method for the preset calculation weights includes: The first variance of the first reference group is calculated based on all the first reference wind speed data contained in the first reference group; The second variance of the second reference group is calculated based on all the second reference wind speed data included in the second reference group; In the formula, To preset the calculation weights, The first variance, The second variance is the first variance, which is negatively correlated with the preset calculation weights, and the second variance is positively correlated with the preset calculation weights.

2. A wind speed prediction system for wind farms in complex terrain, characterized in that, The system for performing the method as described in claim 1 includes: The acquisition module (210) is used to acquire historical monitoring data, climate temperature data, and wind farm geographical data within the wind farm; the historical monitoring data includes historical wind speed data; the historical wind speed data includes multiple historical average wind speed data carrying historical time tags; the historical time tags include year, month, day, and time period data; the wind farm geographical data includes wind measurement tower location data and hilly terrain data; Analysis module (220) is used to analyze wind speed prediction impact data based on the climate temperature data and the wind farm geographic data; the wind speed prediction impact data is associated with the climate temperature data, the wind tower location data and the hilly terrain data; The determination module (230) is used to determine the basic predicted wind speed data based on the historical monitoring data; the basic predicted wind speed data is associated with the historical average wind speed data. The calculation module (240) is used to calculate the comprehensive wind speed prediction data based on the wind speed prediction impact data and the basic predicted wind speed data; the comprehensive wind speed prediction data is positively correlated with the wind speed prediction impact data and the basic predicted wind speed data.

Images