A virtual wind measurement method based on wind frequency distribution correction

Abstract

The present invention discloses a virtual wind measurement method based on wind frequency distribution correction, comprising the following steps: step 1, collecting the actual wind measurement tower closest to the virtual wind measurement tower to obtain measured data; step 2, extracting the wind speed and direction data of the m-height layer of the actual wind measurement tower in step 1, performing representative year correction, and constructing the wind frequency data file of the m-height layer of the actual wind measurement tower using the wind speed and direction data; step 3, extracting the wind frequency data files of the virtual wind measurement tower and the m-height layer of the actual wind measurement tower; step 4, constructing a same-proportional wind frequency trend correction algorithm model; step 5, using the established same-proportional wind frequency trend correction algorithm model, correcting the wind frequency trends of each wind speed segment and each wind direction of the m-height layer of the extracted virtual wind measurement tower. The present invention corrects the wind farm-level error of the high-resolution numerical simulation wind field by constructing a wind frequency equal-proportional correction algorithm model, thereby overcoming the deficiency of insufficient accuracy of the virtual wind measurement method based on numerical simulation.

Description

Technical Field

[0001] The invention belongs to the technical field of wind resource assessment in wind farms, and in particular relates to a virtual wind measurement method based on wind frequency distribution correction. Background Art

[0002] With the advent of the era of grid parity, wind power development has gradually shifted from extensive to lean development. Distributed wind power, as one of the future development directions, has the characteristics of dispersed wind resources, complex wind conditions, and lack of wind measurement data. In particular, the lack of wind measurement data has greatly increased the risk of project development.

[0003] In recent years, virtual wind measurement technology has emerged to solve the problems of lack of wind tower data, high development costs, and long development cycles. Virtual wind measurement technology has been widely developed and applied because it does not require the establishment of wind towers, saves tower erection costs and observation time. At present, a relatively recognized virtual wind measurement method is to use a mesoscale numerical weather forecast model (such as the weather forecast model WRF) to simulate the wind field in the target area, obtain the wind field results at the target wind tower through interpolation calculation, and construct a virtual wind tower. However, the spatial resolution of the wind field simulated by the mesoscale numerical model is low, and the simulation accuracy is not enough to meet the needs of wind resource assessment. In this case, a small-scale numerical model (such as the small-scale dynamic diagnostic model CALMET) is usually used to further downscale the simulated wind field through the dynamic downscaling method to obtain high-resolution wind field simulation results. Although this numerical simulation method avoids the high cost of setting up wind towers and shortens the development cycle of wind farms, and has assimilated some observation data during model simulation and used wind tower data to correct the simulation results, it is limited by the number of wind tower data and the uneven spatial distribution. As a result, there are still large deviations between the processed simulation results and the actual observations in some areas, especially in some complex mountainous areas in southwest my country. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a virtual wind measurement method based on wind frequency distribution correction. The present invention is based on the simulated wind field data of the mesoscale numerical weather forecast model WRF and the small-scale dynamic diagnosis model WALMET, and utilizes a wind frequency proportional correction algorithm to perform wind farm-level error correction on the high-resolution numerical simulation wind field, thereby overcoming the defect of insufficient accuracy of the virtual wind measurement method based on numerical simulation.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] The present invention provides a virtual wind measurement method based on wind frequency distribution correction, comprising the following steps:

[0007] Step 1: Collect the actual wind towers closest to the virtual wind towers to obtain the measured data of the actual wind towers;

[0008] Step 2: extract the wind speed and direction data of the m-height layer of the actual wind measuring tower in step 1, and make corrections for the representative year, and use the wind speed and direction data corrected for the representative year to construct a wind frequency data file of the m-height layer of the actual wind measuring tower;

[0009] Step 3: Read and parse the existing high-resolution refined grid data files, and extract the wind frequency data files of the virtual wind tower and the actual wind tower at the m-height layer through interpolation calculation;

[0010] Step 4: Based on the extracted wind frequency data file of the m-height layer of the actual wind measurement tower and the measured wind frequency data file of the m-height layer of the actual wind measurement tower, a same-scale wind frequency trend correction algorithm model is constructed;

[0011] Step 5: Use the established same-proportional wind frequency trend correction algorithm model to correct the wind frequency trends of each wind speed segment and each wind direction at the m-height layer of the extracted virtual wind measurement tower, and obtain the corrected wind frequency data file of the m-height layer of the virtual wind measurement tower.

[0012] Furthermore, in step one, the terrain of the actual wind measurement tower and the virtual wind measurement tower is similar, and the altitude difference does not exceed 300m; in flat terrain, the distance between the virtual wind measurement tower and the actual wind measurement tower does not exceed 30 kilometers, and in complex mountainous areas, the distance between the virtual wind measurement tower and the actual wind measurement tower does not exceed 5 kilometers.

[0013] Furthermore, in the step 2, the collected measured data of the actual wind tower are quality controlled, processed and handled, that is, the wind tower observation data are quality controlled by screening to remove frozen data, faulty data, missing data and invalid data.

[0014] Furthermore, in the step 2, if there is no data for the m-level layer, the wind speed data for the m-level layer is obtained by interpolating the wind shear index using the wind speed data for the adjacent layer, and the wind direction data is replaced by the adjacent layer.

[0015] Furthermore, in step 2, the representative annual WS of wind speed and direction data is 代表年 The correction process is as follows:

[0016] S2.1. By calculating the average wind speed data of the m-height layer during the observation period of the actual wind tower, the reference annual measured wind speed WS of the m-height layer of the actual wind tower is obtained. 实测 ;

[0017] S2.2. Obtain the monthly average reanalysis data of the wind speed and direction database for the past 30 years at the corresponding height layer m near the actual wind tower site, calculate the average wind speed during the corresponding observation period of the actual wind tower, and obtain the wind speed WS of the height layer m of the wind tower at the same period 同期 ;

[0018] S2.3. According to the start and end months of the observation period of the actual wind tower, the corresponding months are selected as the start and end time points of the past 30 years, and the average of the monthly average wind speed in the past 30 years is calculated to obtain the 30-year average wind speed WS of the mesoscale m-height layer of the wind tower. 中尺度 ;

[0019]

[0020] Furthermore, in the step 2, the wind frequency data file of the m-th level of the wind tower is actually set to include: wind frequency parameters of each wind direction, wind frequency parameters of each wind speed range, and calculation and processing of wind frequency parameter standardization;

[0021] The wind direction and wind frequency parameters include: statistical calculation of wind direction frequency values ​​of 16 sectors of wind measurement data at the m-height layer of the wind measurement tower, with a wind direction range of 0 to 360° and an interval of 22.5°;

[0022] The wind frequency parameters of each wind speed segment include: statistical calculation of wind speed frequency values ​​of 50 wind speed segments of wind measurement data at the m-height layer of the wind measurement tower, with a wind speed range of 0 to 50 m / s and an interval of 1 m / s;

[0023] The calculation and processing of wind frequency parameter standardization include: standardizing the wind direction frequency values ​​of 16 sectors and the wind speed frequency values ​​of 50 wind speed segments, so that the sum of the wind direction frequency values ​​of the 16 sectors is 100, and the sum of the wind speed frequencies of the 50 wind speed segments in each sector is 1000, and constructing the wind frequency data file of the m-height layer of the actual wind measurement tower.

[0024] Furthermore, in the step three, based on reading and parsing the data in the existing high-resolution refined grid data file, a weighted interpolation algorithm with inverse distance weight coefficients of the nearest 4 grid points is used to extract the wind frequency data files of the m-height layer of the actual wind tower and the virtual wind tower.

[0025] The production of the high-resolution refined grid wind frequency data file is based on the sounding and ground measured data of the sounding meteorological station as the classification basis, and the wind speed, wind direction and daily maximum mixing layer height are used as classification factors. It is divided into 256 types in total, and a typical day is selected from each type for numerical simulation calculation.

[0026] The numerical simulation system adopts the mesoscale meteorological model WRF and the microscale meteorological module Calmet model. During the simulation, the real field of NCEP reanalysis data fused with the measured data of meteorological stations and wind towers is used as the initial and boundary conditions of the model. The WRF model adopts double nesting with horizontal resolutions of 27km and 9km. The overlap rate between the second nesting simulation areas of the WRF model is 33%. The horizontal resolution of the Calmet model is 200m. The simulation calculation takes a typical day as an example. The output time interval is 1h. There are 9 vertical height layers (70m / 80m / 90m / 100m / 110m / 120m / 130m / 140m / 150m). The model adopts LCC projection, with standard latitudes of 25° and 45°, and the central origin is 105° east longitude and 37° north latitude. At the same time, the model output results are processed into a wind frequency TAB file to obtain a high-resolution refined grid wind frequency data product.

[0027] Furthermore, in step 4, the construction of the same-proportional wind frequency trend correction algorithm model includes the following steps:

[0028] S4.1. Calculate the wind frequency ratio coefficients of different wind speed segments from the wind frequency data file of the m-height layer of the actual wind tower and the wind frequency values ​​of 50 wind speed segments in the wind frequency data file of the m-height layer of the actual wind tower, and calculate the wind frequency ratio coefficients of different wind speed segments:

[0029]

[0030] Among them, the wind frequency fo of 16 sectors of the actual wind tower is measured. WD (i), i = 1, 2, ..., 16; the wind frequency fo of 50 wind speed ranges in 16 sectors of the actual wind tower WS (i, j), i = 1, 2, ..., 16, j = 1, 2, ..., 50. The wind frequency fg of the 16 sectors of the actual wind tower is extracted WD (i), i = 1, 2, ..., 16, the wind frequency fg of 50 wind speed ranges in 16 sectors of the actual wind tower WS (i, j), i=1, 2,..., 16, j=1, 2,..., 50;

[0031] S4.2. Calculate the wind frequency ratio coefficients of different sectors based on the wind frequency data file of the m-height layer of the actual wind tower and the wind direction frequency values ​​of 16 sectors in the wind frequency data file of the m-height layer of the actual wind tower, as measured:

[0032]

[0033] The wind frequency fo of 16 sectors of the actual wind tower WD (i), i = 1, 2, ..., 16; the wind frequency fg of the 16 sectors of the actual wind tower is extractedWD (i), i=1, 2, ..., 16;

[0034] Furthermore, in the step 5, the wind frequency trends of each wind speed segment and each wind direction at the m-height layer of the extracted virtual wind tower are corrected as follows:

[0035] fc WS (i, j) = fv WS (i, j)×R WS (j);

[0036] fc WD (i) = fv WD (i)×R WD (i);

[0037] Extracted wind frequencies fv of 16 sectors of the virtual wind tower WD (i), i = 1, 2, ..., 16; the wind frequency fv of 50 wind speed segments in 16 sectors of the virtual wind tower extracted WS (i, j), j = 1, 2, ..., 50. The wind frequency fc of the 16 sectors after the virtual wind tower correction WD (i), i = 1, 2, ..., 16; wind frequency fc of 50 wind speed segments in 16 sectors after correction by the virtual wind tower WS (i,j), i=1,2,...,16, j=1,2,...,50.

[0038] The patent of this invention is based on the simulated wind field data of the mesoscale numerical weather forecast model WRF and the small-scale dynamic diagnostic model CALMET. By constructing a wind frequency proportional correction algorithm model, the error correction of the high-resolution numerical simulation wind field at the wind farm level is performed, thus overcoming the deficiency of insufficient accuracy of the virtual wind measurement method based on numerical simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A flowchart of a virtual wind measurement method according to an embodiment of the present invention;

[0041] Figure 2 Schematic diagram of a wind frequency data file in an embodiment of the present invention;

[0042] Figure 3The Weibull wind frequency distribution diagram actually measured at the 100m height layer of the virtual wind tower in the embodiment of the present invention;

[0043] Figure 4 It is a Weibull wind frequency distribution diagram extracted from high-resolution grid points at a height of 100m of a virtual wind tower in an embodiment of the present invention;

[0044] Figure 5 It is a corrected Weibull wind frequency distribution diagram of the 100m height layer of the virtual wind measurement tower in the embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Reference Figure 1 , a flowchart of a virtual wind measurement method based on wind frequency distribution correction in an embodiment of the present invention. In this embodiment, the virtual wind measurement method based on wind frequency distribution correction includes:

[0047] Step 1: Collect the data of the actual wind tower closest to the virtual wind measurement point. To obtain the actual measured data of the actual wind tower, the following conditions must be met:

[0048] (1) The terrain of the actual wind tower is similar to that of the virtual wind tower, and the altitude difference does not exceed 300m;

[0049] (2) In flat terrain, the distance between the actual wind tower and the virtual wind tower shall not exceed 30 kilometers; in complex mountainous terrain, the distance between the actual wind tower and the virtual wind tower shall not exceed 5 kilometers;

[0050] (3) The measured data from the actual wind tower must be revised for the representative year, and the data completeness rate must be greater than 90%.

[0051] Step 2: Quality control, processing and handling of the collected wind data from the actual wind tower, extracting the wind speed and direction data of the m-level layer of the actual wind tower, and making corrections for the representative year. If there is no data for the m-level layer, the wind speed data of the adjacent layer is used to obtain the wind speed data of the m-level layer through wind shear index interpolation, and the wind direction data is replaced by the adjacent layer, including:

[0052] The wind tower observation data is screened and quality controlled to remove frozen data (abnormal data caused by sensor icing), faulty data (sensor failure), missing data (continuous empty data), and invalid data (data exceeding a reasonable range).

[0053] Using the mesoscale reanalysis data, the actual wind tower observation data after screening and quality control are corrected for the representative year. The specific calculation is as follows:

[0054] By calculating the average wind speed data of the m-height layer during the observation period of the actual wind tower, the reference annual measured wind speed WS of the m-height layer of the actual wind tower is obtained. 实测 ;

[0055] Download the MERRY monthly average reanalysis data of the past 30 years at the corresponding height layer m near the wind tower site, calculate the average wind speed of the wind tower in the corresponding time period, and obtain the wind speed WS of the wind tower m height layer at the same period 同期 ;

[0056] According to the start and end months of the observation period of the actual wind tower, the corresponding months are selected as the start and end time points of the past 30 years, and the average of the monthly average wind speed in these 30 years is calculated to obtain the 30-year average wind speed WS of the m-height layer of the actual wind tower. 中尺度 ;

[0057] According to the representative year calculation formula:

[0058]

[0059] In this way, the representative annual corrected wind speed data of the m-height layer of the actual wind measurement tower is obtained.

[0060] Step 3, using the actual wind tower wind data corrected for the representative year, construct the wind frequency TAB file of the m-height layer of the actual wind tower, including: calculation and processing of wind frequency parameters for each wind direction, wind frequency parameters for each wind speed range, and wind frequency parameter standardization.

[0061] The calculation of each wind direction wind frequency parameter includes: statistically calculating the wind direction frequency values ​​of 16 sectors of the wind measurement data of the m-height layer of the actual wind measurement tower, the wind direction range is 0 to 360°, the interval is 22.5°, that is, the wind directions of 0° to 22.5°, 22.5° to 45°, 45° to 67.5°, ..., 315° to 337.5°, and 337.5° to 360°.

[0062] The calculation of wind frequency parameters in each wind speed segment includes: statistically calculating the wind speed frequency values ​​of 50 wind speed segments of wind measurement data at a height layer of an actual wind measurement tower, with a wind speed range of 0 to 50 m / s and an interval of 1 m / s, namely, wind speeds of 0 m / s to 1 m / s, 1 m / s to 2 m / s, 2 m / s to 3 m / s, ..., 48 to 49 m / s, and 49 to 50 m / s.

[0063] The calculation of wind frequency parameter standardization includes: standardizing the wind direction frequency values ​​of 16 sectors and the wind speed frequency values ​​of 50 wind speed segments so that the sum of the wind direction frequency values ​​of the 16 sectors is 100, and the sum of the wind speed frequencies of the 50 wind speed segments in each sector is 1000, and constructing a wind frequency TAB file of the wind measurement data at the m-height layer of the wind measurement tower.

[0064] Step 4: Read and parse the high-resolution refined grid wind frequency data file, and use the weighted interpolation algorithm with the inverse distance weight coefficient of the nearest four grid points to extract the TAB files of the m-height layer of the actual wind tower and the virtual wind tower. The high-resolution refined grid wind frequency data product is produced based on long-term (nearly 30 years) wind field data, so there is no need to make a representative year correction, including:

[0065] The file of the high-resolution refined grid wind frequency database is based on the sounding and ground measurement data of the sounding meteorological station as the classification basis, and the wind speed, wind direction and daily maximum mixing layer height as the classification factors. It is divided into 256 types in total, and typical days are selected from each type for numerical simulation calculations.

[0066] The numerical simulation system adopts the mesoscale meteorological model WRF and the microscale meteorological module Calmet mode. During the simulation, the real field of the NCEP reanalysis data fused with the measured data of the meteorological station and the wind tower is used as the initial and boundary conditions of the model. The WRF model adopts double nesting, with a horizontal resolution of 27km and 9km, and the overlap rate between the second nesting simulation areas of the WRF model is 33%. The horizontal resolution of the Calmet model is 200m, the simulation calculation takes a typical day as an example, the output time interval is 1h, and the vertical height layer has 9 layers (70m / 80m / 90m / 100m / 110m / 120m / 130m / 140m / 150m). The model adopts LCC projection, the standard latitude is 25° and 45°, and the central origin is 105° east longitude and 37° north latitude. At the same time, the model output results are processed into a wind frequency TAB file to obtain a high-resolution refined grid wind frequency data product.

[0067] Step 5: Based on the wind frequency TAB file of the m-height layer measured by the actual wind tower and the extracted high-resolution refined grid wind frequency TAB file of the m-height layer of the actual wind tower, a wind frequency trend correction algorithm model with the same scale is constructed, including:

[0068] According to the wind speed frequency values ​​of 50 wind speed segments measured at the m-height layer of the actual wind tower and the grid wind frequency TAB file, the wind frequency proportional coefficients of different wind speed segments are calculated:

[0069]

[0070] The wind frequency of 16 sectors observed by the actual wind tower is fo WD(i), i = 1, 2, ..., 16; the wind frequency fo of 50 wind speed segments in 16 sectors observed by the actual wind tower WS (i, j), i = 1, 2, ..., 16, j = 1, 2, ..., 50. The wind frequency fg of the 16 sectors of the actual wind tower extracted from the grid wind frequency data WD (i), i = 1, 2, ..., 16, the wind frequency fg of 50 wind speed ranges in 16 sectors of the actual wind tower extracted from the grid wind frequency data WS (i,j), i=1,2,...,16, j=1,2,...,50.

[0071] According to the wind direction frequency values ​​of 16 sectors measured at the m-height layer of the actual wind tower and the grid wind frequency TAB file, the wind frequency proportional coefficients of different sectors are calculated:

[0072]

[0073] The wind frequency of 16 sectors observed by the actual wind tower is fo WD (i), i = 1, 2, ..., 16; the wind frequency fg of the 16 sectors of the actual wind tower extracted from the grid wind frequency data WD (i), i=1, 2, ..., 16.

[0074] Based on the wind frequency proportional coefficients of different wind speed sections at the m-height layer of the actual wind measurement tower and the wind frequency proportional coefficients of different sectors, a wind frequency trend correction algorithm model with the same proportion is constructed:

[0075]

[0076] The wind frequency fv of 16 sectors of the virtual wind tower extracted from the grid wind frequency data WD (i), i = 1, 2, ..., 16; wind frequency fv of 50 wind speed segments in 16 sectors of the virtual wind tower extracted from the grid wind frequency data WS (i, j), j = 1, 2, ..., 50. The wind frequency fc of the 16 sectors after the virtual wind tower correction WD (i), i = 1, 2, ..., 16; wind frequency fc of 50 wind speed segments in 16 sectors after correction by the virtual wind tower WS (i,j), i=1,2,...,16, j=1,2,...,50.

[0077] Step 6, based on the extracted high-resolution refined grid wind frequency TAB file of the m-height layer of the virtual wind measurement tower, use the established proportional wind frequency trend correction algorithm model to correct the wind frequency trends of each wind speed segment and each wind direction at the m-height layer of the virtual wind measurement tower, and obtain the corrected wind frequency TAB file of the m-height layer of the virtual wind measurement tower.

[0078] Based on the above embodiment, a specific description is given below in combination with a practical example.

[0079] The following examples are used to illustrate the present invention, but are not used to limit the scope of protection of the present invention. The virtual wind tower selected in the example is located in Ruyang County, Luoyang City, Henan Province, with longitude and latitude coordinates of (112°23'36", 34°13'16.08"), an altitude of 836m, and is located in mountainous terrain.

[0080] The specific implementation steps of the virtual wind measurement method based on wind frequency distribution correction are as follows:

[0081] Example step 1: collect the data of the wind tower closest to the virtual wind tower to obtain the actual measured data of the wind tower:

[0082] Virtual wind towers and wind towers must meet the following conditions:

[0083] (1) Similar terrain, with an altitude difference of no more than 300m;

[0084] (2) The distance between two points in flat terrain shall not exceed 30 kilometers, and the distance between two points in complex mountainous terrain shall not exceed 5 kilometers;

[0085] (3) Wind towers must be corrected for a representative year and the data integrity rate must be greater than 90%.

[0086] According to the conditions that the virtual wind tower and wind tower need to meet, a wind tower 8002# is found 3.66 kilometers away from the virtual wind measurement point. The wind measurement period of the 8002# wind tower is from August 28, 2015 to August 27, 2016, and the wind measurement height is 80m / 70m / 50m / 30m / 15m. Its altitude is 749m, and the altitude difference with the virtual wind measurement point is 87m.

[0087] In step 2 of the example, the data of the m-height layer of the wind tower is extracted through the collected observation data of the wind tower, and the data is quality controlled, processed, and corrected for the representative year. If the data of the m-height layer does not exist, the wind speed data of the adjacent layer is used to obtain the wind speed data of the m-height layer through wind shear index interpolation, and the wind direction data is replaced by the adjacent layer:

[0088] Sub-step 2.1, it is necessary to extract the wind direction and speed data of the 100m height layer of the wind tower 8002#. Since there is no data at the 100m height layer of 8002#, the wind speed data of the 80m height layer of 8002# is used to obtain the wind speed data of the 100m height layer of 8002# through wind shear index interpolation calculation, and the wind direction data is replaced by the wind direction data of the 80m height layer. The specific calculation is as follows:

[0089]

[0090] Sub-step 2.2, by screening the wind measurement data of the 100m height layer of the 8002# wind tower, quality control processing was carried out to remove frozen data (abnormal data of sensor icing), fault data (sensor failure), missing data (continuous empty data), and invalid data (data beyond the reasonable range). After quality control processing, the amount of valid data of 8002# reached 94.5%, which met the conditions for data collection.

[0091] Sub-step 2.3 uses the mesoscale reanalysis data to perform a representative year correction on the wind tower observation data after screening and quality control. The specific calculation is as follows:

[0092] Calculate the average wind speed data at the 100m height layer of the 8002# wind tower from August 28, 2015 to August 27, 2016, and obtain the reference annual measured wind speed WS at the 100m height layer of the 8002# wind tower 实测 ;

[0093] Download the MERRY monthly average reanalysis data for the 100m height layer near the 8002# wind tower station for nearly 30 years from 1991 to 2021, calculate the average wind speed of the 8002# wind tower from August 2015 to August 2016, and obtain the wind speed WS of the 100m height layer of the 8002# wind tower during the same period 同期 ;

[0094] According to the start and end months of the observation period of the 8002# wind tower, August 1991 and August 2021 were selected as the start and end time points of the past 30 years, and the average of the 30-year monthly average wind speed from August 1991 to August 2021 was calculated to obtain the 30-year mesoscale average wind speed WS at the 100m height layer of the 8002# wind tower. 中尺度 ;

[0095] According to the representative year calculation formula:

[0096]

[0097] In this way, the representative annual corrected wind speed data of the 100m height layer of the 8002# wind tower is obtained.

[0098] Example step 3, using the 8002# wind tower wind data corrected for the representative year, construct the wind frequency TAB file of the 8002# wind tower at the 100m height layer:

[0099] Sub-step 3.1, statistically calculate the wind direction frequency values ​​of 16 sectors of the wind measurement data of the 8002# wind tower at the 100m altitude layer. The wind direction range is 0 to 360°, with an interval of 22.5°, that is, wind directions of 0° to 22.5°, 22.5° to 45°, 45° to 67.5°, ..., 315° to 337.5°, and 337.5° to 360°.

[0100] Sub-step 3.2, statistically calculate the wind speed frequency values ​​of 50 wind speed segments of the wind measurement data of the 8002# wind measurement tower at a height of 100m. The wind speed range is 0 to 50m / s, with an interval of 1m / s, namely, wind speeds of 0m / s to 1m / s, 1m / s to 2m / s, 2m / s to 3m / s, ..., 48 to 49m / s, and 49 to 50m / s.

[0101] Sub-step 3.3, standardize the wind direction frequency values ​​of the 16 sectors and the wind speed frequency values ​​of the 50 wind speed segments so that the sum of the wind direction frequency values ​​of the 16 sectors is 100, and the sum of the wind speed frequencies of the 50 wind speed segments in each sector is 1000, and construct the wind frequency TAB file of the wind measurement data at the 100m height layer of the 8002# wind tower, as shown in the figure.

[0102] In example step 4, read and parse the high-resolution refined grid wind frequency data product, use the weighted interpolation algorithm with the inverse distance weight coefficient of the nearest 4 grid points, and extract the TAB file of the 100m altitude layer of the wind tower point 8002# (112°25'58.08″, 34°13'31.74″) and the virtual wind measurement point (112°23'36″, 34°13'16.08″). The high-resolution refined grid wind frequency data product is produced based on long-term (nearly 30 years) wind field data, so there is no need to make representative year corrections.

[0103] In step 5 of the example, based on the constructed wind frequency TAB file measured by the 8002# wind tower and the extracted high-resolution refined grid wind frequency TAB file of the 8002# wind tower, a wind frequency trend correction algorithm model with the same scale is constructed:

[0104] Sub-step 5.1, calculate the wind frequency proportional coefficients of different wind speed segments based on the wind speed frequency values ​​of 50 wind speed segments in the 8002# (112°25'58.08″, 34°13'31.74″) 100m altitude layer actual measurement and the grid wind frequency TAB file:

[0105]

[0106] Wind frequencies in 16 sectors observed by the wind tower WD (i), i = 1, 2, ..., 16; wind frequency fo of 50 wind speed segments in 16 sectors observed by the wind tower WS (i, j), i = 1, 2, ..., 16, j = 1, 2, ..., 50. The wind frequency fg of the 16 sectors of the wind tower extracted from the grid wind frequency data WD (i), i = 1, 2, ..., 16, the wind frequency fg of 50 wind speed segments in 16 sectors of the wind tower extracted from the grid wind frequency data WS(i,j), i=1,2,...,16, j=1,2,...,50.

[0107] Sub-step 5.2, calculate the wind frequency ratio coefficients of different sectors based on the wind direction frequency values ​​of 16 sectors in the 100m altitude layer measured in 8002# (112°25'58.08″, 34°13'31.74″) and the grid wind frequency TAB file:

[0108]

[0109] Wind frequencies in 16 sectors observed by the wind tower WD (i), i = 1, 2, ..., 16; wind frequency fg of 16 sectors of the wind tower extracted from the grid wind frequency data WD (i), i=1, 2, ..., 16.

[0110] Sub-step 5.3, based on the wind frequency proportional coefficients of different wind speed segments at the 100m altitude layer of the 8002# wind tower (112°25'58.08″, 34°13'31.74″) and the wind frequency proportional coefficients of different sectors, a wind frequency trend correction algorithm model with the same proportion is constructed:

[0111]

[0112] The wind frequency fv of 16 sectors of the virtual wind tower extracted from the grid wind frequency data WD (i), i = 1, 2, ..., 16; wind frequency fv of 50 wind speed segments in 16 sectors of the virtual wind tower extracted from the grid wind frequency data WS (i, j), j = 1, 2, ..., 50. The wind frequency fc of the 16 sectors after the virtual wind tower correction WD (i), i = 1, 2, ..., 16; wind frequency fc of 50 wind speed segments in 16 sectors after correction by the virtual wind tower WS (i,j), i=1,2,...,16, j=1,2,...,50.

[0113] Example step 6, based on the extracted high-precision grid TAB file of the virtual wind measurement (112°23'36″, 34°13'16.08″) 100m altitude layer, using the established same-scale wind frequency trend correction algorithm model, the wind direction frequency of 16 sectors and the wind speed frequency trend of 50 wind speed segments of the virtual wind measurement tower (112°23'36″, 34°13'16.08″) 100m altitude layer are corrected to obtain the corrected wind frequency TAB file of the virtual wind measurement tower (112°23'36″, 34°13'16.08″) 100m altitude layer.

[0114] The accuracy improvement effect of the average wind speed after correction of the high-resolution grid wind frequency data at the 100m height layer of the virtual wind tower (112°23'36″, 34°13'16.08″) is shown in Table 1. The measured average wind speed at the 100m height layer of the virtual wind tower (corrected for the representative year) is 6.315m / s, and the average wind speed extracted from the high-resolution grid is 5.315m / s. Before correction, the average grid wind speed deviation of the virtual wind tower reached 1.18m / s. After the same proportion of wind frequency trend correction, the average grid wind speed of the virtual wind tower was 6.049m / s. After correction, the average grid wind speed deviation of the virtual wind tower was reduced to 0.266m / s, and the accuracy of the average wind speed was improved by 77.5%. The comparison of the Weibull distribution of the high-resolution grid wind frequency data at the 100m altitude layer at the virtual wind tower (112°23'36", 34°13'16.08") before and after correction is shown in the figure below. Figure 3 , Figure 4 , Figure 5 As shown, Figure 3 is the measured Weibull wind frequency distribution at the virtual wind tower. Figure 4 It is a virtual wind tower grid point to extract Weibull wind frequency distribution. Figure 5 This is the corrected Weibull wind frequency distribution of the virtual wind tower. After correction, the Weibull wind frequency distribution of the virtual wind tower is closer to the actual measurement, especially the frequency of strong winds above 12.5m / s has been significantly improved.

[0115] Table 1

[0116]

[0117] Although the above embodiments have been described in detail for the present invention, it should be understood by those skilled in the art that modifications or improvements may be made based on the contents disclosed in the present invention without departing from the spirit and scope of the present invention, and such modifications and improvements are within the spirit and scope of the present invention.

Claims

1. A virtual wind measurement method based on wind frequency distribution correction, It is characterized in that The following steps are involved: Step 1: Collect the actual wind towers closest to the virtual wind towers to obtain the measured data of the actual wind towers; Step 2: extract the wind speed and direction data of the m-height layer of the actual wind measuring tower in step 1, and make corrections for the representative year, and use the wind speed and direction data corrected for the representative year to construct a wind frequency data file of the m-height layer of the actual wind measuring tower; Step 3: Read and parse the existing high-resolution refined grid data files, and extract the wind frequency data files of the virtual wind tower and the actual wind tower at the m-height layer through interpolation calculation; Step 4: Based on the extracted wind frequency data file of the actual wind tower m-height layer and the measured wind frequency data file of the actual wind tower m-height layer, a same-scale wind frequency trend correction algorithm model is constructed. The construction of the same-scale wind frequency trend correction algorithm model includes the following steps: S4.

1. Calculate the wind frequency ratio coefficients of different wind speed segments from the wind frequency data file of the m-height layer of the actual wind tower and the wind frequency values ​​of 50 wind speed segments in the wind frequency data file of the m-height layer of the actual wind tower, and calculate the wind frequency ratio coefficients of different wind speed segments: Among them, the wind frequency fo of 16 sectors of the actual wind tower is measured. WD (i), i = 1, 2, ..., 16; the wind frequency fo of 50 wind speed ranges in 16 sectors of the actual wind tower WS (i, j), i = 1, 2, ..., 16, j = 1, 2, ..., 50; the wind frequency fg of the 16 sectors of the actual wind tower is extracted WD (i), i = 1, 2, ..., 16, the wind frequency fg of 50 wind speed ranges in 16 sectors of the actual wind tower WS (i, j), i=1, 2,..., 16, j=1, 2,..., 50; S4.

2. Calculate the wind frequency ratio coefficients of different sectors based on the wind frequency data file of the m-height layer of the actual wind tower and the wind direction frequency values ​​of 16 sectors in the wind frequency data file of the m-height layer of the actual wind tower, as measured: The wind frequency fo of 16 sectors of the actual wind tower WD (i), i = 1, 2, ..., 16; the wind frequency fg of the 16 sectors of the actual wind tower is extracted WD (i), i=1, 2, ..., 16; Step 5: Use the established same-proportional wind frequency trend correction algorithm model to correct the wind frequency trends of each wind speed segment and each wind direction at the m-height layer of the extracted virtual wind measurement tower. Correct the wind frequency trends of each wind speed segment and each wind direction at the m-height layer of the extracted virtual wind measurement tower as follows: fc WS (i, jj)=fv WS (i,j)×R WS (j); fc WD (i)=fv WD (i)×R WD (i); Extracted wind frequencies fv of 16 sectors of the virtual wind tower WD (i), i = 1, 2, ..., 16; the wind frequency fv of 50 wind speed segments in 16 sectors of the virtual wind tower extracted WS (i, j), j = 1, 2, ..., 50; wind frequency fc of the 16 sectors after correction by the virtual wind tower WD (i), i = 1, 2, ..., 16; wind frequency fc of 50 wind speed segments in 16 sectors after correction by the virtual wind tower WS (i, j), i = 1, 2, ..., 16, j = 1, 2, ..., 50, and obtain the corrected wind frequency data file of the virtual wind tower m height layer.

2. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In the step 1, the terrain of the actual wind measurement tower and the virtual wind measurement tower is similar, and the altitude difference does not exceed 300m; in flat terrain, the distance between the virtual wind measurement tower and the actual wind measurement tower does not exceed 30 kilometers, and in complex mountainous areas, the distance between the virtual wind measurement tower and the actual wind measurement tower does not exceed 5 kilometers.

3. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In the step 2, the measured data collected from the actual wind tower are quality controlled, processed and handled, that is, the wind tower observation data are quality controlled by screening to remove frozen data, faulty data, missing data and invalid data.

4. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In the step 2, if there is no data for the m-level layer, the wind speed data for the m-level layer is obtained by interpolating the wind shear index using the wind speed data of the adjacent layer, and the wind direction data is replaced by the adjacent layer.

5. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In step 2, the representative annual WS of wind speed and direction data 代表年 The correction process is as follows: S2.

1. By calculating the average wind speed data of the m-height layer during the observation period of the actual wind tower, the reference annual measured wind speed WS of the m-height layer of the actual wind tower is obtained. 实测 ; S2.

2. Obtain the monthly average reanalysis data of the wind speed and direction database for the past 30 years at the corresponding height layer m near the actual wind tower site, calculate the average wind speed during the corresponding observation period of the actual wind tower, and obtain the wind speed WS of the height layer m of the wind tower at the same period 同期 ; S2.

3. According to the start and end months of the observation period of the actual wind tower, the corresponding months are selected as the start and end time points of the past 30 years, and the average of the monthly average wind speed in the past 30 years is calculated to obtain the 30-year average wind speed WS of the mesoscale m-height layer of the wind tower. 中尺度 ; 6. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In the step 2, the wind frequency data file of the m-th level of the wind tower is actually set up, including: wind frequency parameters of each wind direction, wind frequency parameters of each wind speed range, and calculation and processing of wind frequency parameter standardization; The wind direction and wind frequency parameters include: statistical calculation of wind direction frequency values ​​of 16 sectors of wind measurement data at the m-height layer of the wind measurement tower, with a wind direction range of 0 to 360° and an interval of 22.5°; The wind frequency parameters of each wind speed segment include: statistical calculation of wind speed frequency values ​​of 50 wind speed segments of wind measurement data at the m-height layer of the wind measurement tower, with a wind speed range of 0 to 50 m / s and an interval of 1 m / s; The calculation and processing of wind frequency parameter standardization include: standardizing the wind direction frequency values ​​of 16 sectors and the wind speed frequency values ​​of 50 wind speed segments, so that the sum of the wind direction frequency values ​​of the 16 sectors is 100, and the sum of the wind speed frequencies of the 50 wind speed segments in each sector is 1000, and constructing the wind frequency data file of the m-height layer of the actual wind measurement tower.

7. A virtual wind measurement method based on wind frequency distribution correction as claimed in claim 1, It is characterized in that In the step three, based on reading and parsing the data in the existing high-resolution refined grid data file, a weighted interpolation algorithm with inverse distance weight coefficients of the nearest four grid points is used to extract the wind frequency data files of the m-height layer of the actual wind tower and the virtual wind tower.

Images