A method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy

Abstract

The application discloses a method for quantifying seasonal variation characteristics of wind speed in different regions based on ordering entropy, which comprises the following steps: performing data standardization processing and basic quality control on original data files of monthly wind speed data; obtaining annual average values of the standardized monthly wind speed data, and performing spatial grid interpolation processing to obtain spatial distribution characteristics of the annual wind speed values; respectively calculating ordering entropy values of the standardized monthly wind speed time series of all weather stations; performing spatial grid interpolation processing on the ordering entropy values of all weather stations to obtain spatial distribution characteristics of the ordering entropy values; and comprehensively analyzing and comparing the spatial distribution maps of steps (2) and (4) to obtain the seasonal variation characteristics of wind speed in different regions in China. The application provides a quantitative characterization parameter or index (ordering entropy of monthly average wind speed) for describing the seasonal variation characteristics of wind speed in regions across China.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of research on seasonal variation characteristics of wind speed in different regions. Specifically, it is a method for quantifying seasonal variation characteristics of wind speed in different regions based on permutation entropy. BACKGROUND

[0002] Under the background of global climate change, extreme weather such as high temperature, drought, heavy rain and strong wind occurs frequently, which has a great impact and damage on the ecological environment of some places. As a basic meteorological element, wind is a kind of renewable clean energy. Accurate description of the seasonal variation characteristics of wind can reveal the regional characteristics and distribution of wind speed seasonal variation, which has important practical significance for fully developing and utilizing wind energy resources, meteorological disaster prevention and reduction, protecting ecological environment, ensuring production and construction, guiding agriculture and animal husbandry, and promoting economic development.

[0003] The climate system itself is a multi-level, nonlinear, non-scale and strong dissipation complex system. Therefore, it is appropriate to use some mathematical statistics that can reflect the uncertainty characteristics of atmospheric variables to analyze the climatological characteristics. However, the existing methods are mostly based on the average value and spatial distribution of meteorological elements, which intentionally or unintentionally ignores some high-order anomaly structure change characteristics in wind speed. Entropy itself is a basic physical quantity in thermodynamics and statistical physics, which can be used to describe the degree of chaos of a system. Specifically, the higher the entropy value, the more chaotic the system, and the less regular characteristics of certainty. Permutation entropy (PE) is another form of entropy used to quantify the organization level of time series. Bandt and Pompe proposed the concept of permutation entropy based on the symbolic representation of time series. This method gives the ordering structure by comparing the size of adjacent values in the original sequence, and obtains the permutation entropy by counting the probability of different ordering structures. This method makes up for the defect of previous methods that only consider the average value of meteorological elements. Its advantages are easy to calculate, simple and practical, and good robustness to noise. Existing research results show that the permutation entropy algorithm has become a useful tool for detecting dynamic mutations of complex systems. However, there is no research on using permutation entropy to quantify the seasonal variation characteristics of wind speed in different regions. At the same time, existing research on seasonal variation of wind speed using other statistical methods mostly focuses on a certain region or station, and there is no research on quantitatively describing the seasonal variation characteristics of wind speed in the whole country. SUMMARY

[0004] Therefore, the technical problem to be solved by the present application is to provide a method for quantifying seasonal variation characteristics of wind speed in different regions based on permutation entropy.

[0005] To solve the above technical problems, the present application provides the following technical solutions:

[0006] A method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, comprising the following steps:

[0007] (1) The original data file of the wind speed monthly value data is subjected to data standardization processing and basic quality control;

[0008] (2) The annual average value of the obtained standardized wind speed monthly value data is calculated, and spatial grid interpolation processing is performed to obtain the spatial distribution characteristics of the wind speed annual value;

[0009] (3) The ranking entropy value of the standardized wind speed monthly value time series of all weather stations is calculated;

[0010] (4) The ranking entropy values of all weather stations are subjected to spatial grid interpolation processing to obtain the spatial distribution characteristics of the ranking entropy values;

[0011] (5) Comprehensive analysis is performed to compare the spatial distribution maps of steps (2) and (4) to obtain the seasonal variation characteristics of wind speed in different regions of China.

[0012] The above method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, in step (1),

[0013] (1-1) The wind speed monthly value data file contains the average wind speed monthly value data of all stations in different time periods, and the data length of all stations in the extracted time period needs to be ensured to be the same, and then the data is subjected to standardization processing;

[0014] (1-2) The data subjected to standardization processing is subjected to basic quality control to eliminate obvious error data in the data to avoid affecting the final result; the quality control includes limit value check, extreme value check, internal consistency check, space-time consistency check and manual intervention check;

[0015] (1-3) The data subjected to quality control check is retained for subsequent processing.

[0016] The above method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, in step (2),

[0017] (2-1) All data of a year in the average wind speed monthly value data of all stations in different time periods in step (1) are selected, and the annual average wind speed value of each station in the year is calculated using the 12-month average wind speed data of all stations;

[0018] (2-2) Using a spatial grid interpolation method to interpolate the annual average wind speed data of all stations to latitude and longitude grid points, and drawing by AUSPLINE or SURFER drawing software; the spatial grid interpolation method includes Kriging interpolation, nearest neighbor interpolation, bilinear interpolation, thin plate spline interpolation;

[0019] (2-3) Retain the drawing results for comparative study and analysis with the subsequent ranking entropy results;

[0020] The above method for quantifying the seasonal variation characteristics of wind speed in different regions based on ranking entropy, in step (3),

[0021] (3-1) For a certain station's monthly average wind speed time series {x i} in the monthly average wind speed data of all stations in different time periods in step (1), first perform phase space reconstruction to obtain the reconstructed monthly average wind speed time series X(i):

[0022] X(i) = [x(i), x(i+n), …, x(i+(D-1)n)];

[0023] In the above formula, D and n = 2 h are the embedding dimension and delay time respectively, h = 1, 2, …, 10, h is called the scale factor;

[0024] (3-2) Rearrange the m reconstructed components x(i), x(i+n), …, x(i+(D-1)n) of the reconstructed monthly average wind speed time series X(i) in ascending order to obtain a group of symbol sequences A(g), A(g) = [j1, j2, …, j D ], where g = 1, 2, …, k, and k ≤ D!; where D different symbols j1, j2, …, j D There are D! different ordering structures, and the probability P1, P2, …, P D！ of each ordering structure is calculated;

[0025] (3-3) Calculate the ranking entropy value of the reconstructed monthly value wind speed time series of all stations, similar to the form of cannon information entropy, then the ranking entropy H s [P] is defined as:

[0026]

[0027] Where i = 1, 2, …, N; Pi is the probability of the i-th ordering structure;

[0028] 0 ≤ H s [P] ≤ 1.

[0029] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy,

[0030] When H s When [P]=1, it corresponds to a completely random system, and the probability of occurrence of all D! ranking structures is the same.

[0031] If there is a specific organizational structure in the reconstructed monthly wind speed time series, then H s The value of [P] will be less than 1.

[0032] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, the embedding dimension D is selected according to the length of the monthly average wind speed time series data, and D! is much less than N to obtain reliable results.

[0033] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, D=3, 4, …, 7.

[0034] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, D=4.

[0035] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, in step (4),

[0036] (4-1) The same spatial grid interpolation method as step (2-2) is used to interpolate the ranking entropy values of all stations obtained in step (3-3) to the latitude and longitude grid, and drawing is performed through AUSPLINE or SURFER software.

[0037] (4-2) Select representative stations for comparative study;

[0038] (4-3) Analyze the distribution characteristics of the ranking entropy values of wind speed in different regions by point-to-area, and give quantitative analysis and conclusion.

[0039] The method for quantifying seasonal variation characteristics of wind speed in different regions based on ranking entropy, in step (5),

[0040] (5-1) Comprehensive analysis and comparative study of the spatial distribution maps obtained in steps (2-3) and (4-3) to analyze the differences between the two methods of “only statistical average” and “considering ranking structure” in describing the fine structure of wind speed seasonal variation;

[0041] (5-2) Comparative analysis and research are carried out by station and region, and the characteristics of wind speed seasonal variation in different regions of China are comprehensively obtained, and the uncertainty differences of wind speed seasonal variation in different regions and the influencing factors behind them are analyzed.

[0042] The technical scheme of the present application has the following beneficial technical effects:

[0043] The application provides a method for quantifying seasonal variation characteristics of wind speed in different regions based on ordering entropy, which quantifies the seasonal variation characteristics of wind speed in different regions. Compared with the existing researches which are mostly limited to simple qualitative description or limited to a certain region or station, the application provides a quantitative representation parameter or index (ordering entropy of monthly mean wind speed) for describing the seasonal variation characteristics of wind speed in the whole country. The statistical analysis and law research of atmospheric boundary layer wind speed by using the ordering entropy method are helpful to further deepen the understanding of the structure and variation of wind speed. The regional characteristics and distribution law of the seasonal variation of wind speed are revealed, which has important practical significance for fully developing and utilizing wind energy resources, meteorological disaster prevention and reduction, ecological environment protection, production and construction guarantee, guidance of agriculture and animal husbandry, and promotion of economic development. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 Flowchart in the application;

[0045] Figure 2 Spatial distribution of the annual mean wind speed after interpolation;

[0046] Figure 3 The monthly mean wind speed time series of station 5108 in the period of 1981-2010, PE=0.755, and the red numbers in the figure represent months;

[0047] Figure 4 The monthly mean wind speed time series of station 5054 in the period of 1981-2010, PE=0.848, and the red numbers in the figure represent months;

[0048] Figure 5 The monthly mean wind speed time series of station 5766 in the period of 1981-2010, PE=0.999;

[0049] Figure 6 Spatial distribution of the ordering entropy of monthly mean wind speed after interpolation. DETAILED DESCRIPTION

[0050] I. Research data

[0051] The data used in the application is derived from the wind speed data set of the national ground station in China, which is publicly released by the National Meteorological Information Center and put into business use. The data set corrects the breakpoints caused by various human reasons such as station migration, wind speed instrument replacement, instrument replacement, manual observation to automatic observation and observation time change, and improves the consistency of the time series of ground wind speed data.

[0052] The data of the present application is derived from the "China National Ground Meteorological Station Homogenized Wind Speed Monthly Value Dataset (V1.0)" published by the National Meteorological Information Center, which contains wind speed monthly value data of 2400 national meteorological stations in China. Basic quality control is performed on the data used, and the specific aspects of quality control include: limit value check, extreme value check, internal consistency check, spatiotemporal consistency check, and manual intervention check, etc. Obvious error data in the data is removed to avoid affecting the final result.

[0053] II. Quantifying the seasonal variation characteristics of wind speed in different regions based on ranking entropy

[0054] Ranking entropy can be used as an indicator to reflect the complexity of a system. For example, in a climate system, the ranking entropy value of meteorological elements (such as wind speed time series) in some areas is larger, indicating that the higher the uncertainty, i.e. it is difficult to predict or forecast the wind speed in the next month based on the wind speed value in the previous month. Places with smaller ranking entropy values have better certainty. Since the calculation of ranking entropy is based on the arrangement of wind speed amplitude on the time axis, while the calculation of statistical quantities such as mean value is only based on a single statistical analysis of wind speed amplitude within a time period, the distribution of ranking entropy better reflects the climate change characteristics of wind speed in a region than the distribution of mean wind speed.

[0055] As shown in Figure 1 , a technical flowchart used in the present study is given. Specifically, the following steps are included.

[0056] (1) Perform data standardization processing and basic quality control on the wind speed monthly value data original data file;

[0057] (1-1) The wind speed monthly value data original data file contains average wind speed monthly value data of all stations in different time periods, and the data length of all stations in the extracted time period needs to be ensured to be the same, and then the data is standardized; in this embodiment, all data from 1981 to 2010 is selected, so that the data length of all stations is the same, effectively avoiding the influence of data length difference on the final statistical result;

[0058] (1-2) Perform basic quality control on the standardized data to remove obvious error data in the data to avoid affecting the final result; quality control includes: limit value check, extreme value check, internal consistency check, spatiotemporal consistency check, and manual intervention check;

[0059] (1-3) The data after quality control check is retained for subsequent processing.

[0060] (2) Calculate the annual mean value of the obtained standardized wind speed monthly value data, and perform spatial grid interpolation processing to obtain the spatial distribution characteristics of wind speed annual value;

[0061] (2-1) Select all data of a certain year (for example, 2008) from all the monthly average wind speed data of different time periods of all stations in step (1), and obtain the annual average wind speed value of each station in 2008 for all stations by using the 12-month average wind speed data of the year;

[0062] (2-2) Interpolate the annual average wind speed data of all stations to latitude and longitude grid points by using a spatial grid interpolation method, and draw the graph by using AUSPLINE or SURFER drawing software; the spatial grid interpolation method includes Kriging interpolation, nearest neighbor interpolation, bilinear interpolation, and thin plate spline interpolation;

[0063] (2-3) Keep the drawing results for comparative analysis with the subsequent ranking entropy results;

[0064] (3) Obtain the ranking entropy value of the standardized monthly average wind speed time series of all weather stations;

[0065] (3-1) For the monthly average wind speed time series {x i} of a certain station in the monthly average wind speed data of different time periods of all stations in step (1), first perform phase space reconstruction to obtain the reconstructed monthly average wind speed time series X(i):

[0066] X(i) = [x(i), x(i+n), …, x(i+(D-1)n)];

[0067] In the above formula, D and n = 2 h are the embedding dimension and the delay time, respectively, h = 1, 2, …, 10, and h is called the scale factor;

[0068] (3-2) Rearrange the m reconstructed components x(i), x(i+n), …, x(i+(D-1)n) of the reconstructed monthly average wind speed time series X(i) in ascending order to obtain a group of symbol sequences A(g), A(g) = [j1, j2, …, j D ], where g = 1, 2, …, k, and k ≤ D!; where D different symbols j1, j2, …, j D There are D! different ranking structures, and the probabilities P1, P2, …, P D！ of each ranking structure are calculated;

[0069] (3-3) Obtain the ranking entropy value of the reconstructed monthly average wind speed time series of all stations, which is similar to the form of Shannon information entropy, and the ranking entropy H s [P] is defined as:

[0070]

[0071] where i = 1, 2, …, N; Pi is the probability of the i-th order structure;

[0072] 0≤H s [P]≤1.

[0073] (3-4) When H s [P] = 1, it corresponds to a completely random system, in which all D! order structures have the same probability. If there is a specific organization structure in the time series, then H s [P] will be less than 1.

[0074] The embedding dimension D plays an important role in estimating the probability of order structure, because it determines the total number of all possible order structures.

[0075] In fact, the choice of D depends on the length of the data, and should satisfy D! « N to get a reliable result.

[0076] In practical applications, Bandt and Pompe suggest that the value of D should be D = 3, 4, …, 7. In this embodiment, the value of D is taken as 4.

[0077] (4) For all the order entropy values of the meteorological stations, spatial grid interpolation is performed to obtain the spatial distribution characteristics of the order entropy values;

[0078] (4-1) The same spatial grid interpolation method as in step (2-2) is used to interpolate the order entropy values of all stations obtained in step (3-3) to the latitude and longitude grid points, and plotting is performed through AUSPLINE or SURFER software. It is worth mentioning that the same interpolation method as in the previous step (2-2) must be used here to exclude the influence of different interpolation methods on the final result.

[0079] (4-2) Representative stations are selected for comparative study; for example, the geographical location difference and the difference in climate characteristics between the station with the minimum order entropy value and the station with the maximum order entropy value are compared and analyzed.

[0080] (4-3) Point-to-area analysis is performed to analyze the distribution characteristics of the wind speed order entropy values in different regions, and quantitative analysis and conclusions are given.

[0081] (5) Comprehensive analysis is performed to compare the spatial distribution maps obtained in steps (2) and (4) to obtain the seasonal variation characteristics of wind speed in different regions of China.

[0082] (5-1) Comprehensive analysis is performed to compare and study the spatial distribution maps obtained in steps (2-3) and (4-3) to analyze the differences in the fine structure description of the seasonal variation of wind speed between the two methods of "only statistical average" and "considering order structure";

[0083] (5-2) Sub-site and regional comparative analysis research and development, comprehensive acquisition of different regions of China wind speed seasonal variation characteristics, analysis of the uncertainty of different regional wind speed seasonal variation and its influencing factors.

[0084] III. Results and analysis

[0085] 3.1, the spatial distribution characteristics of the annual wind speed

[0086] Figure 2 The spatial distribution of the annual average wind speed after interpolation processing in step (2) is given. It can be seen that the maximum value of the annual average wind speed is located in the south of Kunlun Mountain in the Qinghai-Tibet Plateau, and the minimum value of the national average wind speed is located in the three famous basins of Tarim Basin, Qaidam Basin and Sichuan Basin. It can be said that the size of the annual average wind speed is significantly related to the local altitude. At the same time, in the eastern coastal areas, it is also affected by the marine climate, and the closer to the coast, the larger the annual average wind speed.

[0087] At the same time, it can be obtained that the spatial distribution of the annual average wind speed is significantly different in different geographical regions, which is due to the comprehensive influence of local topography and atmospheric circulation pattern on wind speed, for example, the Inner Mongolia region presents a distribution pattern of high in the west and low in the east, for example, Hohhot is located in the mid-temperate semi-arid continental monsoon climate zone, which is one of the areas with serious sand-dust weather phenomenon in northern China. Xinjiang region is located in the upper reaches of the westerlies, and its wind speed distribution is significantly related to the topographic features of "three mountains sandwiching two basins" in the region.

[0088] However, only through the spatial distribution of the annual average wind speed, it is not enough to accurately depict the regional distribution characteristics of wind speed. Especially for the annual seasonal variation characteristics of wind speed in different regions, Figure 2 it is even more difficult to depict in depth. Therefore, it is necessary to analyze and study the annual seasonal variation characteristics of wind speed in different regions. As follows Figures 3-5 Three representative sites are selected, and their annual variation time series are given, which can clearly show their different variation characteristics.

[0089] 3.2, the annual variation characteristics of wind speed in different regions

[0090] Through the study of the annual variation characteristics of wind speed in different regions, it is found that, in general:

[0091] (1) In Xinjiang and part of the Qinghai-Tibet Plateau, the annual variation of wind speed presents a single peak distribution, that is, the wind speed is the largest in June, as Figure 3 shown;

[0092] Figure 3This is the monthly average wind speed time series of the Fuyun meteorological station (51087) in Altay Prefecture, Xinjiang, from 1981 to 2010. The annual seasonal variation of wind speed shows a clear unimodal distribution, with the highest average wind speed in June, gradually decreasing thereafter, and the lowest average wind speed in December, gradually increasing thereafter. Located deep in the Eurasian continent at a high altitude, this area is most significantly affected by continental monsoons, experiencing seasonal strong winds every spring and summer, while wind speeds are lowest in winter. This provides unique climatic conditions for high-altitude snow sports in the region (high mountains, low temperatures, but low wind speeds). Its ordinal entropy value is 0.755, the lowest among all stations, indicating that the seasonal variation of wind speed in the area represented by this station has a significant regularity, exhibiting a clear annual cyclical variation characteristic with minimal uncertainty.

[0093] (2) In the vast regions of Northeast, North China, and eastern Northwest China, the annual variation of wind speed exhibits a bimodal distribution pattern—that is, the peak wind speed periods are April / May and September / October, respectively. Figure 4 As shown;

[0094] Figure 4 This is a time series of monthly average wind speeds from the Xiaoergou meteorological station (station number 50548) in Hulunbuir City, Inner Mongolia, from 1981 to 2010. The annual seasonal variation of wind speed shows a distinct bimodal distribution, with the highest wind speed during the Spring Festival (April), followed by autumn (October), and the lowest in summer and winter. This characteristic is significantly different from the unimodal distribution observed at the Fuyun meteorological station (station number 51087) in Altay Prefecture, Xinjiang. The strong spring winds observed at the Xiaoergou station are due to the continuous southward movement of active Siberian cold air. This area is arid and semi-arid, and rapid continental warming in spring easily leads to the characteristic of strong spring winds and weak summer winds. The ordination entropy value is 0.848, indicating that the seasonal variation of wind speed in the region represented by this station follows a certain regularity, exhibiting a distinct bimodal variation with relatively low uncertainty.

[0095] (3) In the vast southern region, the annual variation in wind speed does not have obvious characteristics, and the wind speed in each month is roughly the same, such as Figure 5 As shown.

[0096] Figure 5 This is the monthly average wind speed time series of the Taojiang meteorological station (station number 57666) in Yiyang City, Hunan Province, from 1981 to 2010. It can be seen that the seasonal differences in average wind speed are not significant, which represents the seasonal variation characteristics of wind speed in most parts of southeastern China. Its sorting entropy value is 0.999, the highest among all stations, meaning that the seasonal variation of wind speed in the area represented by this station does not exhibit significant regularity and has the greatest uncertainty.

[0097] In summary Figures 3-5As can be seen, wind speeds in different regions exhibit varying seasonal characteristics throughout the year; some show a unimodal pattern, some a bimodal pattern, and others show no seasonal variation at all. While it is feasible to analyze the seasonal variations of each of the more than 2,400 meteorological stations across the country by mapping and analyzing each station individually, this qualitative and descriptive research would be extremely labor-intensive. Therefore, it is essential to design a characterizing parameter or index to quantitatively depict or describe the seasonal variations of wind speeds at the stations used, in order to deepen our understanding of the patterns of seasonal wind speed variations in different regions of the country. The sorting entropy method used in this paper, by considering the relative magnitudes of adjacent values ​​in the wind speed time series, can effectively characterize the structural features of this complex system.

[0098] 3.3 Spatial Distribution Characteristics of Monthly Average Wind Speed ​​Ranking Entropy Values ​​Nationwide

[0099] Depend on Figure 6 It can be seen that the distribution of wind speed ordering entropy has obvious regional characteristics, further illustrating that the uncertainty of wind speed has significant geographical distribution characteristics, which has important reference value for the opening and utilization of wind energy resources.

[0100] The regions with higher ordination entropy values ​​are located in the southeast, while those with lower values ​​are in the central and western regions, particularly in Xinjiang and parts of the Qinghai-Tibet Plateau, where ordination entropy values ​​are lowest. This aligns with previous findings. The higher ordination entropy in the southeast indicates poor seasonal variation in wind speed, with little difference between months and significant uncertainty (it's difficult to predict the wind speed of the following month based on the wind speed of the current month). Conversely, the seasonal wind speed variation in regions with higher ordination entropy values ​​exhibits either a unimodal or bimodal pattern, showing less uncertainty (the wind speed of the following month can be predicted with a high probability based on the wind speed of the current month).

[0101] Compared to Figure 2 In other words, Figure 6 The differences in wind speed characteristics reflected between the north and south are more significant. It can be seen that, with the Qinling-Huaihe line as the boundary, the ordination entropy values ​​differ significantly between the north and south, but this characteristic is not reflected in... Figure 2 This further illustrates that, since the calculation of sorting entropy is based on the arrangement of wind speed amplitudes over time, while the calculation of statistics such as the average is based on only one statistical analysis of wind speed amplitudes within a time period, the distribution of sorting entropy better reflects the climate change characteristics of wind speed in a region than the distribution of the wind speed mean.

[0102] It can be further seen that the sorting entropy size has a significant correlation with the altitude and the terrain. The comprehensive effect of atmospheric circulation and large terrain makes the sorting entropy present a unique spatiotemporal distribution pattern, which reflects the spatiotemporal complexity of the wind speed change in different regions and well explains that the local meteorological elements are affected by the comprehensive effect of atmospheric circulation and local terrain.

[0103] IV. Conclusion and discussion

[0104] The application provides a characterization parameter or index for quantifying the seasonal change characteristics of the wind speed in different regions. The sorting entropy method can effectively measure the complexity of the climate system simply and directly. The spatial distribution pattern of the wind speed entropy value is given in the paper, which reflects the complexity of atmospheric circulation and its regional differences, and to some extent, embodies the influence effect of the local average atmospheric circulation form and local terrain. The method of the application is only taken as an example of the meteorological element of wind speed, and has the same universality for other meteorological elements. However, how the method is applicable to other meteorological elements, whether the internal correlation between the changes of various meteorological elements can be quantified, how the method is applicable to meteorological elements of different resolutions, whether new feature rules and conclusions can be found, and the like, are worth further research and in-depth discussion.

[0105] It is worth mentioning that the time scale of the wind speed change focused in the paper is the annual scale, and the seasonal change of the monthly average wind speed within a year is mainly discussed, and how to quantitatively analyze the method of sorting entropy. Other existing researches mainly focus on the interannual or interdecadal time scale (tens of years or even dozens of years), and point out that under the background of climate change, the average wind speed in most regions of China presents a decreasing trend, and the atmospheric circulation change and climate warming are the possible reasons for the decrease of the wind speed. The conclusions of the existing research results are not in conflict with the conclusions of the paper. It can be said that the sorting entropy method proposed herein plays a good expansion and improvement role in accurately describing the change characteristics and regional distribution of the wind speed, especially on the seasonal time scale within a year.

[0106] The dependence of human beings on clean energy such as wind energy is increasing, and therefore, it is of great significance to study the certainty or predictability of wind energy resources within a year. The application quantifies the seasonal change characteristics of the wind speed in the nationwide range by using the sorting entropy, which has certain guiding significance for wind speed prediction modeling. By predicting the seasonal wind speed, the wind energy resource evaluation can be effectively carried out in advance, especially when the monthly average wind speed suddenly increases or decreases, causing the contradiction between the supply and demand of wind energy, timely measures are taken for prevention, and the systematic planning and development of the nationwide wind energy resources are effectively improved.

[0107] Obviously, the above embodiments are only examples for clearly illustrating the present application and are not intended to limit the present application. Based on the above description, one of ordinary skill in the art can make other different forms of changes or modifications. Here, it is not necessary or possible to enumerate all the embodiments. The obvious changes or modifications derived from the above should be covered in the protection scope of the present application.

Claims

1. A method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy, characterized in that, It includes the following steps: (1) Perform data standardization processing and basic quality control on the original data file of monthly wind speed data. (2) For the standardized monthly wind speed data obtained, calculate the annual average value and perform spatial grid interpolation processing to obtain the spatial distribution characteristics of annual wind speed values. (3) For the standardized monthly wind speed time series of all meteorological stations, calculate their permutation entropy values respectively. (4) Perform spatial grid interpolation processing on the permutation entropy values of all meteorological stations obtained to obtain the spatial distribution characteristics of permutation entropy values. (5) Through comprehensive analysis, compare the spatial distribution characteristics in steps (2) and (4) to obtain the seasonal change characteristics of wind speed in different regions of China. In step (3), (3-1) For the monthly average wind speed time series of a certain station in the monthly average wind speed data of all stations in different time periods in step (1) {x i };i=1,2,…,N, First, phase space reconstruction is performed to obtain the reconstructed monthly average wind speed time series X(i): X(i) = [x(i), x(i + n), …, x(i + (D - 1)n)]; In the above formula, D and n = 2 h Here, h represents the embedding dimension and hysteresis time, respectively, where h = 1, 2, ..., 10, and h is called the scaling factor. (3-2) Rearrange the m reconstructed components x(i), x(i+n), ..., x(i+(D-1)n) of the reconstructed monthly average wind speed time series X(i) in ascending order to obtain a symbol sequence A(g), A(g) = [j1,j2,...,jn]. D ], where g = 1, 2, ..., k, and k ≤ D!; where D distinct symbols j1, j2, ..., j D There are D! different sorting structures. Calculate the probability P1, P2, ..., P of each sorting structure. D！ ; (3-3) Calculate the sorting entropy value of the reconstructed monthly wind speed time series for all stations, similar to the form of cannon information entropy, then the sorting entropy H s [P] is defined as: where i = 1, 2, …, N; Pi is the probability of the i-th sorting structure; 0≤H s [P]≤1。 2. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 1, characterized in that, In step (1), (1-1) The original data file of monthly wind speed data contains the average monthly wind speed data of all stations in different time periods, and it is necessary to ensure that the data lengths of all stations in the extracted time period are the same, and then perform data standardization processing. (1-2) For the data after standardization processing, perform basic quality control to剔除 obvious error data in the data to avoid affecting the final result. The quality control includes: boundary value check, extreme value check, internal consistency check, spatio-temporal consistency check and manual intervention check. (1-3) Retain the data after quality control check for subsequent processing.

3. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 1, characterized in that, In step (2), (2-1) Select all the data of a certain year from the average monthly wind speed data of all stations in different time periods in step (1). For all stations, use the average monthly wind speed data of 12 months in this year to calculate the annual average wind speed value of each station in this year. (2-2) Use a spatial grid interpolation method to interpolate the annual average wind speed data of all stations to longitude and latitude grids, and draw a graph through drawing software such as AUSPLINE or SURFER; the spatial grid interpolation method includes Kriging interpolation, nearest neighbor interpolation, bilinear interpolation, thin plate spline interpolation. (2-3) Retain the drawing result for comparative research and analysis with the subsequent permutation entropy result.

4. The method for quantifying the seasonal change characteristics of wind speed in different regions based on permutation entropy according to claim 1, wherein When H s When [P] = 1, it corresponds to a completely random system, in which all D! sorting structures have the same probability of appearing; If the reconstructed monthly wind speed time series has a specific organizational structure, then H s The value of [P] will be less than 1.

5. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 4, characterized in that, The selection of the embedding dimension D depends on the length of the monthly average wind speed time series data, satisfying D! << N to obtain a reliable result.

6. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 5, characterized in that, D＝3,4,…,7。 7. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 6, characterized in that, D＝4。 8. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 1, characterized in that, In step (4); (4-1) Adopt the same spatial grid interpolation method as in step (2-2) to interpolate the permutation entropy values of all stations obtained in step (3-3) to longitude and latitude grids, and draw a graph through software such as AUSPLINE or SURFER. (4-2) Select representative stations for comparative research. (4-3) Analyze the distribution characteristics of the permutation entropy values of wind speed in different regions from point to area, and give quantitative analysis and conclusions.

9. The method for quantifying the seasonal variation characteristics of wind speed in different regions based on sorting entropy according to claim 8, characterized in that, In step (5): (5-1) Comprehensive analysis: Compare and contrast the spatial distribution maps obtained in steps (2-3) and (4-3) to analyze the differences in the fine structure characterization of seasonal wind speed variations between the two methods: "only statistical average" and "considering the sorting structure". (5-2) Comparative analysis and research were conducted at different sites and regions to comprehensively obtain the seasonal variation characteristics of wind speed in different regions of China, and to analyze the uncertainty differences in seasonal wind speed variation in different regions and the influencing factors behind them.

Images