An entropy-based method for quantitatively characterizing the diurnal variation of wind speed in the hundred-li wind region of Xinjiang
By using an entropy-based approach and analyzing wind speed data with information entropy and sorting entropy, the problem of characterizing the daily variation of wind speed in the Hundred-Mile Wind Zone of Xinjiang was solved. This approach enabled accurate characterization and detailed capture of daily wind speed variations, thereby improving the accuracy of wind disaster forecasting and early warning.
Patent Information
- Application Number
- CN202310106069.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-02-10
- Publication Date
- 2025-11-25
- Estimated Expiration
- 2043-02-10
AI Technical Summary
Existing technologies are insufficient to accurately depict the daily variation characteristics of wind speed in the Xinjiang Hundred-Mile Wind Zone, resulting in inadequate wind disaster forecasting and early warning services, which cannot meet the needs of sudden strong winds.
An entropy-based approach is used to quantitatively characterize the daily variation of wind speed in a 100-mile wind zone through preprocessing of wind speed observation data, calculation of information entropy and sorting entropy, including wind speed level matching, information entropy calculation and sorting entropy analysis, and to reconstruct wind speed sequences to capture the details of the changes.
It has achieved a precise depiction of the daily variation of wind speed in a 100-mile wind zone, revealing the fluctuations and details of wind speed changes, and improving the accuracy and service capabilities of wind forecasting and early warning.
Smart Images
Figure CN116186347B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of wind speed diurnal variation characteristics quantification. Specifically, it is a method for quantitatively describing the wind speed diurnal variation characteristics in the hundred-mile wind area of Xinjiang based on entropy. BACKGROUND
[0002] The hundred-mile wind area is the strongest wind area along the railway in China, located in the transitional zone between the East Tianshan Uplift and Hami Depression north of 40°N. It refers to the section from Hongqikan to Liangdun station of the Lanxin Railway in Xinjiang, with a total length of about 123 km. The terrain of the wind area is high in the north and low in the south, with the Qijiaojing gap between Bogda Mountain and Balkun Mountain in the north. The trumpet-shaped valley is mainly composed of low hills, stone erosion plains, and piedmont alluvial sand and gravel, showing a dry land surface state.
[0003] The annual average number of days with gale of eight or above reaches 200 days, with characteristics of high wind speed, long wind period, strong seasonality, stable wind direction, and fast wind speed. The wind power is the largest in the whole region, and it is also characterized by rapid outbreak of gale and strong low-level wind speed. Since the Lanxin Railway was laid in 1959 and passed through the hundred-mile wind area, it has been threatened by gale disasters. Strong winds cause train delays and even overturning. Gale disasters have become the main meteorological disasters that seriously affect the safe transportation in the hundred-mile wind area.
[0004] The Shisanjianfang National Meteorological Observation Station established in the central section of the hundred-mile wind area has accumulated accurate, timely, stable, reliable, and continuous wind observation data for the research on the characteristics of the wind in the hundred-mile wind area. It is an important data source for the research on the wind in the hundred-mile wind area and plays an important role in the research on the prediction and early warning services and disaster prevention of gale in the hundred-mile wind area. Especially since 2005, with the automation of ground meteorological observation, the time resolution of wind observation data in the hundred-mile wind area has been continuously improved, providing more and better basic research data for the accurate prediction and fine service of local wind.
[0005] However, the wind observation data obtained by the meteorological observation station business software is only calculated by averaging the existing data, picking the maximum wind speed, maximum wind speed and other methods. The focus is mainly on the causes of strong wind in the 100-mile wind area and the monthly variation of wind speed or the characteristics of single weather process. Due to the special terrain conditions of the 100-mile wind area and the complex land-air interaction, the scientific problems in the distribution, evolution process and abnormal change of ground wind are very prominent. In the face of strong wind speed, it is difficult to capture the gustiness change and master the fine structure characteristics of the wind by recording data with 6-hour interval; The analysis data obtained cannot accurately describe the diurnal variation characteristics of the 100-mile wind area, and does not involve the details of the wind speed variation process. But the wind disaster is a sudden natural disaster, and the intermittent is strong. Therefore, the previous research is difficult to meet the current needs of meteorological technology services such as wind forecast and warning in the 100-mile wind area. In view of this problem, it is necessary to develop an objective and effective method to describe the diurnal variation and detailed characteristics of the 100-mile wind area in Xinjiang, so as to express and accurately describe the diurnal variation of wind speed in the 100-mile wind area in the most intuitive way under the strong interference of strong wind, and meet the different application needs. SUMMARY
[0006] Therefore, the technical problem to be solved by the present application is to provide a quantitative description method of diurnal variation characteristics of wind speed in the 100-mile wind area in Xinjiang based on entropy, in order to solve the problem that there is no detailed description and analysis of the diurnal variation characteristics in the 100-mile wind area in the prior art, and it is difficult to meet the current needs of meteorological technology services such as wind forecast and warning in the 100-mile wind area.
[0007] To solve the above technical problems, the present application provides the following technical scheme:
[0008] A quantitative description method of diurnal variation characteristics of wind speed in the 100-mile wind area in Xinjiang based on entropy, comprising the following steps:
[0009] Step P1: acquisition and preprocessing of wind speed observation data;
[0010] Step P2: information entropy distribution of different times and different grades of wind speed in the 100-mile wind area;
[0011] Step P3: analysis of diurnal variation distribution characteristics of wind speed in the 100-mile wind area;
[0012] Step P4: reconstruction of wind speed sequence in the 100-mile wind area and calculation of sorting entropy;
[0013] Step P5: detailed description of each stage change in the diurnal variation process of wind speed in the 100-mile wind area;
[0014] Step P6: comprehensive analysis of information entropy results and sorting entropy results, and quantitative description of diurnal variation and detailed characteristics of wind speed in the 100-mile wind area.
[0015] The wind speed observation data in step P1 includes 2-minute average wind speed V n , daily maximum wind speed V max , and daily maximum wind speed V maxi . V n represents the 2-minute average wind speed value at the Nth hour, n = 21, 22, 23, 0, 1,..., 20.
[0016] The above-mentioned quantitative characterization method of wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang based on entropy includes the following steps:
[0017] Step P101: selecting wind speed observation data observed by a meteorological station in the hundred-li wind area in a certain time period;
[0018] Step P102: arranging 2-minute average wind speed sequences of each day and each time in the time period of the wind speed observation data, and calculating average wind speeds of four different time periods, i.e., morning, afternoon, day, and night.
[0019] The above-mentioned quantitative characterization method of wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang based on entropy includes the following steps:
[0020] Step P201: matching wind force grades for the 2-minute average wind speed sequences of each day and each time obtained in step P102;
[0021] Step P202: according to the wind speed grades of each day and each time determined in step P201, respectively counting the number of times of occurrence of different wind speed grades of each day and each time;
[0022] Step P203: calculating the wind force grade information entropy of the first time;
[0023] Step P204: repeating step P203 to form a wind force grade information entropy sequence of 24 times, and recording and classifying the intermediate process results in detail.
[0024] In step P201 of the above-mentioned quantitative characterization method of wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang based on entropy, according to the wind speed range corresponding to each wind force grade of 0-12 grades in the wind force grade table of the Ground Meteorological Observation Specification, the wind speed grade of each day and each time is matched, which is R n , and n is the wind force grade corresponding to 0-12 grades.
[0025] In step P203 of the above-mentioned quantitative characterization method of wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang based on entropy, the calculation formula of the wind force grade information entropy is:
[0026]
[0027] In formula (1), P(x i) is the probability of the occurrence of a certain event; n is the wind force level, and takes values from 0 to 12.
[0028] The step P3 of the quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-kilometer wind area in Xinjiang based on entropy specifically comprises the following steps:
[0029] Step P301: Statistics are made on the time of occurrence of the ground daily maximum wind speed V max and the ground daily maximum wind speed V maxi Step P301: Statistics are made on the time of occurrence of the ground daily maximum wind speed V max and the ground daily maximum wind speed V maxi Step P301: Statistics are made on the time of occurrence of the ground daily maximum wind speed V
[0030] Step P302: According to the information entropy result of the wind force level, the wind speed diurnal variation characteristics of the hundred-kilometer wind area are characterized.
[0031] The step P4 of the quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-kilometer wind area in Xinjiang based on entropy specifically comprises the following steps:
[0032] Step P401: The 2-minute average wind speed sequence of each day and each time obtained in step P102 is reconstructed to obtain a reconstructed 2-minute average wind speed sequence of each day and each time;
[0033] Step P402: The reconstructed components of the reconstructed 2-minute average wind speed sequence of each day and each time are rearranged in ascending order to obtain a group of symbol sequences; and the probability of occurrence of each kind of ordering structure of the symbol sequence is calculated.
[0034] Step P403: The ordering entropy value of all the reconstructed wind speed time sequences of each day and each time is calculated.
[0035] In the step P401 of the quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-kilometer wind area in Xinjiang based on entropy, the 2-minute average wind speed sequence of each day and each time obtained in step P102 is denoted as {x i}; i=1, 2, …, N; and the reconstructed 2-minute average wind speed sequence X(i) of each day and each time is:
[0036] X(i)=[x(i),x(i+1),…,x(i+(D-1))] (2);
[0037] In formula (2), D is the embedding dimension;
[0038] In step P402, the m reconstructed components x(i), x(i+1), …, x(i+(D-1)) of the reconstructed 2-minute average wind speed sequence X(i) of each day and each time are rearranged in ascending order to obtain a group of symbol sequences A(g), A(g)=[j1,j2,…,j Dwhere g = 1, 2, …, k, and k ≤ D!; D different symbols j1, j2, …, j D There are D! different sorting structures, and the probability P1, P2, …, P D！ .
[0039] The above-mentioned entropy-based quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang, the sorting entropy value H s [P] is defined as:
[0040]
[0041] In formula (3), i = 1, 2, …, N; Pi is the probability of the occurrence of the i-th sorting structure; 0 ≤ H s [P] ≤ 1.
[0042] The technical scheme of the present application achieves the following beneficial technical effects:
[0043] 1. The entropy-based quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang of the present application quantitatively analyzes the wind speed data observed by the stations in the hundred-li wind area through entropy value sequences, and characterizes the diurnal variation characteristics and details of the wind speed in the hundred-li wind area.
[0044] 2. The present application quantitatively characterizes the diurnal variation characteristics of the wind speed in the hundred-li wind area based on the information entropy analysis method, and quantitatively characterizes the detailed characteristics of each stage in the diurnal variation process of the wind speed in the hundred-li wind area by using sorting entropy. By virtue of the advantages of information entropy and sorting entropy analysis methods in processing time series, the fluctuation of the diurnal variation of the wind speed in the hundred-li wind area and the intermittent characteristics of the wind speed in different stages of the diurnal variation are revealed.
[0045] 3. The entropy-based quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang of the present application gives a complete algorithm for expressing the diurnal variation of the wind speed in the hundred-li wind area and the detailed characterization thereof. By using the calculation results of the information entropy and sorting entropy of different time 2-minute average wind speeds, the detailed characterization of the wind speed in the hundred-li wind area is given from different angles, and the characteristics of the wind speed in the hundred-li wind area are comprehensively characterized through the corresponding relationship between the two groups of data. The diurnal variation and the detailed characterization method given by the present application can more comprehensively display the diurnal variation characteristics of the wind speed in the hundred-li wind area, and effectively capture the details of the diurnal variation characteristics. BRIEF DESCRIPTION OF DRAWINGS
[0046] Figure 1 The flowchart of the entropy-based quantitative characterization method of the wind speed diurnal variation characteristics of the hundred-li wind area in Xinjiang in the embodiments of the present application is shown.
[0047] Figure 2The information entropy expression information schematic diagram of the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0048] Figure 3 The sequencing entropy 24 structure schematic diagram of the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0049] Figure 4 The sequencing entropy expression information schematic diagram of the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0050] Figure 5 The information entropy and sequencing entropy daily variation result comprehensive analysis schematic diagram of the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0051] Figure 6 The wind speed daily variation detail schematic diagram for selecting the continuous gale day number in spring by using the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0052] Figure 7 The wind speed daily variation detail schematic diagram for selecting the continuous gale day number in summer by using the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0053] Figure 8 The wind speed daily variation detail schematic diagram for selecting the continuous gale day number in autumn by using the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application;
[0054] Figure 9 The wind speed daily variation detail schematic diagram for selecting the continuous gale day number in winter by using the quantitative characterization method of the entropy-based daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang in the embodiment of the application. DETAILED DESCRIPTION
[0055] 1. Data and methods
[0056] 1.1 Study data
[0057] The 2-minute average wind speed, daily maximum wind speed and daily maximum wind speed observation data of the thirteen rooms national meteorological station in the hundred-li wind area from 2005 to 2020 used in the embodiment come from the Xinjiang Uygur Autonomous Region Meteorological Information Center, and the data has undergone strict quality control, such as main change range check, internal consistency check, time consistency check, spatial consistency check and the like. The daily variation and its detail characterization method in the embodiment is based on the wind speed data after quality control, so as to ensure the accuracy of the data information obtained after the entropy value method processing.
[0058] 1.2 Research methods
[0059] As Figure 1 shown, the embodiment based on the quantitative characterization method of the entropy of the daily variation characteristics of the wind speed in the hundred-li wind area of Xinjiang includes the following steps:
[0060] Step P1: acquisition and preprocessing of basic data;
[0061] Step P101: Establishing wind speed observation data sequence, selecting 2-minute average wind speed, maximum wind speed and maximum wind speed of each day of the thirteen rooms national meteorological observation station in the hundred-li wind area from January 1, 2005 to December 31, 2020.
[0062] Step P102: V n represents the wind speed value of the Nth hour (n=21, 22, 23, 0, 1…20), such as: V 21 21 o'clock 2-minute average wind speed. Organize the 2-minute average wind speed sequence of each day and each time period within the data period, and calculate the average wind speed of different time periods such as morning, afternoon, day and night;
[0063] V max represents the maximum wind speed value of a certain day, which refers to the maximum 10-minute average wind speed value occurring within a certain time period. Organize the maximum wind speed sequence of each day within the data period. Organize the maximum wind speed sequence from January 1, 2005 to December 31, 2020.
[0064] V maxi represents the maximum wind speed value of a certain day, which refers to the maximum 10-minute average wind speed value occurring within a certain time period. Organize the maximum wind speed sequence of each day within the data period. Organize the maximum wind speed sequence from January 1, 2005 to December 31, 2020.
[0065] Step P2: Information entropy distribution of different time and different grade wind speed in the hundred-li wind area;
[0066] Step P201: Wind power grade matching of the 2-minute average wind speed sequence of each day and each time period obtained by step P102.
[0067] According to the wind power grade table of "Ground Meteorological Observation Specification", the wind power is divided into 13 grades (0-12 grades), and the corresponding wind speed range is shown in Table 1.
[0068] Table 1 Wind power grade table
[0069]
[0070] Build function: IF (wind speed value >= 32.7, "12", IF (wind speed value >= 28.5, "11", IF (wind speed value >= 28.5, "11", IF (wind speed value >= 24.5, "10", IF (wind speed value >= 20.8, "9", IF (wind speed value >= 17.2, "8", IF (wind speed value >= 13.9, "7", IF (wind speed value >= 10.8, "6", IF (wind speed value >= 8, "5", IF (wind speed value >= 5.5, "4", IF (wind speed value >= 3.4, "3", IF (wind speed value >= 1.6, "2", IF (wind speed value >= 0.3, "1", "0")).
[0071] Match each day each time wind speed level one by one, each wind speed value corresponds to an R n , n is the wind force level corresponding to 0-12. Complete all observation days all time wind force level matching.
[0072] Step P202: for each day each time wind speed level determined in step P201, build 0-12 level wind each time occurrence frequency statistics function, respectively statistics 0-24 different level wind occurrence frequency. Record the results of the process for saving.
[0073] Step P203: calculate the first time wind force level information entropy. According to the information entropy formula:
[0074]
[0075] In which, P(x i ) is the probability of the occurrence of an event. N is the wind force level, its value is 0-12.
[0076] Step P204: repeat step P203, form 24 times wind force level information entropy sequence, and the intermediate process results are recorded in detail and classified.
[0077] Step P3: analysis of the distribution characteristics of the wind speed in the 100 miles wind area;
[0078] Step P301: build the frequency statistics function for the maximum wind speed and the maximum wind speed occurrence time of each day in the 100 miles wind area meteorological observation station, and count the occurrence frequency of each time maximum wind speed and maximum wind speed.
[0079] Step P302: as Figure 2 shown, the information entropy results of the above each level wind speed are statistically analyzed to depict the wind speed daily variation characteristics in the 100 miles wind area. At the same time, the results are compared with the traditional statistical hourly average wind speed, the occurrence frequency of each time maximum wind speed and the occurrence frequency of each time maximum wind speed (the above four kinds of data are processed by dimensionless).
[0080] Step P4: Reconstruction of wind speed sequence and calculation of sorting entropy in the 100-mile wind zone;
[0081] Step P401: Process the wind speed time series {x} from step P101. i};i=1,2,…,N, First, the sequence is reconstructed to obtain the reconstructed 2-minute average wind speed sequence X(i);
[0082] X(i)=[x(i),x(i+1),…,x(i+(D-1))];
[0083] In the above formula, D is the embedding dimension.
[0084] Step P402: Rearrange the m reconstructed components x(i), x(i+1), ..., x(i+(D-1)) of the reconstructed 2-minute average wind speed sequence X(i) in ascending order to obtain a symbol sequence A(g), A(g) = [j1,j2,...,j...]. 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！ Here, D is set to 4, such as... Figure 3 As shown, there are 24 sorting structures.
[0085] Step P403: As Figure 4 As shown, the number of occurrences of all permutations at each time step is calculated, and the sorting entropy value of the reconstructed wind speed time series at each time step is calculated, similar to the form of information entropy. Then the sorting entropy H is... s [P] is defined as:
[0086]
[0087] Where i = 1, 2, ..., 24; Pi is the probability of appearing in the i-th sorted structure; 0 ≤ H s [P]≤1.
[0088] When H s When [P] = 1, it corresponds to a completely random system, in which all 24 sorting structures have the same probability of appearing.
[0089] If the reconstructed 2-minute average wind speed sequences contain a specific organizational structure, then H s The value of [P] will be less than 1.
[0090] Step P5: Detailed depiction of the daily wind speed variations in the 100-mile wind zone at each stage;
[0091] like Figure 4The sorting entropy results of the wind speed at each time in step P4 are statistically analyzed to depict the details of the wind speed diurnal variation in the 100-kilometer wind area. When the sorting entropy value is significantly lower than other values, it means that the wind speed variation law in the variation process is characterized by the sorting structure.
[0092] Step P6: The information entropy result and the sorting entropy result are fused to comprehensively express the wind speed diurnal variation and the details of the variation in the 100-kilometer wind area.
[0093] As Figure 5 shown, the information entropy and the sorting entropy data sequence of the 2-minute average wind speed at different times are analyzed to analyze the wind speed characteristics and the variation characteristics at each time, and to comprehensively express the wind speed diurnal variation law and the details of the variation in the 100-kilometer wind area.
[0094] 2 Results and analysis
[0095] 2.1 Wind speed diurnal variation characteristics in the 100-kilometer wind area
[0096] The data of this embodiment are selected from the 2-minute average wind speed, the daily maximum wind speed and the daily maximum wind speed data sequence observed by the thirteen-room meteorological station from 2005 to 2020. The data source is the Xinjiang Uygur Autonomous Region Meteorological Information Center, and the data quality is strictly controlled.
[0097] The statistical results show that the diurnal variation laws of the average wind speed, the maximum wind speed and the maximum wind speed in the 100-kilometer wind area are different, and the peak and valley values displayed by each index appear at different times. The high values of the average wind speed and the maximum wind speed appear in the early morning to the morning, and the most frequent time of the maximum wind speed appears in the afternoon. The wind speed extreme value appears frequently in the morning and afternoon, and the wind speed is in a stable change period. Figure 2 The traditional statistical method is shown in the table, which needs to be integrated and analyzed on the basis of multiple statistical results of multiple elements. The method of the present application can directly and completely reflect the high value, the extreme value appearance time and the diurnal variation characteristics of the wind speed in the 100-kilometer wind area.
[0098] 2.2 Sorting entropy analysis of the wind speed diurnal variation and the details of the variation in the 100-kilometer wind area
[0099] By reconstituted after the 100 Li wind area each time 2 minutes average wind speed, analysis of different change details in each time distribution, the results show that the wind speed from large to small, from small to large this law change distribution all day has obvious peak and valley characteristics, wind speed by small increase trend (with the order structure "1234" more typical representative) distribution of the most widely for 20-0 hours, wind speed by large to small trend distribution of the most widely for 8-13 hours in the morning and 17-20 hours in the afternoon (with the order structure "4321" more typical representative). The order entropy results and information entropy analysis results complement each other - the information entropy value of wind speed at a certain time of day is smaller, which reflects the larger probability of the occurrence of the maximum or minimum value of wind speed, while the smaller order entropy value at a certain time of day reflects the larger probability of the occurrence of the order structure of wind speed increasing from small to large (such as "1234") or decreasing from large to small (such as "4321"). Therefore, through these two forms of entropy, the characteristics of the daily variation of wind speed in the 100 Li wind area of Xinjiang can be well quantitatively described. The distribution of wind speed is relatively average throughout the day. That is, during the time period when non-extreme values occur, the wind speed obviously shows regular changes, or a stable increase or a stable decrease trend; but during the time period when extreme values such as maximum wind speed and maximum wind speed occur, the hourly continuity of wind speed is significantly weakened, the burst characteristics are highlighted, and the intermittency is strong.
[0100] Figure 4 It can be seen that the rapid decrease trend (the proportion of the order structure "4321" is significantly higher than that of other order structures) from 9 to 11 o'clock is significant, and the rapid increase trend (the proportion of the order structure "1234" is significantly higher than that of other order structures) from 20 to 23 o'clock is significant; at the same time, the distribution structure with ascending and descending indication in other order structures is also more regular than the disordered arrangement structure.
[0101] Figure 5 It can be seen that the average wind speed in the morning is always high, and the wind speed in the afternoon is obviously enhanced, which is consistent with the situation displayed by the information entropy in step P3.
[0102] Figures 6 to 9 It can be seen that the four gale weather processes in the four seasons all have a common point, that is, there are two wind speed aggregation points around 9 o'clock and 20 o'clock, indicating that the wind has a significant increase (the proportion of the order structure "1234" is significantly higher than that of other order structures) or decrease (the proportion of the order structure "4321" is significantly higher than that of other order structures) in these two time periods, which is consistent with the Figure 4 expressive meaning. That is, the wind speed has a significant high-low change rule in these two time periods, so the order entropy is low, and the other time is not obvious in the change rule, indicating that the wind speed changes intermittently and the data dispersion is high, so the order entropy is high. This conclusion has been well expressed in Figure 5 .
[0103] 3. Conclusion and discussion
[0104] The embodiment is a simple and effective quantitative description method for the diurnal variation of the Baili wind area and its detailed changes. By matching the hourly average wind speed for many years, reconstructing the structure, calculating the wind speed grade information entropy at each time of the day, and describing the diurnal variation of the wind speed in the Baili wind area, the sorting entropy gives the detailed change characteristics of the wind speed in the corresponding period, and the extreme value occurrence time period and the continuous wind speed in other periods are described in detail.
[0105] The entropy value expression method proposed in the embodiment shows that the diurnal variation of the wind speed in the Baili wind area and the intermittent changes between each time have certain distribution rules, and the entropy analysis can realize the intuitive display of the diurnal variation characteristics of the complex wind speed sequence. The method first decomposes the hourly average wind speed, the maximum wind speed, and the extreme wind speed occurrence time in the Baili wind area, reconstructs the intermittent structure of the wind speed at each time, and realizes the quantitative description of the diurnal variation characteristics of the wind speed. In addition, the embodiment gives a description method for the diurnal variation of the wind speed in the Baili wind area and its detailed characteristics. By the different performances of the information entropy and the sorting entropy, the characteristics of the wind speed at each time in the wind area are determined, which can better serve the disaster warning and forecasting. In the following research, different structure reconstruction modes can be combined for individual case analysis of specific weather processes, and the data sequences of different D values and different time scales can be compared in more detail, which can better reflect the practical application effect of the entropy value method in the fine structure description of the wind speed.
Claims
1. A method for quantitatively characterizing the diurnal variation of wind speed in the Xinjiang Hundred-Li Wind Region based on entropy, characterized in that, It comprises the following steps: Step P1: obtaining and preprocessing the wind speed observation data of the Baili wind area; Step P2: information entropy distribution of different time and different grade wind speed in Baili wind area; Step P3: analysis of the distribution characteristics of wind speed daily variation in Baili wind area; Step P4: reconstruction of wind speed sequence in Baili wind area and calculation of ordering entropy; Step P5: detailed description of the change of each stage in the wind speed daily variation process in Baili wind area; Step P6: comprehensive analysis of information entropy results and ordering entropy results, and quantitative description of the wind speed daily variation and its detailed characteristics in Baili wind area; Step P1 specifically comprises the following steps: Step P101: selecting the wind speed observation data observed by the meteorological station in Baili wind area in a certain time period; Step P102: arranging the 2-minute average wind speed sequence of each day and each time in the time period of the wind speed observation data, and calculating the average wind speed of four different time periods, i.e. morning, afternoon, day and night; Step P2 specifically comprises the following steps: Step P201: matching the wind power grade of the 2-minute average wind speed sequence of each day and each time obtained in step P102; Step P202: according to the wind speed grade of each day and each time determined in step P201, the number of different wind speed grades appearing in each day and each time is counted respectively; Step P203: calculating the information entropy of the first time wind power grade; Step P204: repeating step P203 to form a sequence of wind power grade information entropy of 24 times, and recording and saving the intermediate process results in detail and classifying; Step P3 specifically comprises the following steps: Step P301: Statistics of the ground daily maximum wind speed observed by the meteorological observation station in the Bili wind area V max and the ground daily maximum wind speed V maxi Statistics of the ground daily maximum wind speed and the ground daily maximum wind speed V max and the ground daily maximum wind speed V maxi the number of times Step P302: according to the wind power grade information entropy result, the wind speed daily variation characteristics of Baili wind area are described; Step P4 specifically comprises the following steps: Step P401: sequence reconstruction is performed on the 2-minute average wind speed sequence of each day and each time obtained in step P102, to obtain the reconstructed 2-minute average wind speed sequence of each day and each time; Step P402: rearranging the reconstruction components of the reconstructed 2-minute average wind speed sequence of each day and each time in ascending order to obtain a group of symbol sequences; Step P403: calculating the ordering entropy value of all reconstructed wind speed time series of each day and each time. In step P203, the calculation formula of wind power grade information entropy is:
2. The method according to claim 1, wherein, In step P1, the wind speed observation data includes the 2-minute average wind speed at the ground. V n Maximum daily wind speed at ground level V max and the maximum daily wind speed at ground level V maxi ; V n Representing the N The 2-minute average wind speed value at that time. n =21, 22, 23, 0, 1……20. 3.The method according to claim 1, characterized in that, In step P201, according to the wind speed range corresponding to each wind force level of 0~12 levels in the wind force level table of the Ground Meteorological Observation Specification, the wind speed level of each day and each time is matched, and is R n , n the wind force level corresponding to 0~12 levels.
4. The method according to claim 1, wherein, In formula (2), D is the embedding dimension; （1）； In formula (1), P x i is the probability of occurrence of a certain event; n is the wind force level, taking a value of 0 ~ 12. 5. The method according to claim 1, wherein, In step P401, the 2-minute average wind speed sequence of each day and each time obtained in step P102 is denoted as ; and the reconstructed 2-minute average wind speed sequence X(i) of each day and each time is: （2）； In step P402: reconstructing m reconstructed components of the 2-minute average wind speed sequence X(i) of each day and each time in step P401 Re-arranging in ascending order to obtain a set of symbol sequences A(g), A(g)=[j1, j2,…, j D ], where g=1, 2, …, k, and k≤D!; D different symbols j1, j2,…, j D There are D! different sorting structures, and the probability P1, P2, …, P D! of each sorting structure is calculated. 6.The method according to claim 1, wherein, The ranking entropy value H in step P403 s [P] is defined as: （3）； In formula (3), i = 1, 2, …, N; Pi is the occurrence probability of the ith ranking structure; 0≤H s [P]≤1.