Turning meteorological multi-time-scale multi-dimensional feature analysis method, system and device based on historical data and medium

Abstract

The invention discloses a turning meteorological multi-time-scale multi-dimensional feature analysis method, system and device based on historical data and a medium, and belongs to the technical field of electric power system meteorological data analysis, and the method comprises the steps: obtaining historical meteorological data; setting a flag bit for each meteorological parameter according to a preset judgment condition, and identifying whether a turning weather occurs in a time period; calculating a daily time scale feature, a monthly time scale feature and an annual time scale feature based on the flag bit, wherein the monthly time scale feature comprises a time period statistical feature and a day number statistical feature; and outputting the multi-time scale features. According to the invention, analysis is carried out directly from historical meteorological data, dependence on new energy output data is avoided, and original features of turning weather can be reflected objectively; a day-month-year multi-time scale hierarchical feature analysis system is constructed, and accurate support is provided for resource scheduling decisions of power scheduling personnel on different time scales.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system weather data analysis, and in particular relates to a turning weather multi-time scale multi-dimension feature analysis method, system, device and medium based on historical data. BACKGROUND

[0002] With the intensification of global warming, short-time gale, short-time wind speed fluctuation, short-time strong light, long-time rainfall and long-time drought and other dramatic fluctuation type turning weather events occur frequently. The grid-connected power of wind power and photovoltaic units is significantly positively correlated with wind energy and light energy in short time scale, and rainfall process directly affects the reservoir inflow of hydropower stations. There have been many typical cases in reality that turning weather has impacted new energy output and safe operation of power systems. Frequent turning weather has seriously threatened stable output of new energy units and safe operation of power systems. SUMMARY

[0003] In view of the above problems, the present application provides a turning weather multi-time scale multi-dimension feature analysis method, system, device and medium based on historical data.

[0004] Therefore, the technical problem solved by the present application is that only through analysis of new energy output fluctuation, the feature of turning weather itself cannot be analyzed, and the analysis effect is not good for new energy units with strong dispatchability and long time scale such as hydropower; the types of weather factors involved in the analysis method are relatively limited.

[0005] To solve the above technical problems, the present application provides the following technical scheme: a turning weather multi-time scale multi-dimension feature analysis method based on historical data, comprising: obtaining historical weather data, the historical weather data comprising values of multiple weather parameters in multiple time periods of multiple historical dates; For each weather parameter, a flag bit is set for each time period of each historical date according to a preset determination condition, for identifying whether turning weather occurs in the time period; Based on the flag bit, a daily time scale feature is calculated, the daily time scale feature being obtained by quantifying feature indicators of turning weather occurring in each time period within a day; Based on the flag bit, a monthly time scale feature is calculated, the monthly time scale feature comprising time period statistical feature and day number statistical feature, wherein the time period statistical feature reflects the situation of turning weather occurring in each time period within a month, and the day number statistical feature reflects the cumulative situation of turning weather occurring in a month; Based on the monthly time scale features of each month, an annual time scale feature is calculated; The daily time scale feature, the monthly time scale feature and the annual time scale feature are output.

[0006] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, wherein: the setting of the flag bit for each time period of each historical date according to the preset determination condition comprises determining a turning determination type corresponding to the weather parameter, and the turning determination type comprises a determination type based on time sequence change and a determination type based on numerical value range. The corresponding determination condition is set according to the turning determination type. The weather parameter value of each time period of the historical date is compared with the determination condition, and a binary flag bit is set for the time period.

[0007] The beneficial effect of the preferred technical scheme is that by distinguishing the determination type based on time sequence change and the determination type based on numerical value range, the turning characteristics of different weather parameters can be identified differently. The time sequence change type is suitable for capturing rapid fluctuation phenomena such as wind speed mutation and temperature mutation, and the numerical value range type is suitable for identifying over-limit states such as gale and high temperature, thereby improving the accuracy of turning weather identification.

[0008] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, wherein: the calculation of the daily time scale feature based on the flag bit comprises traversing all historical dates in the historical weather data, and extracting the flag bit corresponding to each time period of each date. For each time period within a day, the number of flag bits that meet the turning identification state in all historical dates is counted. The number is compared with the total number of historical dates to obtain the turning occurrence frequency of the time period.

[0009] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, wherein: the time period statistical feature comprises counting, for each time period in each month, the number of flag bits that meet the turning identification state in all dates in the month. The number is compared with the total number of dates in the month to obtain the turning occurrence frequency of the time period in the month. The day number statistical feature is obtained by the following steps: for each month, it is determined whether there is at least one time period whose flag bit meets the turning identification state in each date in the month, and the number of dates that meet the condition is counted.

[0010] The beneficial effects of the preferred technical scheme are as follows: through the double-dimension statistics of the time period statistical characteristics and the day number statistical characteristics, the occurrence frequency distribution law of the turning weather in each time period in a month can be reflected, and the cumulative day number condition of the turning weather in the month can be reflected.

[0011] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, when the turning determination type is the determination type based on time sequence change, the determination condition is that the absolute value of the difference between the meteorological parameter value of the current time period and the meteorological parameter value of the adjacent time period exceeds the first preset threshold value. When the turning determination type is the determination type based on numerical value range, the determination condition is that the meteorological parameter value of the current time period exceeds the second preset threshold value.

[0012] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, the annual time scale feature based on the monthly time scale features of each month comprises the following steps: extracting the time period statistical characteristics of each month in the order of months, and splicing the time period statistical characteristics of each month to form an annual time period statistical matrix, wherein the first dimension of the matrix corresponds to the month, and the second dimension corresponds to the time period in a day. According to the order of months, the day number statistical characteristics of each month are extracted, and the day number statistical characteristics of each month are spliced to form an annual day number statistical vector.

[0013] As a preferred scheme of the turning weather multi-time scale multi-dimension feature analysis method based on historical data, the plurality of meteorological parameters comprises wind speed, rainfall intensity, temperature and light intensity. For wind speed, the turning weather comprises wind speed mutation and strong wind. For rainfall intensity, the turning weather comprises excessive rainfall and insufficient rainfall. For temperature, the turning weather comprises temperature mutation and high temperature. For light intensity, the turning weather comprises light fluctuation and strong light.

[0014] The application provides a turning weather multi-time scale multi-dimension feature analysis system based on historical data.

[0015] To solve the above technical problems, the application provides the following technical scheme: a turning weather multi-time scale multi-dimension feature analysis system based on historical data, comprising: a data acquisition module, configured to acquire historical meteorological data, wherein the historical meteorological data comprises numerical values of a plurality of meteorological parameters in a plurality of time periods on a plurality of historical dates. The flag setting module is used to set flags for each historical date and each time period for each meteorological parameter according to preset judgment conditions, so as to indicate whether a turning point in the weather occurs during that time period; The daily feature calculation module is used to calculate the daily time scale features based on the flag bit. The daily time scale features are obtained by statistically analyzing the feature quantification indicators of the weather turning points in each time period of the day. The monthly feature calculation module is used to calculate monthly time scale features based on the flag bits. The monthly time scale features include time period statistical features and day statistical features. The time period statistical features reflect the weather changes that occur in each time period within the month, and the day statistical features reflect the cumulative weather changes that occur within the month. The annual feature calculation module is used to calculate annual time scale features based on the monthly time scale features of each month. The output module is used to output the daily timescale features, monthly timescale features, and yearly timescale features.

[0016] The present invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the method for analyzing the multi-timescale and multi-dimensional features of turning point weather based on historical data.

[0017] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the method for analyzing the multi-timescale and multi-dimensional features of turning point weather based on historical data.

[0018] The beneficial effects of this invention are: it analyzes historical meteorological data directly without relying on new energy output data, and can objectively reflect the original characteristics of changing weather conditions. It is particularly suitable for new energy units such as hydropower, which have strong dispatchability and weak correlation between output and meteorological factors.

[0019] Differentiated judgment methods are established to address the transitional characteristics of different meteorological factors, covering various meteorological parameters such as wind speed, rainfall intensity, temperature, and light intensity, which can comprehensively capture various transitional meteorological features.

[0020] A hierarchical feature analysis system with multiple time scales from day to month to year is constructed to provide precise support for power dispatchers to make resource scheduling decisions at different time scales, thereby ensuring the safe and stable operation of the power system. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 The following is an overall flowchart of a method for analyzing the multi-timescale and multi-dimensional characteristics of turning point meteorological events based on historical data, provided as an embodiment of the present invention. Detailed Implementation

[0023] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0024] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for analyzing the multi-timescale, multi-dimensional features of meteorological turning points based on historical data, including: Step 1: Obtain historical meteorological data, which includes the values ​​of various meteorological parameters at multiple time periods on multiple historical dates; Step 2: For each meteorological parameter, set a flag for each time period of each historical date according to the preset judgment conditions to indicate whether a turning point in the weather occurred during that time period; Step 3: Calculate the daily timescale characteristics based on the flag, which are obtained by statistically analyzing the characteristic quantification indicators of weather changes occurring in each time period of the day; Step 4: Calculate the monthly time scale features based on the flag bits. The monthly time scale features include time period statistical features and day statistical features. The time period statistical features reflect the weather changes that occur in each time period within the month, and the day statistical features reflect the cumulative weather changes that occur within the month. Step 5: Calculate the annual time scale features based on the monthly time scale features of each month; Step 6: Output the daily timescale features, monthly timescale features, and yearly timescale features.

[0025] In power system dispatching, transitional weather significantly impacts the output of renewable energy units. However, existing analysis methods primarily rely on renewable energy output data for inference, failing to directly capture the original characteristics and patterns of transitional weather. This is particularly problematic for dispatchable units like hydropower, whose output is heavily influenced by human dispatch, rendering output-based weather analysis ineffective. This embodiment addresses this by directly analyzing historical meteorological data, setting judgment conditions and establishing a flag mechanism for each meteorological parameter to objectively identify the timing of transitional weather events. This method does not rely on renewable energy output data but rather on the temporal variation and numerical range characteristics of the meteorological parameters themselves. This allows for accurate capture of the occurrence patterns of various transitional meteorological phenomena such as sudden wind speed changes, abnormal rainfall, sudden temperature changes, and fluctuations in sunlight, solving the problems of poor analysis results for dispatchable units and limited meteorological factor types inherent in traditional methods.

[0026] This embodiment constructs a hierarchical analysis system across multiple time scales (day-month-year) to comprehensively reveal the occurrence patterns of transitional weather at different time scales. At the daily time scale, by statistically analyzing the frequency of transitions in each time period, high-incidence periods of transitional weather within a day are identified, providing refined guidance for intraday dispatching. At the monthly time scale, dual statistics from both the time period and day dimensions reflect both the distribution of transitional frequency across different time periods within a month and the cumulative intensity of transitional weather within a month, providing a comprehensive basis for monthly dispatching plans. At the annual time scale, by splicing monthly features to form matrices and vectors, the seasonal variation patterns of transitional weather throughout the year are revealed, providing support for medium- and long-term resource planning. This method achieves a complete transformation from historical data to multi-scale features, providing power dispatchers with objective and accurate decision-making basis for formulating dispatching strategies, allocating resources, and preventing risks at different time dimensions, ensuring the safe and stable operation of the power system under transitional weather conditions.

[0027] Example 2, an embodiment of the present invention, provides a method for analyzing the multi-timescale, multi-dimensional features of turning point meteorological data based on historical data, based on the previous embodiment, including: Furthermore, in step 1: acquire historical meteorological data, which includes various meteorological parameters, including wind speed, rainfall intensity, temperature, and light intensity; regarding wind speed, the transitional meteorological data includes sudden wind speed changes and strong winds; regarding rainfall intensity, the transitional meteorological data includes excessive rainfall and insufficient rainfall; regarding temperature, the transitional meteorological data includes sudden temperature changes and high temperatures; regarding light intensity, the transitional meteorological data includes light intensity fluctuations and strong light.

[0028] Step 2: For each meteorological parameter, a flag is set for each time period of each historical date according to preset judgment conditions to indicate whether a turning point in meteorological conditions has occurred during that time period, including the following steps A1-A3: A1: Determine the turning point judgment type corresponding to the meteorological parameter. The turning point judgment type includes a judgment type based on time series changes and a judgment type based on numerical range. A2: Set corresponding judgment conditions according to the aforementioned transition judgment type; A3: Compare the meteorological parameter values ​​for each time period of the historical date with the determination conditions, and set a binary flag bit for the time period.

[0029] In this embodiment of the application, in step 2, the preset judgment condition is determined by: determining the threshold parameter according to the turning point judgment type of the meteorological parameter; when the turning point judgment type is a judgment type based on time sequence change, a first preset threshold is set to determine whether the parameter change between adjacent time periods exceeds the limit; when the turning point judgment type is a judgment type based on numerical range, a second preset threshold is set to determine whether the parameter value of the current time period exceeds the normal range.

[0030] In an optional implementation, in step 2, the preset judgment condition can be achieved by: statistically analyzing the quantiles of meteorological parameters in historical meteorological data, setting the first preset threshold to the 90th quantile of the distribution of differences between adjacent time periods, and setting the second preset threshold to the 95th quantile of the distribution of absolute values ​​of parameters, thereby realizing adaptive threshold determination based on the distribution characteristics of historical data.

[0031] In another optional implementation, in step 2, the preset judgment conditions can also be set by: combining the physical characteristics of meteorological parameters and power system operation experience to set thresholds, and adopting differentiated threshold standards for different meteorological parameters, such as setting the wind speed change threshold to 3 m / s, the strong wind threshold to 15 m / s, the temperature change threshold to 5 degrees Celsius, and the high temperature threshold to 35 degrees Celsius.

[0032] In this embodiment of the application, in step A1, the transition determination type is determined by: analyzing the physical characteristics of meteorological parameters and their impact mechanism on new energy units. For parameters such as wind speed, temperature, and light intensity, rapid fluctuations can lead to drastic changes in the output of new energy units. Therefore, a determination type based on time-series changes is set to identify abrupt changes. At the same time, a determination type based on numerical range is set to identify over-limit states. For rainfall intensity parameters, the time scale is relatively long and the fluctuations are relatively weak. Only a determination type based on numerical range is set to identify excessive rainfall and insufficient rainfall states.

[0033] In an optional implementation, in step A1, the transition determination type can be determined by: dividing the impact of meteorological parameters on the power system according to the time scale; parameters with short time scale impact, such as wind speed and sunlight, are captured by a determination type based on time-series changes to capture rapid fluctuations; parameters with long time scale impact, such as rainfall, are identified by a determination type based on numerical range to identify persistent abnormal states.

[0034] In another optional implementation, in step A1, the transition determination type can also be determined by setting multiple determination type combinations for the same meteorological parameter. For example, for wind speed, a determination type based on time-series changes can be set to identify sudden changes in wind speed, and a determination type based on numerical range can be set to identify strong wind weather, thereby achieving comprehensive identification of different transition characteristics of the same parameter.

[0035] Step 3: Calculate the daily timescale characteristics based on the aforementioned flag. These daily timescale characteristics are obtained by statistically analyzing the characteristic quantification indicators of weather transitions occurring in different time periods within the day, including the following steps B1-B3: B1: Iterate through all historical dates in the historical meteorological data and extract the flag bits corresponding to each time period of each date; B2: For each time period within the day, count the number of times the flag bit of the time period satisfies the transition indicator state across all historical dates; B3: Calculate the ratio of the quantity to the total number of historical dates to obtain the frequency of the turning point in the time period.

[0036] In this embodiment of the application, in step 3, the feature quantification index is obtained by: statistically analyzing the frequency of turning point weather in each time period of the day; for each time period of the day, dividing the number of times the flag bit of that time period is in the turning point status in all historical dates by the total number of historical dates to obtain the turning point frequency of that time period, thereby forming a frequency distribution feature that reflects the pattern of turning point weather occurrence in each time period of the day.

[0037] In an optional implementation, in step 3, the feature quantification index can be obtained by: calculating the probability of the occurrence of turning weather in each time period of the day, normalizing the statistically obtained frequency so that the sum of the probabilities of all time periods is 1, thereby obtaining the probability distribution characteristics of the occurrence of turning weather in the day.

[0038] In another optional implementation, in step 3, the feature quantification index can also be obtained by: calculating the turning point intensity index for each time period within the day; combining the turning point amplitude information with the statistical frequency; extracting the change in meteorological parameters or the degree of exceeding the limit for the time period whose marker position meets the turning point identification state; and calculating the weighted average value as the turning point intensity feature for that time period.

[0039] In this embodiment of the application, in step B3, the frequency of the turning point is obtained by dividing the number of flag bits that meet the turning point status by the total number of days in the historical data S. The calculation formula is the sum of the flag bits divided by S, where S is the total number of days included in the historical meteorological data. This frequency value reflects the historical statistical probability of the turning point weather occurring during a specific period of the day.

[0040] In one alternative implementation, the frequency of turning points can be obtained by: statistically calculating the frequency of turning points separately for different seasons, grouping historical dates into four seasons (spring, summer, autumn, and winter), and independently calculating the frequency of turning points for each time period in each season, thereby revealing the seasonal differences in intraday turning point meteorological patterns.

[0041] In another alternative implementation, the frequency of turning points can also be obtained by dynamically calculating using a sliding window method, selecting historical data from the most recent few years as the calculation window, updating the data within the window over time and recalculating the frequency, thereby obtaining dynamic frequency characteristics that reflect the evolution trend of the weather patterns of turning points.

[0042] Step 4: Calculate the monthly timescale features based on the flag bits. The monthly timescale features include time period statistical features and day count statistical features. The time period statistical features reflect the weather transitions that occur in different time periods within the month, and the day count statistical features reflect the cumulative weather transitions that occur within the month. The time period statistical features include the following steps C1-C3: C1: For each time period within a month, count the number of times the flag bit of the time period satisfies the transition indicator state among all dates in the month; C2: Calculate the ratio of the quantity to the total number of days in the month to obtain the frequency of the transition in the period within the month; C3: The day count statistics feature is obtained through the following steps: For each month, determine whether there is at least one time period flag that satisfies the transition indicator state for each date in the month, and count the number of dates that meet the condition.

[0043] Step 5: Calculating the annual timescale features based on the monthly timescale features of each month includes the following steps D1-D2: D1: Extract the time period statistical features of each month in order of month, and concatenate the time period statistical features of each month to form an annual time period statistical matrix. The first dimension of the matrix corresponds to the month, and the second dimension corresponds to the intraday time period. D2: Extract the statistical features of the number of days in each month in order of month, and concatenate the statistical features of the number of days in each month to form an annual statistical vector of the number of days.

[0044] Example 3, an embodiment of the present invention, provides a method for analyzing the multi-timescale, multi-dimensional features of turning point meteorological data based on historical data, based on the previous embodiment, including: Furthermore, in steps 2 and 3, a time-scale characteristic analysis of the turning point meteorological day is performed, a flag is set for each meteorological parameter, and the frequency of turning point occurrence in each time period of the day is calculated.

[0045] Regarding wind speed parameters, the weather transitions can be categorized into two typical scenarios: sudden wind speed changes and strong winds, both of which significantly impact wind power output. When wind speed changes abruptly, due to the inertia of the wind turbine blades, their rotational speed often lags behind the wind speed change, resulting in a significant deviation between actual and predicted wind power output. In the event of sudden strong winds, to prevent mechanical damage to the blades, the wind turbine will activate its self-protection mechanism, actively shutting off from the grid and drastically reducing turbine output, leading to actual wind power output far below predicted power, thus posing a potential risk to the safe operation of the power system. To analyze the diurnal timescale characteristics of wind speed transition weather, historical data can be used to statistically analyze the frequency of transitions at different times of the day and study the intraday trend of transition frequency changes. A flag indicating a wind speed transition is set as... , It can be calculated using the following formula: (1-1) (1-2) In the formula: and The first in historical data Wind speed The binary flags indicate sudden changes in wind speed and strong winds during certain periods. The first in historical data sky Wind speed during the period; and This is the threshold for determining sudden changes in wind speed.

[0046] Furthermore, the frequency of wind speed changes during each time period of the day was obtained as follows: (1-3) (1-4) In the formula: S is the number of historical data. and for There is a frequency of wind speed changes during certain periods. Based on and It can further reveal the changing trends of wind speed at different times of the day.

[0047] Furthermore, in step 2, regarding the rainfall intensity parameter, the turning point weather includes excessive rainfall and insufficient rainfall.

[0048] Because rainfall intensity changes occur on a longer timescale and exhibit less volatility compared to temperature and wind speed, relying on changes in rainfall intensity between adjacent periods to characterize rainfall transitions is inaccurate. When analyzing the diurnal timescale characteristics of rainfall transitions, the focus is on identifying periods of excessively high and low rainfall intensity. These typical periods have greater reference value in the scheduling and planning of hydropower stations, and can more effectively guide decision-making and resource allocation. A transition marker for excessively high and low rainfall intensity is defined as follows: , It can be calculated using the following formula: (1-5) (1-6) In the formula: and The first in historical data sky Binary flag indicating whether rainfall intensity is high or low during a given period; The first in historical data sky Rainfall intensity over a given period; and These are the thresholds for determining whether the rainfall intensity is too high or too low.

[0049] Furthermore, the frequency of rainfall transitions during each time period within the day was obtained as follows: (1-7) (1-8) In the formula: and for The frequency of two types of transitional weather (rainfall and change) exists within a given time period, which can further reveal the changing trends of rainfall and change in weather at different times of the day.

[0050] Regarding temperature parameters, the transitional weather includes sudden temperature changes and high temperatures.

[0051] Temperature transitions are characterized by rapid temperature fluctuations and sudden high temperatures, both of which significantly impact photovoltaic (PV) power generation performance. Rapid temperature fluctuations cause PV module conversion efficiency to change swiftly, exacerbating the volatility and uncertainty of PV array output. Conversely, high temperatures significantly reduce PV module efficiency, resulting in lower-than-expected output and further affecting the stability and reliability of PV power generation. This study analyzes the diurnal timescale characteristics of temperature transition weather, using historical data to obtain the frequency of transitions at each time period and the trend of these frequencies throughout the day. A transition marker is set as... , It can be calculated using the following formula: (1-9) (1-10) In the formula: , The first sky The time period contains binary flags indicating sudden temperature changes and high temperatures; For the first sky Temperature during the period; and These are the thresholds for determining sudden temperature changes and transitions to high-temperature weather, respectively.

[0052] Furthermore, the frequency of temperature abrupt changes and high temperatures occurring in each time period of the day was obtained as follows: (1-11) (1-12) In the formula: , They are respectively The frequency of sudden temperature changes and high-temperature transitions within a given time period was obtained using historical data. Based on , It can further reveal the changing trends of temperature shifts at various times of the day.

[0053] Regarding the light intensity parameter, the transitional weather includes light fluctuations and strong light.

[0054] The transition characteristic of solar illumination is mainly manifested in the drastic fluctuations in solar intensity caused by rapid changes in cloud cover. Solar intensity directly affects the output power of photovoltaic arrays, thus impacting their power generation efficiency. This change not only exacerbates the volatility of photovoltaic output but also increases its uncertainty, posing a challenge to the stable operation of the power system. Furthermore, strong sunlight can indirectly lead to temperature changes in photovoltaic panels, further affecting their power generation. Therefore, strong sunlight is also a factor that should be considered during solar illumination transitions. Let the binary variable representing the transition of rapidly changing solar illumination be defined as... The binary variable representing the transition of strong light is It can be calculated using the following formula: (1-13) (1-14) In the formula: and The first Sunlight intensity The binary flags indicate periods of rapid fluctuation and strong light. For the first sky Light intensity during a given time period; and The threshold for determining rapid fluctuations in light intensity and strong light.

[0055] Furthermore, the frequency of the two light transitions occurring in each time period during the day was obtained as follows: (1-15) (1-16) In the formula: and They are respectively The frequency of light intensity fluctuations and strong light transitions during certain time periods can be further analyzed to obtain the changing trends of light transition frequencies at different times of the day.

[0056] Furthermore, in step 4, a monthly timescale characteristic analysis of transitional meteorological events is performed. The following analyses all use the monthly timescale characteristics of temperature abrupt changes / high temperatures as examples. In the monthly timescale characteristic analysis, existing data can be combined to calculate and analyze the monthly / seasonal timescale characteristics. Definitions and For the first Month The frequency of sudden temperature changes and high-temperature transitions within a given time period. The calculation method can be expressed as: (1-17) (1-18) In the formula: For this is the first Number of days in a month (e.g.) It is 31. (For 28 or 29). (Through) and It can obtain the frequency of corresponding weather changes occurring in each time period of each month.

[0057] In addition, the cumulative number of days with monthly transitional weather patterns can be obtained in the monthly time-scale feature analysis, defined as follows: and For the first The number of days with sudden temperature changes / high temperatures during the month is calculated as follows: (1-19) (1-20) (1-21) (1-22) In the formula: and For the first There are markers for temperature fluctuations and high temperatures in the sky.

[0058] The above analysis yields monthly timescale characteristics such as the frequency of transitional weather events in each time period of each month and the number of days with such events in each month. The analysis method for transitional weather events, such as wind speed, is similar to the above process and will not be elaborated upon here.

[0059] Furthermore, in step 5, a timescale characteristic analysis of the transitional meteorological year is performed. This analysis can be obtained by splicing and combining monthly timescale characteristics, again using the annual timescale characteristics of temperature abrupt changes and high temperatures as an example. This is achieved by... , By splicing the features, we can obtain the annual timescale characteristics, which can be defined as follows: and It can be expressed in the following matrix form.

[0060] (1-23) Through matrix , Clearly identify the periods when the weather changes each month. Similarly, by... , By splicing the data, we can obtain annual time-scale features regarding the frequency of occurrence, if defined as... and It can be expressed in the following matrix form: (1-24) pass and It can reveal the changes in the frequency of transitional weather events throughout the year, providing a reference for resource allocation and planning on medium and long time scales.

[0061] Example 4 is an embodiment of the present invention. This embodiment provides a multi-timescale, multi-dimensional feature analysis system for turning point meteorological data based on historical data, including: a data acquisition module for acquiring historical meteorological data, wherein the historical meteorological data includes the values ​​of various meteorological parameters at multiple time periods on multiple historical dates; The flag setting module is used to set flags for each historical date and each time period for each meteorological parameter according to preset judgment conditions, so as to indicate whether a turning point in the weather occurs during that time period; The daily feature calculation module is used to calculate the daily time scale features based on the flag bit. The daily time scale features are obtained by statistically analyzing the feature quantification indicators of the weather turning points in each time period of the day. The monthly feature calculation module is used to calculate monthly time scale features based on the flag bits. The monthly time scale features include time period statistical features and day statistical features. The time period statistical features reflect the weather changes that occur in each time period within the month, and the day statistical features reflect the cumulative weather changes that occur within the month. The annual feature calculation module is used to calculate annual time scale features based on the monthly time scale features of each month. The output module is used to output the daily timescale features, monthly timescale features, and yearly timescale features.

[0062] This embodiment also provides an electronic device applicable to a method for analyzing the multi-timescale, multi-dimensional features of turning point meteorology based on historical data, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the method for analyzing the multi-timescale, multi-dimensional features of turning point meteorology based on historical data as proposed in the above embodiment.

[0063] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements a method for analyzing the multi-timescale and multi-dimensional characteristics of turning point meteorological data based on historical data, as proposed in the above embodiment.

[0064] The storage medium proposed in this embodiment and the method for implementing a multi-timescale, multi-dimensional feature analysis of turning point meteorology based on historical data proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0065] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0066] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing the multi-timescale and multi-dimensional characteristics of turning point meteorological events based on historical data, characterized by: include, Acquire historical meteorological data, which includes values ​​of various meteorological parameters at multiple time periods on multiple historical dates; For each meteorological parameter, a flag is set for each time period of each historical date according to preset judgment conditions to indicate whether a turning point in the weather occurs during that time period; The daily timescale characteristics are calculated based on the aforementioned flag, and these daily timescale characteristics are obtained by statistically analyzing the characteristic quantitative indicators of weather changes that occur in each time period of the day. The monthly timescale features are calculated based on the flag bits. The monthly timescale features include time period statistical features and day statistical features. The time period statistical features reflect the weather changes that occur in each time period within the month, and the day statistical features reflect the cumulative weather changes that occur within the month. Calculate the annual time scale features based on the monthly time scale features of each month; Output the daily timescale features, monthly timescale features, and yearly timescale features.

2. The method for analyzing the multi-timescale, multi-dimensional features of turning points in meteorology based on historical data as described in claim 1, characterized in that: The step of setting a flag for each time period of each historical date according to preset judgment conditions includes determining the turning point judgment type corresponding to the meteorological parameter, wherein the turning point judgment type includes a judgment type based on time sequence change and a judgment type based on numerical range. Set corresponding judgment conditions according to the aforementioned turning point judgment type; The meteorological parameter values ​​for each time period of the historical date are compared with the determination criteria, and a binary flag is set for the time period.

3. The method for analyzing the multi-timescale, multi-dimensional features of turning points in meteorological data based on historical data as described in claim 2, characterized in that: The step of calculating the daily timescale features based on the flag bit includes traversing all historical dates in the historical meteorological data and extracting the flag bit corresponding to each time period of each date. For each time period within a day, count the number of times the flag bit of that time period satisfies the transition indicator state across all historical dates; The frequency of turning points in the time period is obtained by calculating the ratio of the quantity to the total number of historical dates.

4. The method for analyzing the multi-timescale and multi-dimensional characteristics of turning point meteorological events based on historical data as described in claim 3, characterized in that: The statistical characteristics of the time period include, For each time period within a month, count the number of times the flag bit of the time period satisfies the transition indicator state among all dates in the month; The frequency of transitions within the month is obtained by calculating the ratio of the stated quantity to the total number of days in the month. The day count characteristics are obtained through the following steps: For each month, determine whether there is at least one time period flag that satisfies the transition indicator state for each date in the month, and count the number of dates that meet the condition.

5. The method for analyzing the multi-timescale, multi-dimensional characteristics of turning point meteorological events based on historical data as described in claim 4, characterized in that: When the turning point determination type is a determination type based on time sequence change, the determination condition is: the absolute value of the difference between the meteorological parameter value of the current time period and the meteorological parameter value of the adjacent time period exceeds the first preset threshold. When the turning point determination type is a determination type based on numerical range, the determination condition is: the meteorological parameter value of the current time period exceeds the second preset threshold.

6. The method for analyzing the multi-timescale, multi-dimensional features of turning point meteorological events based on historical data as described in claim 5, characterized in that: The calculation of annual timescale features based on monthly timescale features includes, Extract the time period statistical features of each month in order of month, and concatenate the time period statistical features of each month to form an annual time period statistical matrix. The first dimension of the matrix corresponds to the month, and the second dimension corresponds to the intraday time period. Extract the statistical features of the number of days in each month in order of month, and concatenate the statistical features of the number of days in each month to form an annual statistical vector of the number of days.

7. The method for analyzing the multi-timescale and multi-dimensional characteristics of turning point meteorological events based on historical data as described in claim 6, characterized in that: The various meteorological parameters include wind speed, rainfall intensity, temperature, and light intensity; Regarding wind speed, the aforementioned weather changes include sudden changes in wind speed and strong winds; Regarding rainfall intensity, the transitional weather includes excessive rainfall and insufficient rainfall; Regarding temperature, the transitional weather includes sudden temperature changes and high temperatures; Regarding light intensity, the changing weather conditions include light fluctuations and strong light.

8. A system for analyzing the multi-timescale, multi-dimensional features of meteorological turning points based on historical data, employing the multi-timescale, multi-dimensional feature analysis method for meteorological turning points based on historical data as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire historical meteorological data, which includes the values ​​of various meteorological parameters at multiple time periods on multiple historical dates; The flag setting module is used to set flags for each historical date and each time period for each meteorological parameter according to preset judgment conditions, so as to indicate whether a turning point in the weather occurs during that time period; The daily feature calculation module is used to calculate the daily time scale features based on the flag bit. The daily time scale features are obtained by statistically analyzing the feature quantification indicators of the weather turning points in each time period of the day. The monthly feature calculation module is used to calculate monthly time scale features based on the flag bits. The monthly time scale features include time period statistical features and day statistical features. The time period statistical features reflect the weather changes that occur in each time period within the month, and the day statistical features reflect the cumulative weather changes that occur within the month. The annual feature calculation module is used to calculate annual time scale features based on the monthly time scale features of each month. The output module is used to output the daily timescale features, monthly timescale features, and yearly timescale features.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method for analyzing the multi-timescale and multi-dimensional features of turning point meteorology based on historical data, as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for analyzing the multi-timescale and multi-dimensional features of turning point meteorology based on historical data, as described in any one of claims 1 to 7.

Images