Virtual power plant resource optimization management method and system based on big data

Abstract

The invention discloses a virtual power plant resource optimization management method and system based on big data, relates to the field of clean energy, and solves the problem of poor management efficiency of the existing virtual power plant resource optimization management technology, and the method comprises the steps: S1, carrying out the photovoltaic monitoring of a target resource region, creating a target resource optimization period according to a monitoring result, s2, carrying out photovoltaic monitoring on a target resource region, carrying out historical wind power period matching on the target resource optimization period according to a monitoring result, and obtaining photovoltaic period matching data according to a matching result, and carrying out photovoltaic monitoring on the target resource region, and carrying out historical wind power period matching on the target resource optimization period according to the monitoring result. And S3, according to the wind energy period matching data and the photovoltaic period matching data, performing electric energy storage optimization on the target virtual power plant in the target resource optimization period. According to the method, the operation stability and the new energy consumption capability of the virtual power plant can be effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of clean energy and relates to resource management technology, specifically a virtual power plant resource optimization management method and system based on big data. Background Technology

[0002] Existing power resource optimization management technologies have the following shortcomings when performing resource optimization management on virtual power plants: (1) Existing power resource optimization management technology still has certain limitations in the predictive scheduling of new energy power generation and energy storage. Wind power generation is affected by the intermittent changes in wind speed and direction and complex climate conditions, while solar power generation is highly dependent on meteorological factors such as sunshine duration, intensity and dynamic cloud distribution. Although existing technologies can build prediction models, their accuracy is limited when capturing these random fluctuation characteristics, resulting in deviations between the predicted power generation results and the actual values. It is difficult to accurately predict the spatiotemporal distribution patterns of wind and solar power output. At the same time, the effective scheduling of energy storage systems needs to be based on the dynamic matching of power generation prediction and load demand. However, the current technology is not perfect in its ability to quantify the uncertainty of wind and solar power output. The prediction of key periods of energy storage demand has problems such as coarse time granularity and delayed response. The real-time assessment of energy storage capacity also lacks a dynamic adjustment mechanism, resulting in low utilization efficiency of energy storage resources. (2) In the prediction of photovoltaic and wind power generation, when the optimization period is short, if daily, monthly or even seasonal data are used as samples, the prediction will be delayed because the rapid changes in intraday weather are ignored. If the optimization period is long, but only short-term high-frequency data are relied on, it is difficult to capture seasonal climate patterns and cause trend deviation. The mismatch between the sample period and the optimization target makes it impossible for the model to accurately extract key features, which ultimately reduces the adaptability of the prediction results to the actual power generation fluctuations and results in a lack of accuracy in the power generation prediction results.

[0003] To address this, a virtual power plant resource optimization management method and system based on big data is proposed. Summary of the Invention

[0004] In view of the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for optimizing virtual power plant resources based on big data, the main purpose of which is to improve the utilization efficiency of virtual power plant energy storage resources and the utilization efficiency of clean energy.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a virtual power plant resource optimization management method based on big data, the specific steps of which are as follows: Step S1: Conduct photovoltaic monitoring on the target resource area, create a target resource optimization cycle based on the monitoring results, match the target resource optimization cycle with historical photovoltaic cycles based on historical photovoltaic monitoring results, and obtain photovoltaic cycle matching data based on the matching results; Step S2: Conduct photovoltaic monitoring on the target resource area, match the historical wind cycle with the target resource optimization cycle based on the monitoring results, and obtain wind energy cycle matching data based on the matching results; Step S3: Optimize the power storage of the target virtual power plant in the target resource optimization cycle based on the wind energy cycle matching data and the photovoltaic cycle matching data.

[0006] Furthermore, in step S1, the specific steps are as follows: Step S11: Obtain the virtual power plant that needs to be optimized and managed by resources, obtain the target virtual power plant, obtain the resource adjustment duration of the target virtual power plant, obtain the preset resource adjustment duration, take the time point corresponding to the current moment as the start time point of the cycle, take the preset resource adjustment duration as the cycle duration, create a resource optimization cycle, and obtain the target resource optimization cycle. Step S12: Perform photovoltaic resource analysis on the target virtual power plant in the target resource optimization cycle, and obtain the photovoltaic initial screening matching cycle corresponding to the target resource optimization cycle based on the analysis results; Step S13: Randomly select a sample photovoltaic screening matching period from the multiple historical photovoltaic screening matching periods obtained, match the sample photovoltaic screening matching period with the target resource optimization period to the periodic change of solar radiation, and obtain the periodic radiation consistency based on the matching result. Step S14: Obtain the periodic radiation consistency degree corresponding to each initial photovoltaic matching cycle, set the periodic radiation consistency degree benchmark interval. If the periodic radiation consistency degree is within the periodic radiation consistency degree benchmark interval, the corresponding initial photovoltaic matching cycle is divided into an effective photovoltaic matching cycle. If the periodic radiation consistency degree is not within the periodic radiation consistency degree benchmark interval, the corresponding initial photovoltaic matching cycle is divided into an ineffective photovoltaic matching cycle, thus obtaining photovoltaic periodic matching data.

[0007] Furthermore, in step S12, the specific steps are as follows: The photovoltaic power generation area corresponding to the target virtual power plant is marked as the target photovoltaic area. The calendar date of the target resource optimization cycle is obtained, and the collected calendar date is converted into an annual day to obtain the target annual day. The noon solar altitude angle corresponding to the target resource optimization cycle is calculated based on the target annual day. The solar altitude angle at noon corresponding to the target resource optimization cycle and the target year accumulated days are used to calculate the solar radiation outside the atmosphere at noon corresponding to the target resource optimization cycle, and the target solar radiation outside the atmosphere at noon is obtained. For the target photovoltaic area, several historical sunshine monitoring cycles are selected, and the noon solar radiation outside the atmosphere corresponding to each historical sunshine monitoring cycle is obtained. The difference between the noon solar radiation outside the atmosphere and the target noon solar radiation outside the atmosphere is calculated, and the absolute value of the obtained difference is calculated to obtain multiple noon solar radiation deviations. A preset reasonable value for the solar radiation deviation is set. If the noon solar radiation deviation is less than or equal to the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic initial screening matching cycle. If the noon solar radiation deviation is greater than the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic matching anomaly cycle.

[0008] Furthermore, in step S13, the specific steps are as follows: In the existing Cartesian coordinate system, a periodic radiation coordinate system is created by using the periodic time value as the x-axis and the actual solar radiation as the y-axis. The solar radiation of the target photovoltaic area during the target resource optimization cycle is predicted by weather forecasting equipment. Based on the prediction results, the target radiation variation curve is created in the cycle radiation coordinate system. The actual solar radiation of the target photovoltaic area during the sample photovoltaic initial screening matching cycle is collected, and the sample radiation variation curve is plotted based on the collection results. In the periodic radiation coordinate system, arbitrarily select a characteristic curve point in the target radiation variation curve, and create a vertical comparison line with the characteristic curve point as the center point. If the vertical comparison line can cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an effective coverage point. If the vertical comparison line cannot cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an invalid coverage point. The longitudinal comparison line is used to traverse each curve point in the target radiation change curve. Based on the traversal results, the curve points are divided into effective coverage points and ineffective coverage points. The quantity value corresponding to the effective coverage points is counted as A1, and the quantity value corresponding to the ineffective coverage points is counted as A2. A1 / (A1+A2) is calculated to obtain the periodic radiation consistency degree corresponding to the sample photovoltaic initial screening matching cycle.

[0009] Furthermore, in step S2, the specific steps are as follows: Step S21: Obtain the target resource optimization cycle and mark the wind power generation area corresponding to the target virtual power plant as the target wind power area; Step S22: Predict the satellite cloud image of the target wind area within the target resource optimization period, and mark the predicted satellite cloud images as T1 target satellite cloud image, T2 target satellite cloud image, ... Ta target satellite cloud image in chronological order to obtain the target satellite cloud image sequence; Step S23: Collect multiple historical wind cycles for the target wind area, match the historical wind cycles with the target resource optimization cycle using satellite cloud images, and screen the wind energy initial matching cycle based on the matching results; Step S24: Match the wind energy initial screening matching period with the target resource optimization period to the wind energy resource quantity period change. Obtain the periodic wind energy consistency based on the matching results. Divide the wind energy initial screening matching period into effective wind energy matching period and ineffective wind energy matching period based on the periodic wind energy consistency to obtain wind energy period matching data.

[0010] Furthermore, in step S23, the specific steps are as follows: Step S231: Randomly select one sample historical wind cycle from the multiple historical wind cycles obtained; Step S232: Collect historical satellite cloud images corresponding to the historical wind force cycles of the samples, and label them in chronological order of collection time as T1 historical satellite cloud image, T2 historical satellite cloud image, ... Ta historical satellite cloud image; Step S233: Perform meteorological complex identification on the target satellite cloud image. If a meteorological complex exists in the target satellite cloud image, compare the non-trajectory parameters of the T1 historical satellite cloud image and the T1 target satellite cloud image. If the comparison results are consistent, continue to compare the T2 historical satellite cloud image and the T2 target satellite cloud image in the same way. If the comparison results are inconsistent, directly determine that the historical wind force period corresponding to the sample is an invalid wind force matching period. Continue in this manner until the consistency comparison between the Ta historical satellite cloud image and the Ta target satellite cloud image is completed. Step S234: If the non-trajectory parameters of the historical satellite cloud image and the target satellite cloud image are completely consistent, then perform a meteorological complex trajectory consistency analysis on the historical satellite cloud image and the target satellite cloud image. Based on the analysis results, collect an invalid cloud image control group for the target satellite cloud image. If there is no invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind energy initial screening matching cycle. If there is an invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind force invalid matching cycle.

[0011] Furthermore, in step S234, the specific steps are as follows: One meteorological complex is randomly selected from the meteorological complexes appearing in the target satellite cloud image as a sample meteorological complex. Two target satellite cloud images with adjacent acquisition times are obtained for the sample meteorological complex and combined to set up the target cloud image control group. In this way, multiple target cloud image control groups are obtained. Two historical satellite cloud images corresponding to the acquisition time of the target cloud image control group are obtained to obtain multiple historical cloud image control groups. The target satellite cloud images in the target cloud image control group were named the first target control cloud image and the second target control cloud image according to the order of acquisition time. The historical satellite cloud images in the historical cloud image control group were named the first historical control cloud image and the second historical control cloud image according to the order of acquisition time. In the first target comparison cloud map, the cloud map area covered by the sample meteorological complex is divided into several meteorological pixels. Then, one sample meteorological pixel is randomly selected from these pixels, and the geometric center of the cloud map area corresponding to the target wind region is acquired to obtain the center point of the wind region. A Cartesian coordinate system is created by taking the center point of the wind area as the origin and the due north direction as the positive y-axis, thus obtaining the first target coordinate system. Repeat the process of creating the first target coordinate system, and create Cartesian coordinate systems for the second target reference cloud map, the first historical reference cloud map, and the second historical reference cloud map respectively, to obtain the second target coordinate system, the first historical coordinate system, and the second historical coordinate system.

[0012] Furthermore, in step S234, the specific steps are as follows: The coordinate positions of the sample meteorological pixels in the first target coordinate system, the second target coordinate system, the first historical coordinate system, and the second historical coordinate system are collected to obtain the first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates. The first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates are used to create coordinate points in the first target comparison cloud map to obtain the first pixel target point, the second pixel target point, the first pixel historical point, and the second pixel historical point. The closed area determined by the first pixel target point, the second pixel target point, and the center point of the wind area is marked as the sample pixel target area. The closed area determined by the first pixel historical point, the second pixel historical point, and the center point of the wind area is marked as the sample pixel historical area. The overlapping area between the sample pixel target area and the sample pixel historical area is marked as the target historical overlapping area. The ratio of the area value of the target historical overlapping area to the sample pixel target area is calculated to obtain the path overlap area ratio corresponding to the sample meteorological pixel. Repeat the process of obtaining the path overlap area ratio corresponding to the sample meteorological pixels. Obtain the path overlap area ratio for each meteorological pixel covered by the sample meteorological complex, and set a qualified value for the path overlap area ratio. If the path overlap area ratio is greater than or equal to the qualified value, the meteorological pixel is classified as a path consistent pixel. If the path overlap area ratio is less than the qualified value, the meteorological pixel is classified as a path inconsistent pixel. The path consistency of the historical cloud map control group is obtained by calculating the ratio of the number of path-consistent pixels to the number of meteorological pixels. A path consistency benchmark interval is set. If the path consistency is within the path consistency benchmark interval, the historical cloud map control group is classified as a valid cloud map control group. If the path consistency is not within the path consistency benchmark interval, the historical cloud map control group is classified as an invalid cloud map control group. For each meteorological complex in the target satellite cloud image, the validity of the cloud image control group is judged, and the invalid cloud image control group is obtained based on the judgment results.

[0013] Furthermore, in step S3, the specific steps are as follows: Acquire photovoltaic cycle matching data, and obtain multiple effective photovoltaic matching cycles based on the photovoltaic cycle matching data; acquire wind energy cycle matching data, and obtain multiple effective wind energy matching cycles based on the wind energy cycle matching data. Based on the effective matching period of photovoltaic power, the photovoltaic power generation of the target virtual power plant is predicted, and multiple predicted photovoltaic power generation are obtained based on the prediction results. Create an energy storage analysis coordinate system. In the energy storage analysis coordinate system, connect the coordinate points with the characteristic time point on the horizontal axis and the predicted power generation on the vertical axis in sequence to obtain the power generation cycle prediction curve. The power generation demand of the target virtual power plant at different characteristic time points is collected and created as a power generation periodic demand curve in the energy storage analysis coordinate system; In the energy storage analysis coordinate system, when the power generation cycle prediction curve is higher than the power generation cycle demand curve, the closed area enclosed by the power generation cycle prediction curve and the power generation cycle demand curve is set as the energy storage dispatch area. The x-axis region covered by the energy storage dispatch area is set as the energy storage demand period, and the area value of the region covered by the energy storage dispatch area is set as the energy storage space demand corresponding to the energy storage demand period. The energy storage space demand corresponding to the energy storage demand period is fed back to the energy storage control terminal, and the energy storage control terminal performs energy storage allocation for the target virtual power plant.

[0014] A virtual power plant resource optimization management system based on big data includes: Photovoltaic matching module: performs photovoltaic monitoring on the target resource area, creates a target resource optimization cycle based on the monitoring results, performs historical photovoltaic cycle matching on the target resource optimization cycle based on historical photovoltaic monitoring results, and obtains photovoltaic cycle matching data based on the matching results; Wind matching module: Monitors photovoltaic power in the target resource area, matches the target resource optimization cycle with historical wind cycles based on the monitoring results, and obtains wind energy cycle matching data based on the matching results; Resource optimization module: Optimizes the energy storage of target virtual power plants in the target resource optimization cycle based on wind energy cycle matching data and photovoltaic cycle matching data.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: (1) This invention obtains the target noon extra-layer radiation by collecting the annual accumulated days and noon solar altitude angle corresponding to the target resource optimization cycle. Based on the target noon extra-layer radiation, the photovoltaic initial screening matching cycle is matched with the target resource optimization cycle. The photovoltaic initial screening matching cycle is further screened by performing time-period solar radiation consistency analysis, thereby obtaining the effective photovoltaic matching cycle. This can fully guarantee the adaptability of the photovoltaic power generation prediction sample cycle, thus providing a guarantee for the accuracy of photovoltaic power prediction. (2) This invention compares non-trajectory meteorological parameters and analyzes the consistency of meteorological complex trajectory by collecting satellite cloud images corresponding to the target resource optimization cycle and historical cycle satellite cloud images. Based on the analysis results, it matches the wind power initial screening matching cycle for the target resource optimization cycle, and performs time-based wind energy resource consistency analysis on the wind power initial screening matching cycle to further screen the wind power initial screening matching cycle, thereby obtaining the effective wind power matching cycle. It can accurately match the effective wind power cycle that is highly consistent with the meteorological characteristics of the target cycle, reduce prediction deviation, and significantly improve the accuracy and reliability of wind power generation prediction. (3) The present invention creates a power generation cycle demand curve for the target virtual power plant in the target resource optimization cycle based on the effective matching cycle of wind power and the effective matching cycle of photovoltaic power, and optimizes the power storage of the target virtual power plant based on the power generation cycle demand curve. This enables the on-demand dynamic allocation of power storage, effectively improving the operation stability of the virtual power plant and the new energy absorption capacity. Attached Figure Description

[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0017] Figure 1 This is a diagram illustrating the implementation steps of the present invention; Figure 2 This is an overall system block diagram of the present invention; Figure 3 This is the periodic radiation coordinate system in this invention; Figure 4 This is the coordinate system for energy storage analysis in this invention. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1 Please see Figure 1 This invention provides a technical solution: a virtual power plant resource optimization management method based on big data, the specific steps of which are as follows: Step S1: Conduct photovoltaic monitoring on the target resource area, create a target resource optimization cycle based on the monitoring results, match the target resource optimization cycle with historical photovoltaic cycles based on historical photovoltaic monitoring results, and obtain photovoltaic cycle matching data based on the matching results; The specific steps in step S1 are as follows: Step S11: Obtain the virtual power plant that needs to be optimized and managed by resources, obtain the target virtual power plant, obtain the resource adjustment duration of the target virtual power plant, obtain the preset resource adjustment duration, take the time point corresponding to the current moment as the start time point of the cycle, take the preset resource adjustment duration as the cycle duration, create a resource optimization cycle, and obtain the target resource optimization cycle. Step S12: Perform photovoltaic resource analysis on the target virtual power plant in the target resource optimization cycle, and obtain the photovoltaic initial screening matching cycle corresponding to the target resource optimization cycle based on the analysis results; In step S12, the specific steps are as follows: The photovoltaic power generation area corresponding to the target virtual power plant is marked as the target photovoltaic area. The calendar date of the target resource optimization cycle is obtained, and the collected calendar date is converted into an annual day to obtain the target annual day. The noon solar altitude angle corresponding to the target resource optimization cycle is calculated based on the target annual day. The solar altitude angle at noon corresponding to the target resource optimization cycle and the target year accumulated days are used to calculate the solar radiation outside the atmosphere at noon corresponding to the target resource optimization cycle, and the target solar radiation outside the atmosphere at noon is obtained. For the target photovoltaic area, several historical sunshine monitoring cycles are selected, and the noon solar radiation outside the atmosphere corresponding to each historical sunshine monitoring cycle is obtained. The difference between the noon solar radiation outside the atmosphere and the target noon solar radiation is calculated, and the absolute value of the obtained difference is calculated to obtain multiple noon solar radiation deviations. A preset reasonable value for the solar radiation deviation is set. If the noon solar radiation deviation is less than or equal to the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic initial screening matching cycle. If the noon solar radiation deviation is greater than the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic matching abnormal cycle. Step S13: Randomly select a sample photovoltaic screening matching period from the multiple historical photovoltaic screening matching periods obtained, match the sample photovoltaic screening matching period with the target resource optimization period to the periodic change of solar radiation, and obtain the periodic radiation consistency based on the matching result. In step S13, the specific steps are as follows: In the existing Cartesian coordinate system, a periodic radiation coordinate system is created by using the periodic time value as the x-axis and the actual solar radiation as the y-axis. The solar radiation of the target photovoltaic area during the target resource optimization cycle is predicted by weather forecasting equipment. Based on the prediction results, the target radiation variation curve is created in the cycle radiation coordinate system. The actual solar radiation of the target photovoltaic area during the sample photovoltaic initial screening matching cycle is collected, and the sample radiation variation curve is plotted based on the collection results. In the periodic radiation coordinate system, arbitrarily select a characteristic curve point in the target radiation variation curve, and create a vertical comparison line with the characteristic curve point as the center point. If the vertical comparison line can cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an effective coverage point. If the vertical comparison line cannot cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an invalid coverage point. The longitudinal comparison line is used to traverse each curve point in the target radiation change curve. Based on the traversal results, the curve points are divided into effective coverage points and ineffective coverage points. The quantity value corresponding to the effective coverage points is counted as A1, and the quantity value corresponding to the ineffective coverage points is counted as A2. A1 / (A1+A2) is calculated to obtain the periodic radiation consistency degree corresponding to the sample photovoltaic initial screening matching cycle. Step S14: Obtain the periodic radiation consistency degree corresponding to each initial photovoltaic matching cycle, set the periodic radiation consistency degree benchmark interval. If the periodic radiation consistency degree is within the periodic radiation consistency degree benchmark interval, the corresponding initial photovoltaic matching cycle is divided into an effective photovoltaic matching cycle. If the periodic radiation consistency degree is not within the periodic radiation consistency degree benchmark interval, the corresponding initial photovoltaic matching cycle is divided into an ineffective photovoltaic matching cycle, and photovoltaic periodic matching data is obtained. Step S2: Conduct photovoltaic monitoring on the target resource area, match the historical wind cycle with the target resource optimization cycle based on the monitoring results, and obtain wind energy cycle matching data based on the matching results; In step S2, the specific steps are as follows: Step S21: Obtain the target resource optimization cycle and mark the wind power generation area corresponding to the target virtual power plant as the target wind power area; Step S22: Predict the satellite cloud image of the target wind area within the target resource optimization period, and mark the predicted satellite cloud images as T1 target satellite cloud image, T2 target satellite cloud image, ... Ta target satellite cloud image in chronological order to obtain the target satellite cloud image sequence; Step S23: Collect multiple historical wind cycles for the target wind area, match the historical wind cycles with the target resource optimization cycle using satellite cloud images, and screen the wind energy initial matching cycle based on the matching results; In step S23, the specific steps are as follows: Step S231: Randomly select one sample historical wind cycle from the multiple historical wind cycles obtained; Step S232: Collect historical satellite cloud images corresponding to the historical wind force cycles of the samples, and label them in chronological order of collection time as T1 historical satellite cloud image, T2 historical satellite cloud image, ... Ta historical satellite cloud image; Step S233: Perform meteorological complex identification on the target satellite cloud image. If a meteorological complex exists in the target satellite cloud image, compare the non-trajectory parameters of the T1 historical satellite cloud image and the T1 target satellite cloud image. If the comparison results are consistent, continue to compare the T2 historical satellite cloud image and the T2 target satellite cloud image in the same way. If the comparison results are inconsistent, directly determine that the historical wind force period corresponding to the sample is an invalid wind force matching period. Continue in this manner until the consistency comparison between the Ta historical satellite cloud image and the Ta target satellite cloud image is completed. Step S234: If the non-trajectory parameters of the historical satellite cloud image and the target satellite cloud image are completely consistent, then perform a meteorological complex trajectory consistency analysis on the historical satellite cloud image and the target satellite cloud image. Based on the analysis results, collect an invalid cloud image control group for the target satellite cloud image. If there is no invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind energy initial screening matching cycle. If there is an invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind force invalid matching cycle. In step S234, the specific steps are as follows: One meteorological complex is randomly selected from the meteorological complexes appearing in the target satellite cloud image as a sample meteorological complex. Two target satellite cloud images with adjacent acquisition times are obtained for the sample meteorological complex and combined to set up the target cloud image control group. In this way, multiple target cloud image control groups are obtained. Two historical satellite cloud images corresponding to the acquisition time of the target cloud image control group are obtained to obtain multiple historical cloud image control groups. The target satellite cloud images in the target cloud image control group were named the first target control cloud image and the second target control cloud image according to the order of acquisition time. The historical satellite cloud images in the historical cloud image control group were named the first historical control cloud image and the second historical control cloud image according to the order of acquisition time. In the first target comparison cloud map, the cloud map area covered by the sample meteorological complex is divided into several meteorological pixels. Then, one sample meteorological pixel is randomly selected from these pixels, and the geometric center of the cloud map area corresponding to the target wind region is acquired to obtain the center point of the wind region. A Cartesian coordinate system is created by taking the center point of the wind area as the origin and the due north direction as the positive y-axis, thus obtaining the first target coordinate system. Repeat the process of creating the first target coordinate system, and create Cartesian coordinate systems for the second target reference cloud map, the first historical reference cloud map, and the second historical reference cloud map respectively, to obtain the second target coordinate system, the first historical coordinate system, and the second historical coordinate system; The coordinate positions of the sample meteorological pixels in the first target coordinate system, the second target coordinate system, the first historical coordinate system, and the second historical coordinate system are collected to obtain the first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates. The first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates are used to create coordinate points in the first target comparison cloud map to obtain the first pixel target point, the second pixel target point, the first pixel historical point, and the second pixel historical point. The closed area determined by the first pixel target point, the second pixel target point, and the center point of the wind area is marked as the sample pixel target area. The closed area determined by the first pixel historical point, the second pixel historical point, and the center point of the wind area is marked as the sample pixel historical area. The overlapping area between the sample pixel target area and the sample pixel historical area is marked as the target historical overlapping area. The ratio of the area value of the target historical overlapping area to the sample pixel target area is calculated to obtain the path overlap area ratio corresponding to the sample meteorological pixel. Repeat the process of obtaining the path overlap area ratio corresponding to the sample meteorological pixels. Obtain the path overlap area ratio for each meteorological pixel covered by the sample meteorological complex, and set a qualified value for the path overlap area ratio. If the path overlap area ratio is greater than or equal to the qualified value, the meteorological pixel is classified as a path consistent pixel. If the path overlap area ratio is less than the qualified value, the meteorological pixel is classified as a path inconsistent pixel. The path consistency of the historical cloud map control group is obtained by calculating the ratio of the number of path-consistent pixels to the number of meteorological pixels. A path consistency benchmark interval is set. If the path consistency is within the path consistency benchmark interval, the historical cloud map control group is classified as a valid cloud map control group. If the path consistency is not within the path consistency benchmark interval, the historical cloud map control group is classified as an invalid cloud map control group. For each meteorological complex in the target satellite cloud image, the validity of the cloud image control group is judged, and the invalid cloud image control group is obtained based on the judgment results; Step S24: Match the wind energy initial screening matching period with the target resource optimization period to the wind energy resource quantity period change. Obtain the periodic wind energy consistency based on the matching results. Divide the wind energy initial screening matching period into effective wind energy matching period and ineffective wind energy matching period based on the periodic wind energy consistency to obtain wind energy period matching data. Step S3: Optimize the energy storage of the target virtual power plant in the target resource optimization cycle based on the wind energy cycle matching data and the photovoltaic cycle matching data; In step S3, the specific steps are as follows: Acquire photovoltaic cycle matching data, and obtain multiple effective photovoltaic matching cycles based on the photovoltaic cycle matching data; acquire wind energy cycle matching data, and obtain multiple effective wind energy matching cycles based on the wind energy cycle matching data. Based on the effective matching period of photovoltaic power, the photovoltaic power generation of the target virtual power plant is predicted, and multiple predicted photovoltaic power generation are obtained based on the prediction results. Create an energy storage analysis coordinate system. In the energy storage analysis coordinate system, connect the coordinate points with the characteristic time point on the horizontal axis and the predicted power generation on the vertical axis in sequence to obtain the power generation cycle prediction curve. The power generation demand of the target virtual power plant at different characteristic time points is collected and created as a power generation periodic demand curve in the energy storage analysis coordinate system; In the energy storage analysis coordinate system, when the power generation cycle prediction curve is higher than the power generation cycle demand curve, the closed area enclosed by the power generation cycle prediction curve and the power generation cycle demand curve is set as the energy storage dispatch area. The x-axis region covered by the energy storage dispatch area is set as the energy storage demand period, and the area value of the region covered by the energy storage dispatch area is set as the energy storage space demand corresponding to the energy storage demand period. The energy storage space demand corresponding to the energy storage demand period is fed back to the energy storage control terminal, and the energy storage control terminal performs energy storage allocation for the target virtual power plant.

[0020] Example 2 Please see Figure 2 Based on another concept of the same invention, a virtual power plant resource optimization management system based on big data is proposed, as follows: The photovoltaic matching module monitors the photovoltaic area of ​​the target resource area, creates a target resource optimization cycle based on the monitoring results, performs historical photovoltaic cycle matching on the target resource optimization cycle in combination with historical photovoltaic monitoring results, and obtains photovoltaic cycle matching data based on the matching results. Specifically as follows: The system acquires virtual power plants that require resource optimization management, obtains target virtual power plants, acquires resource adjustment duration for target virtual power plants, obtains preset resource adjustment duration, uses the current time point as the cycle start time point, uses the preset resource adjustment duration as the cycle duration, creates a resource optimization cycle, and obtains the target resource optimization cycle. It should be noted here that: In this invention, the power generation corresponding to the target virtual power plant is specifically solar power generation and wind power generation, and the preset resource adjustment time involved here is specifically 24 hours.

[0021] Photovoltaic resource analysis is performed on the target virtual power plant in the target resource optimization cycle, and the photovoltaic initial screening matching cycle corresponding to the target resource optimization cycle is obtained based on the analysis results; Specifically as follows: The photovoltaic power generation area corresponding to the target virtual power plant is marked as the target photovoltaic area. The calendar date of the target resource optimization cycle is obtained, and the collected calendar date is converted into a year-day to obtain the target year-day. It should be noted here that: In this invention, if the calendar date of the target resource optimization period is January 1, then the target year-day is 1; if the calendar date of the target resource optimization period is December 31 and the year corresponding to the target resource optimization period is a leap year, then the target year-day is 366; if the year corresponding to the target resource optimization period is not a leap year, then the target year-day is 365.

[0022] The noon solar altitude angle corresponding to the target resource optimization cycle is calculated based on the target year's accumulated days, using the following formula: ; in, Gdj The noon solar altitude angle corresponding to the target resource optimization cycle. Mrj For target year-to-date accumulation; It should be noted here that: In this invention, 23.45° is an approximation of the obliquity of the ecliptic, and 284 is a phase adjustment parameter in the formula, used to correct the difference between the calculation starting point (vernal equinox) and the annual day counting method, so as to ensure that the calculation results are consistent with actual astronomical observations.

[0023] The solar altitude angle at noon corresponding to the target resource optimization cycle and the target year accumulated days are used to calculate the solar radiation outside the atmosphere at noon corresponding to the target resource optimization cycle, and the target solar radiation outside the atmosphere at noon is obtained. The specific formula for calculating the target's noon extra-layer radiation is as follows: ; in, Edq The target noon extra-layer radiation amount. Esc The solar constant, Mrj For the target year accumulated days, Gdj The noon solar altitude angle; It should be noted here that: In this invention, the solar constant is 1367 W / m. 2 .

[0024] For the target photovoltaic area, several historical sunshine monitoring cycles are selected, and the noon solar radiation outside the atmosphere corresponding to each historical sunshine monitoring cycle is obtained. The difference between the noon solar radiation outside the atmosphere and the target noon solar radiation is calculated, and the absolute value of the obtained difference is calculated to obtain multiple noon solar radiation deviations. A preset reasonable value for the solar radiation deviation is set. If the noon solar radiation deviation is less than or equal to the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic initial screening matching cycle. If the noon solar radiation deviation is greater than the preset reasonable value, the historical sunshine monitoring cycle is divided into a photovoltaic matching abnormal cycle. It should be noted here that: In this invention, the matching cycle of the completed initial photovoltaic screening is obtained to obtain the historical photovoltaic initial screening matching cycle. The deviation of the noon extra-layer radiation corresponding to each historical photovoltaic initial screening matching cycle is obtained. The values ​​of the obtained multiple noon extra-layer radiation deviations are compared, and the noon extra-layer radiation deviation with the smallest value is set as the preset reasonable value of the extra-layer radiation deviation.

[0025] It should be noted here that: In this invention, the duration of the initial photovoltaic matching cycle and the duration of the target resource optimization cycle are equal. In this invention, by obtaining the deviation of noon extra-layer radiation corresponding to each historical illumination monitoring cycle, the historical illumination monitoring cycle is initially screened. This ensures that the noon solar altitude angle, sunrise and sunset times, and solar azimuth angle corresponding to the historical photovoltaic initial screening matching cycle are within a reasonable deviation level from the noon solar altitude angle, sunrise and sunset times, and solar azimuth angle corresponding to the target resource optimization cycle, thus avoiding abnormal influences of the noon solar altitude angle, sunrise and sunset times, and solar azimuth angle on the historical photovoltaic cycle matching.

[0026] Randomly select a sample photovoltaic screening matching period from the multiple historical photovoltaic screening matching periods obtained, match the solar radiation period variation with the target resource optimization period, and obtain the periodic radiation consistency based on the matching results. Specifically as follows: In the existing Cartesian coordinate system, a periodic radiation coordinate system is created by using the periodic time value as the x-axis and the actual solar radiation as the y-axis. The solar radiation of the target photovoltaic area during the target resource optimization cycle is predicted by weather forecasting equipment. Based on the prediction results, the target radiation variation curve is created in the cycle radiation coordinate system. The actual solar radiation of the target photovoltaic area during the sample photovoltaic initial screening matching cycle is collected, and the sample radiation variation curve is plotted based on the collection results. Please see Figure 3 In the periodic radiation coordinate system, arbitrarily select a characteristic curve point in the target radiation variation curve, and create a vertical comparison line with the characteristic curve point as the center point. If the vertical comparison line can cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an effective coverage point. If the vertical comparison line cannot cover the sample radiation variation curve in the vertical direction, then the characteristic curve point is marked as an invalid coverage point. It should be noted here that: In this invention, the target radiation change curves divided into effective coverage points are collected to obtain multiple historical effective points. The vertical distance between each historical effective point and the corresponding sample radiation change curve is obtained to obtain multiple effective vertical distances. The multiple effective vertical distances are compared numerically, and the effective vertical distance with the largest value is marked as the preset effective distance. Twice the preset effective distance is set as the length value of the longitudinal comparison line.

[0027] The longitudinal comparison line is used to traverse each curve point in the target radiation change curve. Based on the traversal results, the curve points are divided into effective coverage points and ineffective coverage points. The quantity value corresponding to the effective coverage points is counted as A1, and the quantity value corresponding to the ineffective coverage points is counted as A2. A1 / (A1+A2) is calculated to obtain the periodic radiation consistency degree corresponding to the sample photovoltaic initial screening matching cycle. Repeat the process of obtaining the periodic radiation consistency corresponding to the initial photovoltaic screening matching cycle of the sample. Obtain the periodic radiation consistency corresponding to each initial photovoltaic screening matching cycle. Set the periodic radiation consistency benchmark interval. If the periodic radiation consistency is within the periodic radiation consistency benchmark interval, the corresponding initial photovoltaic screening matching cycle is divided into a valid photovoltaic matching cycle. If the periodic radiation consistency is not within the periodic radiation consistency benchmark interval, the corresponding initial photovoltaic screening matching cycle is divided into an invalid photovoltaic matching cycle. Obtain photovoltaic periodic matching data. It should be noted here that: In this invention, the upper limit of the periodic radiation consistency benchmark interval is 100%, meaning that the curve points corresponding to the initial screening photovoltaic matching cycle are all effective coverage points. The historical initial screening photovoltaic matching cycles, which are divided into effective photovoltaic matching cycles in the historical matching process, are obtained. The periodic radiation consistency corresponding to each historical initial screening photovoltaic matching cycle is obtained, and the values ​​of the multiple obtained periodic radiation consistency are compared. The periodic radiation consistency with the smallest value is set as the lower limit of the periodic radiation consistency benchmark interval.

[0028] The wind matching module monitors the photovoltaic power in the target resource area, matches the historical wind cycle with the target resource optimization cycle based on the monitoring results, and obtains wind energy cycle matching data based on the matching results. Specifically as follows: Obtain the target resource optimization cycle and mark the wind power generation area corresponding to the target virtual power plant as the target wind power area; Meteorological models are used to predict satellite cloud images of the target wind area within the target resource optimization period, and the predicted satellite cloud images are labeled as T1 target satellite cloud image, T2 target satellite cloud image, ... Ta target satellite cloud image in chronological order to obtain the target satellite cloud image sequence; It should be noted here that: In this invention, T1, T2, ..., Ta in the target satellite cloud image T1, T2, ..., Ta are the sequence symbols corresponding to the target satellite cloud images, and a is the quantity value corresponding to the target operational cloud image, and a is an integer greater than 0; Multiple historical wind cycles were collected for the target wind area, and one sample historical wind cycle was randomly selected from the multiple historical wind cycles obtained. The historical wind force cycle of the sample is matched with the target resource optimization cycle using satellite cloud images, and the historical wind force cycle of the sample is classified according to the matching results. Specifically as follows: Historical satellite cloud images corresponding to the historical wind force cycles of the samples were collected and labeled as T1 historical satellite cloud image, T2 historical satellite cloud image, ... Ta historical satellite cloud image according to the order of collection time; It should be noted here that: In this invention, T1, T2, ..., Ta in the historical satellite cloud image T1, T2, ..., Ta are the sequence symbols corresponding to the historical satellite cloud images, and a is the quantity value corresponding to the target running cloud image, and a is an integer greater than 0; In this invention, the acquisition time of the historical satellite cloud image from T1 to Ta and the target satellite cloud image from T1 to Ta within the corresponding wind cycle remains relatively consistent. For example, if the acquisition time of the historical satellite cloud image from T1 in the sample historical wind cycle is 16:00:07, then the acquisition time of the target satellite cloud image from T1 in the target resource optimization cycle is also 16:00:07.

[0029] Meteorological complex identification is performed on the target satellite cloud image. If a meteorological complex exists in the target satellite cloud image, the non-trajectory parameters of the T1 historical satellite cloud image and the T1 target satellite cloud image are compared. If the comparison results are consistent, the same comparison is performed between the T2 historical satellite cloud image and the T2 target satellite cloud image. If the comparison results are inconsistent, the historical wind cycle corresponding to the sample is directly determined to be an invalid wind matching cycle. This process is repeated until the consistency comparison between the Ta historical satellite cloud image and the Ta target satellite cloud image is completed. It should be noted here that: In this invention, the meteorological complex referred to herein specifically refers to weather or climate phenomena in the atmosphere that are formed by the interaction of various physical processes and have specific structures and dynamic characteristics. The meteorological complex referred to herein includes, but is not limited to, extratropical cyclones, fronts, and turbulence. The specific meteorological complex referred to herein is an extratropical cyclone.

[0030] Non-trajectory parameter comparison was performed between historical satellite cloud images of T1 and target satellite cloud images of T1; Specifically as follows: Meteorological complexes appearing in historical T1 satellite cloud images are collected, and a characteristic meteorological complex is randomly selected from the acquired meteorological complexes. It should be noted here that: The area values ​​of the image region occupied by the characteristic meteorological complex in the T1 historical satellite cloud image are collected to obtain the T1 historical area value. The area values ​​of the image region occupied by the characteristic meteorological complex in the T1 target satellite cloud image are collected to obtain the T1 target area value. The area deviation between the T1 historical area value and the T1 target area value is calculated, and the ratio of the obtained difference to the T1 target area value is calculated to obtain the first index deviation degree. The central pressure value of the characteristic meteorological complex in the T1 historical satellite cloud image is obtained to obtain the T1 historical central pressure value. The central pressure value of the characteristic meteorological complex in the T1 target satellite cloud image is obtained to obtain the T1 target central pressure value. The difference between the T1 historical central pressure value and the T1 target central pressure value is calculated. The ratio of the obtained difference to the T1 target central pressure value is used to obtain the second index deviation degree. Set a first indicator deviation preset value and a second indicator deviation preset value respectively. If the first indicator deviation is less than or equal to the first indicator deviation preset value and the second indicator deviation is less than or equal to the second indicator deviation preset value, then the characteristic meteorological complex is marked as a consistent meteorological complex. If the first indicator deviation is greater than the first indicator deviation preset value or the second indicator deviation is greater than the second indicator deviation preset value, then the characteristic meteorological complex is marked as a non-consistent meteorological complex. It should be noted here that: In this invention, the first index deviation is specifically the range deviation corresponding to the meteorological complex, and the second index deviation is specifically the intensity deviation corresponding to the meteorological complex. The preset value of the first index deviation is set to 5%, and the preset value of the second index deviation is set to 3%.

[0031] Repeat the process of classifying the characteristic meteorological complexes. Classify each meteorological complex that appears in the T1 target satellite cloud image. If there are inconsistent meteorological complexes, it is determined that the non-trajectory parameters of the T1 historical satellite cloud image and the T1 target satellite cloud image are inconsistent. If the non-trajectory parameters of the historical satellite cloud image and the target satellite cloud image are completely consistent, then a meteorological complex trajectory consistency analysis will be performed on the historical satellite cloud image and the target satellite cloud image, and the wind energy initial screening matching period will be obtained based on the analysis results. Specifically as follows: One meteorological complex is randomly selected from the meteorological complexes appearing in the target satellite cloud image as a sample meteorological complex. Two target satellite cloud images with adjacent acquisition times are obtained for the sample meteorological complex and combined to set up the target cloud image control group. In this way, multiple target cloud image control groups are obtained. Two historical satellite cloud images corresponding to the acquisition time of the target cloud image control group are obtained to obtain multiple historical cloud image control groups. It should be noted here that: If the two target satellite cloud images in the target cloud image control group were acquired at 15:00:05 and 15:01:05 respectively during the target resource optimization cycle, then the two historical satellite cloud images corresponding to the target satellite cloud images were also acquired at 15:00:05 and 15:01:05 during the sample historical wind cycle, and the target cloud image control group and the historical cloud image control group corresponded one-to-one.

[0032] The target satellite cloud images in the target cloud image control group were named the first target control cloud image and the second target control cloud image according to the order of acquisition time. The historical satellite cloud images in the historical cloud image control group were named the first historical control cloud image and the second historical control cloud image according to the order of acquisition time. In the first target comparison cloud map, the cloud map area covered by the sample meteorological complex is divided into several meteorological pixels. Then, one sample meteorological pixel is randomly selected from these pixels, and the geometric center of the cloud map area corresponding to the target wind region is acquired to obtain the center point of the wind region. A Cartesian coordinate system is created by taking the center point of the wind area as the origin and the due north direction as the positive y-axis, thus obtaining the first target coordinate system. Repeat the process of creating the first target coordinate system, and create Cartesian coordinate systems for the second target reference cloud map, the first historical reference cloud map, and the second historical reference cloud map respectively, to obtain the second target coordinate system, the first historical coordinate system, and the second historical coordinate system; The coordinate positions of the sample meteorological pixels in the first target coordinate system, the second target coordinate system, the first historical coordinate system, and the second historical coordinate system are collected to obtain the first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates. Please see Figure 3The first pixel target coordinates, the second pixel target coordinates, the first pixel historical coordinates, and the second pixel historical coordinates are used to create coordinate points in the first target comparison cloud map to obtain the first pixel target point, the second pixel target point, the first pixel historical point, and the second pixel historical point. The closed area determined by the first pixel target point, the second pixel target point, and the center point of the wind area is marked as the sample pixel target area. The closed area determined by the first pixel historical point, the second pixel historical point, and the center point of the wind area is marked as the sample pixel historical area. The overlapping area between the sample pixel target area and the sample pixel historical area is marked as the target historical overlapping area. The ratio of the area value of the target historical overlapping area to the sample pixel target area is calculated to obtain the path overlap area ratio corresponding to the sample meteorological pixel. It should be noted here that: In this invention, if a sample meteorological pixel disappears in any reference cloud map, then the corresponding cloud map does not have its corresponding coordinate point. At this time, the overlapping area between the target area of ​​the sample pixel and the historical area of ​​the sample pixel does not overlap, so the area value of the target historical overlapping area is 0, that is, the path overlap area ratio corresponding to the sample meteorological pixel is 0.

[0033] Repeat the process of obtaining the path overlap area ratio corresponding to the sample meteorological pixels. Obtain the path overlap area ratio for each meteorological pixel covered by the sample meteorological complex, and set a qualified value for the path overlap area ratio. If the path overlap area ratio is greater than or equal to the qualified value, the meteorological pixel is classified as a path consistent pixel. If the path overlap area ratio is less than the qualified value, the meteorological pixel is classified as a path inconsistent pixel. It should be noted here that: In this invention, the path overlap area ratio is specifically set to 85%.

[0034] The path consistency of the historical cloud map control group is obtained by calculating the ratio of the number of path-consistent pixels to the number of meteorological pixels. A path consistency benchmark interval is set. If the path consistency is within the path consistency benchmark interval, the historical cloud map control group is classified as a valid cloud map control group. If the path consistency is not within the path consistency benchmark interval, the historical cloud map control group is classified as an invalid cloud map control group. It should be noted here that: In this invention, the effective cloud map control group includes cases where the path consistency is at the boundary of the path consistency benchmark interval; The historical cloud map control group is divided into effective cloud map control groups to obtain multiple historical effective control groups. The path consistency degree corresponding to each historical effective control group is obtained. The path consistency degree of the multiple obtained path consistency degrees is compared numerically, and the path consistency degree with the smallest value is set as the lower limit of the path consistency degree benchmark interval. The upper limit of the path consistency degree involved here is 100%, that is, the movement path of the sample meteorological complex in the historical cloud map control group is completely consistent with the path in the target cloud map control group.

[0035] For each meteorological complex in the target satellite cloud image, the validity of the cloud image control group is judged, and the invalid cloud image control group is obtained based on the judgment results; If there is no invalid cloud image control group in the target satellite cloud image, the historical wind force cycle of the sample is divided into the wind energy initial screening matching cycle; if there is an invalid cloud image control group in the target satellite cloud image, the historical wind force cycle of the sample is divided into the wind force invalid matching cycle. Repeat the process of classifying the historical wind cycles of the sample, classify each historical wind cycle into its own type, and obtain multiple wind energy preliminary screening matching cycles based on the classification results. The wind energy initial screening matching period is matched with the target resource optimization period to match the cyclical changes of wind energy resources. The cyclical wind energy consistency is obtained based on the matching results. Based on the cyclical wind energy consistency, the wind energy initial screening matching period is divided into effective wind energy matching period and ineffective wind energy matching period to obtain wind energy cycle matching data. It should be noted here that: In this invention, the screening process for the effective matching cycle of wind energy is the same as that for the screening process of the effective matching cycle of photovoltaic energy, and the amount of wind energy resources here corresponds to the amount of solar radiation. Specifically, the amount of wind energy resources here refers to the power of air flow per unit area.

[0036] The resource optimization module optimizes the energy storage of the target virtual power plant in the target resource optimization cycle based on wind energy cycle matching data and photovoltaic cycle matching data. Specifically as follows: Acquire photovoltaic cycle matching data, and obtain multiple effective photovoltaic matching cycles based on the photovoltaic cycle matching data; acquire wind energy cycle matching data, and obtain multiple effective wind energy matching cycles based on the wind energy cycle matching data. Based on the effective matching period of photovoltaic power, the photovoltaic power generation of the target virtual power plant is predicted, and multiple predicted photovoltaic power generation are obtained based on the prediction results. Specifically as follows: Create a resource optimization timeline covering the time range of the target resource optimization cycle, mark several feature time points in the resource optimization timeline, and arbitrarily select a sample feature time point from the multiple feature time points obtained. It should be noted here that: In this invention, the time interval between any two consecutive characteristic time points is set to 10 seconds.

[0037] Substitute each effective photovoltaic matching cycle into the resource optimization time axis, and then determine the sample characteristic time points of the target virtual power plant in each effective photovoltaic matching cycle on the resource optimization time axis. Obtain the historical photovoltaic power generation corresponding to each sample characteristic time point, and label the obtained historical photovoltaic power generation from small to large as L1 photovoltaic power generation to Lc photovoltaic power generation. Collect the median of L1 photovoltaic power generation to Lc photovoltaic power generation to obtain Li photovoltaic power generation. It should be noted here that: L1, L2, L3...Lc in the photovoltaic power generation range are the marking symbols corresponding to the historical photovoltaic power generation. The variance of photovoltaic power generation from Li to Li+1 is calculated to obtain the variance of Li+1. The variance of photovoltaic power generation from Li+1 to Li-1 is calculated to obtain the variance of Li-1. The variances of Li+1 and Li-1 are compared. If the variance of Li+1 is greater than or equal to the variance of Li-1, the variance of Li-1 is marked as a temporary characteristic variance. If it is less than, the variance of Li+1 is marked as a temporary characteristic variance. The variance of photovoltaic power generation from Li-1 to Li+2 is calculated to obtain the variance of Li+2. If the variance of Li+2 is less than the temporary characteristic variance, the temporary characteristic variance is reassigned to the variance of Li+2. If it is greater than or equal to the variance of Li, the value of the temporary characteristic variance is not changed. The variance of photovoltaic power generation from Li to Li-2 is calculated to obtain the variance of Li-2. If the variance of Li-2 is less than the temporary characteristic variance, the temporary characteristic variance is reassigned to the variance of Li-2. If it is greater than or equal to the variance of Li, the value of the temporary characteristic variance is not changed. Similarly, the variance of photovoltaic power generation from Li-(d-1) to Li+d is calculated to obtain the variance of Li+d. If the variance of Li+d is less than the temporary characteristic variance, the variance of Li+d is reassigned to the temporary characteristic variance. If it is greater than or equal to the variance of Li+d, the value of the temporary characteristic variance is not changed. The variance of photovoltaic power generation from Li+d to Li-d is calculated to obtain the variance of Li-d. If the variance of Li-d is less than the temporary characteristic variance, the variance of Li-d is reassigned to the temporary characteristic variance. If it is greater than or equal to the variance of Li-d, the value of the temporary characteristic variance is not changed. It should be noted here that: In this invention, i+d=c, id=1.

[0038] Perform the above iterative operation, name the final temporary feature variance as the final value feature variance, calculate the average of the photovoltaic power generation power used to calculate the final value feature variance, and obtain the predicted photovoltaic power generation power corresponding to the sample feature time point. It should be noted here that: If the photovoltaic power used to calculate the characteristic variance of the final value is from L8 to L17, then the average of the photovoltaic power from L8 to L17 is calculated to obtain the predicted photovoltaic power corresponding to the characteristic time point of the sample. Repeat the process of obtaining the predicted photovoltaic power generation corresponding to the feature time points of the sample, and obtain the predicted photovoltaic power generation corresponding to each feature time point respectively; Based on the effective matching period of photovoltaic power, wind power generation of the target virtual power plant is predicted, and multiple predicted wind power generation is obtained based on the prediction results. Calculate the sum of the predicted photovoltaic power generation and the predicted wind power generation at each characteristic time point to obtain the predicted power generation of the target virtual power plant at each characteristic time point; Create an energy storage analysis coordinate system. In the energy storage analysis coordinate system, connect the coordinate points with the characteristic time point on the horizontal axis and the predicted power generation on the vertical axis in sequence to obtain the power generation cycle prediction curve. It should be noted here that: In this energy storage analysis coordinate system, the x-axis represents time values, and the y-axis represents power values.

[0039] The power generation demand of the target virtual power plant at different characteristic time points is collected and created as a power generation periodic demand curve in the energy storage analysis coordinate system; Please see Figure 4 In the energy storage analysis coordinate system, when the power generation cycle prediction curve is higher than the power generation cycle demand curve, the closed area enclosed by the power generation cycle prediction curve and the power generation cycle demand curve is set as the energy storage dispatch area. The x-axis region covered by the energy storage dispatch area is set as the energy storage demand period, and the area value of the region covered by the energy storage dispatch area is set as the energy storage space demand corresponding to the energy storage demand period. The energy storage space demand corresponding to the energy storage demand period is fed back to the energy storage control terminal, and the energy storage control terminal performs energy storage allocation for the target virtual power plant.

[0040] Compared to the problems described in the background art, this invention, on the one hand, obtains the target noon extra-layer radiation by analyzing the annual accumulated days and noon solar altitude angle corresponding to the target resource optimization cycle. Based on the target noon extra-layer radiation, a photovoltaic (PV) initial screening matching cycle is matched to the target resource optimization cycle. Furthermore, a time-specific solar radiation consistency analysis is performed on the PV initial screening matching cycle to further refine the matching cycle and obtain an effective PV matching cycle. This fully ensures the adaptability of the PV power generation prediction sample cycle, thus guaranteeing the accuracy of PV power prediction. On the other hand, by collecting satellite cloud images corresponding to the target resource optimization cycle and comparing them with historical cycle satellite cloud images, non-trajectory meteorological parameters are compared, and a meteorological complex trajectory consistency analysis is performed. Based on the analysis results... To match the target resource optimization cycle with the initial wind power matching cycle, a time-based wind energy resource consistency analysis is performed on the initial wind power matching cycle to further refine the matching cycle and obtain the effective wind power matching cycle. This can accurately match the effective wind power cycle that highly matches the meteorological characteristics of the target cycle, reduce prediction bias, and significantly improve the accuracy and reliability of wind power generation prediction. Finally, based on the effective wind power matching cycle and the effective photovoltaic matching cycle, a power generation cycle demand curve is created for the target virtual power plant in the target resource optimization cycle. Based on the power generation cycle demand curve, the power storage of the target virtual power plant is optimized, enabling on-demand dynamic allocation of power storage and effectively improving the operational stability of the virtual power plant and the capacity for renewable energy absorption.

[0041] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A virtual power plant resource optimization management method based on big data, characterized in that, include: Step S1: Conduct photovoltaic monitoring on the target resource area, create a target resource optimization cycle based on the monitoring results, and perform historical photovoltaic cycle matching on the target resource optimization cycle in combination with historical photovoltaic monitoring results to obtain photovoltaic cycle matching data; Step S2: Conduct photovoltaic monitoring on the target resource area, and match the historical wind cycle with the target resource optimization cycle based on the monitoring results to obtain wind energy cycle matching data; Step S3: Optimize the power storage of the target virtual power plant in the target resource optimization cycle based on the wind energy cycle matching data and the photovoltaic cycle matching data.

2. The method for optimizing virtual power plant resource management based on big data according to claim 1, characterized in that, The specific steps in step S1 are as follows: Step S11: Obtain the virtual power plant that needs resource optimization management, obtain the target virtual power plant, and create a target resource optimization cycle for the target virtual power plant; Step S12: Perform photovoltaic resource analysis on the target virtual power plant in the target resource optimization cycle, and obtain the photovoltaic initial screening matching cycle based on the analysis results; Step S13: Randomly select one sample photovoltaic screening matching cycle from the multiple historical photovoltaic screening matching cycles obtained, and perform a periodic variation analysis of solar radiation between the sample photovoltaic screening matching cycle and the target resource optimization cycle. Obtain the periodic radiation consistency based on the analysis results. Step S14: Obtain the periodic radiation consistency degree corresponding to each initial photovoltaic matching cycle, set the periodic radiation consistency degree benchmark interval, if the periodic radiation consistency degree is within the periodic radiation consistency degree benchmark interval, then the corresponding initial photovoltaic matching cycle is divided into a valid photovoltaic matching cycle, if not, then the corresponding initial photovoltaic matching cycle is divided into an invalid photovoltaic matching cycle, and obtain photovoltaic periodic matching data.

3. The method for optimizing virtual power plant resource management based on big data according to claim 2, characterized in that, In step S12, the specific steps are as follows: The photovoltaic power generation area corresponding to the target virtual power plant is marked as the target photovoltaic area. The calendar date of the target resource optimization cycle is obtained and converted into an annual day to obtain the target annual day. The noon solar altitude angle is obtained based on the target annual day. The solar radiation outside the atmosphere at noon corresponding to the target resource optimization cycle is calculated based on the noon solar altitude angle and the target year accumulated days, and the target noon extra-atmosphere radiation is obtained. Obtain the noon solar radiation outside the atmosphere corresponding to several historical illumination monitoring cycles, calculate the difference between the noon solar radiation outside the atmosphere and the target noon solar radiation outside the atmosphere for each cycle, and calculate the absolute value of the difference to obtain multiple noon solar radiation deviations. Set a preset reasonable value for the solar radiation deviation. If the noon solar radiation deviation is less than or equal to the preset reasonable value, the historical illumination monitoring cycle is divided into a photovoltaic initial screening matching cycle. If it is greater than the preset reasonable value, the historical illumination monitoring cycle is divided into a photovoltaic matching anomaly cycle.

4. The method for optimizing virtual power plant resource management based on big data according to claim 2, characterized in that, In step S13, the specific steps are as follows: Create a periodic radiance coordinate system to predict the solar radiation of the target photovoltaic area during the target resource optimization cycle. Based on the prediction results, create a target radiation variation curve in the periodic radiance coordinate system. Collect the actual solar radiation of the target photovoltaic area during the sample photovoltaic initial screening matching cycle and plot the sample radiation variation curve. Select any characteristic curve point in the target radiation change curve, and create a vertical comparison line with the characteristic curve point as the center point. If the vertical comparison line can cover the sample radiation change curve in the vertical direction, the characteristic curve point is marked as an effective coverage point. If it cannot cover the vertical direction, the characteristic curve point is marked as an invalid coverage point. The longitudinal comparison line is used to traverse each curve point in the target radiation variation curve, and the quantity value corresponding to the effective coverage point is counted as A1, and the quantity value corresponding to the ineffective coverage point is counted as A2, and the periodic radiation consistency is calculated.

5. The method for optimizing virtual power plant resource management based on big data according to claim 1, characterized in that, In step S2, the specific steps are as follows: Step S21: Obtain the target resource optimization cycle and mark the wind power generation area corresponding to the target virtual power plant as the target wind power area; Step S22: Predict the satellite cloud image of the target wind area within the target resource optimization period, and mark the predicted satellite cloud images as T1 target satellite cloud image to Ta target satellite cloud image in chronological order to obtain the target satellite cloud image sequence; Step S23: Collect multiple historical wind cycles, match the historical wind cycles with the target resource optimization cycle using satellite cloud images, and obtain the initial wind energy screening matching cycle based on the matching results; Step S24: Match the wind energy initial screening matching period with the target resource optimization period to the wind energy resource quantity period change. Obtain the periodic wind energy consistency based on the matching results. Divide the wind energy initial screening matching period into effective wind energy matching period and ineffective wind energy matching period based on the periodic wind energy consistency to obtain wind energy period matching data.

6. The method for optimizing virtual power plant resource management based on big data according to claim 5, characterized in that, In step S23, the specific steps are as follows: Step S231: Randomly select one sample historical wind cycle from the multiple historical wind cycles obtained; Step S232: Collect historical satellite cloud images corresponding to the historical wind force cycles of the sample to obtain historical satellite cloud images from T1 to Ta. Step S233: Perform meteorological complex identification on the target satellite cloud image. If a meteorological complex exists in the target satellite cloud image, compare the non-trajectory parameters of the T1 historical satellite cloud image and the T1 target satellite cloud image. If the comparison results are consistent, continue to compare the T2 historical satellite cloud image and the T2 target satellite cloud image in the same way. If the comparison results are inconsistent, directly determine that the sample's historical wind force cycle is an invalid wind force matching cycle. Continue in this manner until the consistency comparison between the Ta historical satellite cloud image and the Ta target satellite cloud image is completed. Step S234: If the non-trajectory parameters of the historical satellite cloud image and the target satellite cloud image are completely consistent, then perform a meteorological complex trajectory consistency analysis on the historical satellite cloud image and the target satellite cloud image. Based on the analysis results, collect an invalid cloud image control group for the target satellite cloud image. If there is no invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind energy initial screening matching cycle. If there is an invalid cloud image control group in the target satellite cloud image, then divide the historical wind force cycle of the sample into a wind force invalid matching cycle.

7. The method for optimizing virtual power plant resource management based on big data according to claim 6, characterized in that, In step S234, the specific steps are as follows: One meteorological complex is randomly selected from the meteorological complexes appearing in the target satellite cloud image as a sample meteorological complex. Satellite cloud images are collected for the sample meteorological complex to obtain the target cloud image control group and the historical cloud image control group. The target satellite cloud images in the target cloud image control group were named the first target control cloud image and the second target control cloud image according to the order of acquisition time. The historical satellite cloud images in the historical cloud image control group were named the first historical control cloud image and the second historical control cloud image according to the order of acquisition time. In the first target comparison cloud map, the cloud map area covered by the sample meteorological complex is divided into several meteorological pixels. Then, one sample meteorological pixel is randomly selected from these pixels, and the geometric center of the target wind region is acquired to obtain the center point of the wind region. Using the center point of the wind area as the origin, a Cartesian coordinate system is created to obtain the first target coordinate system; A Cartesian coordinate system is created for the second target comparison cloud map, the first historical comparison cloud map, and the second historical comparison cloud map to obtain the second target coordinate system, the first historical coordinate system, and the second historical coordinate system.

8. The method for optimizing virtual power plant resource management based on big data according to claim 7, characterized in that, In step S234, the specific steps are as follows: Analyze the position of sample meteorological pixels in different coordinate systems to obtain the target area and historical area of ​​sample pixels. Mark the overlapping area of ​​the target area and historical area of ​​sample pixels as the target historical overlapping area. Calculate the ratio of the area of ​​the target historical overlapping area to that of the target area of ​​sample pixels to obtain the path overlap area ratio. The path overlap area ratio is obtained for each meteorological pixel covered by the sample meteorological complex, and a qualified value for the path overlap area ratio is set. If the path overlap area ratio is greater than or equal to the qualified value, the meteorological pixel is classified as a path consistent pixel; if it is less than the qualified value, the meteorological pixel is classified as a path inconsistent pixel. The path consistency of the historical cloud map control group is obtained by calculating the ratio of the number of path-consistent pixels to the number of meteorological pixels. A path consistency baseline range is set. If the path consistency is within the path consistency baseline range, the historical cloud map control group is classified as a valid cloud map control group. If it is not within the baseline range, the historical cloud map control group is classified as an invalid cloud map control group. For each meteorological complex in the target satellite cloud image, the validity of the cloud image control group is judged, and the invalid cloud image control group is obtained based on the judgment results.

9. The method for optimizing virtual power plant resource management based on big data according to claim 1, characterized in that, In step S3, the specific steps are as follows: Acquire photovoltaic cycle matching data, and obtain multiple effective photovoltaic matching cycles based on the photovoltaic cycle matching data; acquire wind energy cycle matching data, and obtain multiple effective wind energy matching cycles based on the wind energy cycle matching data. Based on the effective matching cycle of photovoltaic and wind power, the power generation of the target virtual power plant in the target resource optimization cycle is predicted, and a power generation cycle prediction curve is created. The power generation demand of the target virtual power plant at different characteristic time points is collected and created as a power generation periodic demand curve in the energy storage analysis coordinate system; In the energy storage analysis coordinate system, when the power generation cycle prediction curve is higher than the power generation cycle demand curve, the closed area enclosed by the power generation cycle prediction curve and the power generation cycle demand curve is set as the energy storage dispatch area. The x-axis region covered by the energy storage dispatch area is set as the energy storage demand period. The area value corresponding to the energy storage dispatch area is set as the energy storage space demand corresponding to the energy storage demand period. The energy storage space demand corresponding to the energy storage demand period is fed back to the energy storage control terminal, and the energy storage control terminal performs energy storage allocation for the target virtual power plant.

10. A virtual power plant resource optimization management system based on big data, applicable to the virtual power plant resource optimization management method based on big data as described in any one of claims 1-9, characterized in that, The resource optimization management system includes: Photovoltaic matching module: performs photovoltaic monitoring on the target resource area, creates a target resource optimization cycle based on the monitoring results, performs historical photovoltaic cycle matching on the target resource optimization cycle based on historical photovoltaic monitoring results, and obtains photovoltaic cycle matching data based on the matching results; Wind matching module: Monitors photovoltaic power in the target resource area, matches the target resource optimization cycle with historical wind cycles based on the monitoring results, and obtains wind energy cycle matching data based on the matching results; Resource optimization module: Optimizes the energy storage of target virtual power plants in the target resource optimization cycle based on wind energy cycle matching data and photovoltaic cycle matching data.

Images