New Energy Power Generation Prediction Method and Platform Based on Collected Data

Abstract

The present invention provides a new energy power generation prediction method and platform based on collected data, relating to data processing technology. The method includes: a server obtains a first power generation time point when a first power generation group starts initial power generation, and obtains first synchronization data of the first power generation group's history based on the first power generation time point; decomposes the data volume of the first synchronization data according to a period to obtain first sub-information under different periods, determines a target period based on the current time point, and determines the first sub-information corresponding to the target period to obtain a first prediction sample set; traverses second power generation groups corresponding to the attributes of the first power generation group in a database, and fills in the data of the first prediction sample set based on the second synchronization data of the second power generation group's history to obtain a second prediction sample set; predicts the power generation of the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result. The present invention can improve the accuracy of power generation prediction for power generation groups.

Description

Technical Field

[0001] The present invention relates to data processing technologies, and in particular, to a new energy power generation prediction method and platform based on collected data. Background Art

[0002] With the increasing global demand for renewable energy, new energy power generation systems such as photovoltaic and wind power generation have gradually become an important part of power supply. These systems usually consist of multiple power generation units, each operating under different geographical locations and environmental conditions. To make full use of natural resources and improve power generation efficiency, accurate prediction of power generation is crucial for grid stability and optimal allocation of resources. However, due to the variability of environmental factors and problems such as equipment aging, the prediction of new energy power generation faces multiple challenges.

[0003] Most of the existing new energy power generation prediction methods rely on simple statistical models for prediction. These methods usually fail to fully consider the deployment time of power generation units and the impact of environmental changes on power generation, resulting in low prediction accuracy.

[0004] Therefore, how to improve the accuracy of power generation prediction for power generation units has become an urgent problem to be solved. Summary of the Invention

[0005] An embodiment of the present invention provides a new energy power generation prediction method and platform based on collected data, which can improve the accuracy of power generation prediction for power generation units.

[0006] In a first aspect of an embodiment of the present invention, a new energy power generation prediction method based on collected data is provided, including:

[0007] A server obtains a first power generation time point when a first power generation unit starts initial power generation, and obtains first synchronization data of the first power generation unit's history based on the first power generation time point, where the first synchronization data is the collected data of the first power generation unit at historical moments;

[0008] The data volume of the first synchronization data is decomposed and processed according to a period to obtain first sub-information under different periods, a target period is determined based on the current time point, and the first sub-information corresponding to the target period is determined to obtain a first prediction sample set;

[0009] If it is determined that the prediction sample set does not meet the power generation prediction requirements, then second power generation units corresponding to the attributes of the first power generation unit in the database are traversed, and the first prediction sample set is supplemented with data based on the second synchronization data of the second power generation units' history to obtain a second prediction sample set;

[0010] Based on the first prediction sample set or the second prediction sample set, power generation prediction is performed on the first power generation unit to obtain a new energy power generation prediction result.

[0011] Optionally, decomposing the data volume of the first synchronization data according to a period to obtain first sub-information under different periods, determining a target period based on the current time point, and determining the first sub-information corresponding to the target period to obtain a first prediction sample set, including:

[0012] Extract the earliest time of the first synchronization data, determine the annual time period between the earliest time and the current time based on the earliest time, and establish a first annual data axis corresponding to each annual time period;

[0013] Based on the first annual data axis, establish corresponding display slots, place each first annual data axis in the display slots, establish a division slot below the last display slot, and establish a division data axis corresponding to the first annual data axis in the division slot;

[0014] Based on the division data axis, divide all the first annual data axes to obtain a target period and retrieve the corresponding first sub-information to obtain a first prediction sample set.

[0015] Optionally, the extracting the earliest time of the first synchronization data, determining the annual time period between the earliest time and the current time based on the earliest time, and establishing a first annual data axis corresponding to each annual time period, including:

[0016] Determine the first quantity of pixel points corresponding to the first annual data axis;

[0017] Determine the number of days corresponding to each first annual data axis, perform normalization processing on the number of days of all first annual data axes to obtain a unified number of days, obtain the day time information corresponding to each group of pixel points based on the first quantity and the unified number of days, establish a storage format for the collected data based on the day time information, and each day time information corresponds to a storage unit;

[0018] According to the order of each group of pixel points, add the corresponding day time information to each group of pixel points, and store the information in the collected data corresponding to the day time information and the corresponding group of pixel points, so that each group of pixel points has a corresponding storage unit.

[0019] Optionally, the dividing all the first annual data axes based on the division data axis to obtain a target period and retrieving the corresponding first sub-information to obtain a first prediction sample set, including:

[0020] Determine a target period based on the current time point, each current time point has a preset target period, and extract the initial time point and the cut-off time point of the target period;

[0021] Determine the day time information and pixel points corresponding to the initial time point and the cut-off time point respectively in the division data axis to obtain an initial pixel point and a cut-off pixel point;

[0022] Based on the initial pixel point and the cut-off pixel point, all the first-year data axes are divided, and the first sub-information is retrieved based on the storage format of the collected data to obtain the first prediction sample set.

[0023] Optionally, the step of dividing all the first-year data axes based on the initial pixel point and the cut-off pixel point, and retrieving the first sub-information based on the storage format of the collected data to obtain the first prediction sample set includes:

[0024] An initial division line perpendicular to the divided data axis is established upward starting from the initial pixel point, and a cut-off division line perpendicular to the divided data axis is established upward starting from the cut-off pixel point;

[0025] All pixel points of each first-year data axis located within the initial division line and the cut-off division line are obtained, and the first sub-information of the corresponding storage unit in the storage format is retrieved according to the order of the pixel points to obtain the first prediction sample set.

[0026] Optionally, the step of obtaining all pixel points of each first-year data axis located within the initial division line and the cut-off division line, and retrieving the first sub-information of the corresponding storage unit in the storage format according to the order of the pixel points to obtain the first prediction sample set includes:

[0027] If it is determined that the user selects and moves the initial division line and / or the cut-off division line on the interaction end, the initial division line and / or the cut-off division line are synchronously moved in response to the user's movement;

[0028] After it is determined that the movement of the initial division line and / or the cut-off division line ends, the pixel points of the initial division line and / or the cut-off division line are counted;

[0029] If it is determined that the user double-clicks at any position of the initial division line and / or the cut-off division line on the interaction end, the initial division line and / or the cut-off division line are split at the double-click position to obtain multiple moving split lines;

[0030] If it is determined that the user selects and moves any of the moving split lines on the interaction end, the moving split line is synchronously moved in response to the user's movement, and the pixel points between the moving split lines, between the moving split line and the cut-off division line, or between the initial division line and the moving split line are counted.

[0031] Optionally, the step of, if it is determined that the prediction sample set does not meet the power generation prediction requirements, traversing the second power generation groups corresponding to the attributes of the first power generation group in the database, and complementing the data of the first prediction sample set based on the historical second synchronization data of the second power generation groups to obtain the second prediction sample set includes:

[0032] If the current predicted quantity of the first sub - information in the predicted sample set is less than the predicted threshold quantity, it is determined that the predicted sample set does not meet the power generation prediction requirements;

[0033] Calculate the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity, and traverse the second power generation groups corresponding to the attributes of the first power generation group in the database;

[0034] Generate a display slot and a second - year data axis corresponding to the second power generation group, and complete the data of the first predicted sample set based on the historical second synchronization data of the second power generation group to obtain a second predicted sample set.

[0035] Optionally, the calculating the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity, and traversing the second power generation groups corresponding to the attributes of the first power generation group in the database includes:

[0036] Obtain the attributes of the first power generation group, where the attributes at least include power generation type attributes, power generation specification attributes, power generation region attributes, and power generation environment attributes;

[0037] Based on the power generation type attributes, determine the screening of other power generation groups in the database to determine other power generation groups with the same power generation type;

[0038] Calculate the fusion similarity values of other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes;

[0039] Based on the fusion similarity values and the missing predicted quantity, determine the second power generation group among other power generation groups, and add corresponding prediction offset labels to the second power generation group based on each fusion similarity value.

[0040] Optionally, the calculating the fusion similarity values of other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes includes:

[0041] Conduct discrimination processing on the power generation specification attributes of the power generation group to obtain power generation specification coefficients in different dimensions, where the power generation specification coefficients at least include rated power generation, area information, and / or volume specification information;

[0042] Conduct quantization processing on the power generation region attributes of the power generation group to obtain region quantization coefficients, where the region quantization coefficients at least include plain, hilly, mountainous, and urban;

[0043] Conduct discrimination processing on the power generation environment attributes of the power generation group to obtain power generation environment coefficients in different dimensions, where the power generation environment coefficients at least include temperature coefficient and humidity coefficient;

[0044] Based on the power generation specification coefficients, region quantization coefficients, and power generation environment coefficients, conduct fusion calculation to obtain the fusion similarity values of other power generation groups and the first power generation group.

[0045] Optionally, the fusion calculation based on the power generation specification coefficient, the regional quantization coefficient, and the power generation environment coefficient to obtain the fusion similarity value between the other power generation groups and the first power generation group includes:

[0046] Calculate the fusion similarity value through the following formula

[0047] ;

[0048] where is the fusion similarity value between the c th power generation group and the first power generation group, is the power generation specification weight, is the power generation specification coefficient of the i th dimension, is the specification coefficient weight of the i th dimension, n is the upper limit value of the dimension of the power generation specification, is the regional specification weight, is the regional specification coefficient of the m th dimension, is the specification coefficient weight of the m th dimension, q is the upper limit value of the dimension of the regional specification, is the environmental specification weight, is the environmental specification coefficient of the r th dimension, is the specification coefficient weight of the r th dimension, e is the upper limit value of the dimension of the environmental specification.

[0049] Optionally, determining the second power generation group based on the fusion similarity value and the missing prediction quantity in the other power generation groups, and adding corresponding prediction offset labels to the second power generation group based on each fusion similarity value includes:

[0050] Sort all the other power generation groups based on the fusion similarity value to obtain a similarity value sequence;

[0051] Successively obtain the second-year data axis of the first power generation group in the similarity value sequence and add it above the first-year data axis, and obtain the pixel points and second sub-information corresponding to the second-year data based on the initial division line, the cut-off division line, and the moving split line;

[0052] If the quantity of the second sub-information is greater than or equal to the missing prediction quantity, determine that the currently obtained second sub-information meets the requirements, and select the second power generation group corresponding to the currently obtained second sub-information, and the second sub-information belongs to the second synchronous data of the corresponding second power generation group;

[0053] Determine the corresponding prediction offset label based on the fusion similarity value of the second power generation group, and each similarity value interval has a preset prediction offset label.

[0054] Optionally, it further includes:

[0055] If the number of the second sub-information is less than the number of missing predictions, select other power generation groups again according to the similarity value sequence, and add the second annual data axis of the other power generation groups above the previous second annual data axis;

[0056] Obtain the pixel points and the second sub-information corresponding to the newly added second annual data based on the initial division line, the cut-off division line, and the moving split line until the number of the second sub-information is greater than or equal to the number of missing predictions or there are no unselected power generation groups in the similarity value sequence.

[0057] Optionally, the power generation prediction of the first power generation group based on the first prediction sample set or the second prediction sample set to obtain the new energy power generation prediction result includes:

[0058] If it is determined that only the first prediction sample set exists, no prediction interval is generated, and the new energy power generation prediction result at the predicted target time is calculated based on the first prediction sample set through the following formula,

[0059] ;

[0060] ;

[0061] ;

[0062] Wherein, is the new energy power generation prediction result at the predicted target time calculated based on the first prediction sample set, is the average power generation result of all the first sub-information in the first prediction sample set, is the weight value of the F dimension that affects the power generation result, is the value of the time predicted by the first power generation group in the F dimension, is the average value of the F dimension that affects the power generation result in the first prediction sample set, B is a constant value, G is the upper limit value of the dimension that affects the power generation result, is the power generation result of the th first sub-information in the first prediction sample set, is the upper limit value of the first prediction sample, is the number value of the first prediction sample, is the The value of a first sub-information in the F numerical value of the dimension, D is the upper limit value of the first sub-information in the first prediction sample set, X is the quantity value of the first sub-information in the first prediction sample set;

[0063] If it is determined that there is a second prediction sample set, the first prediction sample subset of the first power generation group and the second prediction sample subset of the second power generation group are obtained by dividing the second prediction sample set;

[0064] The first prediction sample subset is calculated and processed according to the calculation method when only the first prediction sample set exists to obtain the first sub-power generation result, the second sub-power generation result of the second prediction sample subset is calculated, and the new energy power generation prediction result is obtained by fusing the first sub-power generation result, the second sub-power generation result, and the prediction offset label.

[0065] Optionally, the first prediction sample subset is calculated and processed according to the calculation method when only the first prediction sample set exists to obtain the first sub-power generation result, the second sub-power generation result of the second prediction sample subset is calculated, and the new energy power generation prediction result is obtained by fusing the first sub-power generation result, the second sub-power generation result, and the prediction offset label, including:

[0066] The second sub-power generation result is calculated through the following formula,

[0067] ;

[0068] ;

[0069] wherein, is the second sub-power generation result, is the prediction result based on the second electronic information of the I th second power generation group, h is the upper limit value of the second power generation group, is the quantity value of the second power generation group, is the I th average power generation value of the second power generation group, is the average value of the I th second power generation group in the T dimension, is the average value of the first power generation group in the T dimension, is the weight value of the T dimension, N is the upper limit value of the dimension when the first power generation group and the second power generation group are compared, M is the quantity value of the dimension when the first power generation group and the second power generation group are compared;

[0070] Calculate the mean of the first sub-generation result and the second sub-generation result to obtain a predicted mean. Determine the sum of the offset coefficients by adding up the offset coefficients corresponding to each predicted offset label, where each offset label has a preset offset coefficient.

[0071] Perform upward and downward offset processing on the predicted mean based on the sum of the offset coefficients to obtain a predicted result interval, which is used as the new energy power generation prediction result.

[0072] In the second aspect of the embodiments of the present invention, a new energy power generation prediction platform based on collected data is provided, including:

[0073] A time module, configured to obtain, for the server, the first power generation time point when the first power generation group performs initial power generation, and obtain the first synchronization data of the history of the first power generation group based on the first power generation time point, where the first synchronization data is the collected data of the first power generation group at historical moments.

[0074] A decomposition module, configured to perform cycle decomposition processing on the data volume of the first synchronization data to obtain first sub-information under different cycles, determine a target cycle based on the current time point, and determine the first sub-information corresponding to the target cycle to obtain a first prediction sample set.

[0075] A complementing module, configured to, if it is determined that the prediction sample set does not meet the power generation prediction requirements, traverse the second power generation groups corresponding to the attributes of the first power generation group in the database, and perform data complementing on the first prediction sample set based on the second synchronization data of the history of the second power generation groups to obtain a second prediction sample set.

[0076] A prediction module, configured to perform power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result.

[0077] In the third aspect of the embodiments of the present invention, a storage medium is provided, in which a computer program is stored, and when the computer program is executed by a processor, it is used to implement the method described in the first aspect and various possible designs of the first aspect of the present invention.

[0078] The present invention collects and integrates the historical data of the first power generation group and similar power generation groups to form a complete prediction sample set. This process solves the problem of prediction deviation caused by insufficient or incomplete data in traditional methods. By integrating historical data of multi-dimensional environmental factors (such as temperature, humidity, geographical attributes, etc.), the present invention effectively improves the accuracy and adaptability of the prediction model, and significantly improves the accuracy of new energy power generation prediction.

[0079] The present invention introduces a dynamic data filling mechanism. When the prediction sample set is insufficient, it can automatically screen out data of other power generation groups with similar attributes from the database for supplementation. This mechanism ensures the integrity and timeliness of the data set. Combining with the user interaction function, it allows users to flexibly adjust the data range and analysis period on the visualization interface, enabling the prediction system to respond to external environment changes and user needs in real time, and further improving the reliability of the prediction results.

[0080] By establishing an annual data axis and display slots, the present invention realizes the systematic and standardized management of historical data. Using the method of pixel points and dividing lines for data segmentation and cycle analysis makes data processing more intelligent and intuitive. Users can perform intuitive data display and analysis through a friendly interface, effectively supporting the multi-dimensional analysis and management of new energy power generation. This intelligent data processing method ensures the accurate application of data within each time period, supporting more complex prediction models and analysis strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 is a schematic flowchart of a new energy power generation prediction method based on collected data provided by an embodiment of the present invention.

[0082] Figure 2 is a schematic diagram of a display slot provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0084] See Figure 1 , which is a schematic flowchart of a new energy power generation prediction method based on collected data provided by an embodiment of the present invention, including:

[0085] S1, the server obtains the first power generation time point when the first power generation group starts initial power generation, and obtains the first synchronization data of the history of the first power generation group based on the first power generation time point. The first synchronization data is the collected data of the first power generation group at historical moments.

[0086] First of all, it should be noted that in this solution, considering different deployment times of the power generation group, the corresponding predicted power generation results are different. For example, the predicted power generation results corresponding to the deployment of power generation group A five years ago and ten years ago are different. Therefore, this solution will obtain the first power generation time point when the first power generation group starts initial power generation, and the first power generation time point is the deployment time corresponding to the first power generation unit. Among them, the first power generation group can be a new energy power generation equipment group such as photovoltaic or wind energy.

[0087] After obtaining the first power generation time of the first power generation group, the historical data of the first power generation group, that is, the above-mentioned first synchronization data, can be obtained by combining the current time and the first power generation time. It can be understood that the first synchronization data includes acquisition data, and the acquisition data can be data such as power generation amount, temperature, humidity, wind speed, etc.

[0088] S2. Decompose the data volume of the first synchronization data according to the cycle to obtain the first sub-information in different cycles, determine the target cycle based on the current time point, and determine the first sub-information corresponding to the target cycle to obtain the first prediction sample set.

[0089] Since the data volume of the first synchronization data may be large, this solution needs to decompose the data volume of the first synchronization data to obtain multiple first sub-informations. These cycles may be divided according to time cycles such as annual, quarterly, and monthly. Among them, the target cycle corresponds to the current time point. For example, when divided by quarter, if the current time point is January 20th, then the corresponding target cycle can be the quarter corresponding to January 20th.

[0090] In some embodiments, the step of decomposing the data volume of the first synchronization data according to the cycle to obtain the first sub-information in different cycles, determining the target cycle based on the current time point, and determining the first sub-information corresponding to the target cycle to obtain the first prediction sample set includes:

[0091] S21. Extract the earliest moment of the first synchronization data, determine the annual time periods existing between the earliest moment and the current moment based on the earliest moment, and establish the first annual data axis corresponding to each annual time period.

[0092] This solution will extract the earliest moment from the first synchronization data and determine the annual time periods from that moment to the current moment. For example, if the earliest moment is 2020 and the current moment is 2023, then the corresponding annual time periods are 2020, 2021, 2022, and 2023 respectively. For each annual time period, establish a corresponding annual data axis for further data analysis and processing. Among them, see Figure 2 , the annual data axis can be a horizontal data axis.

[0093] Among them, the earliest time for extracting the first synchronization data is used to determine the annual time period between the earliest time and the current time, and a first annual data axis corresponding to each annual time period is established, including:

[0094] S211, determine the first quantity of the pixel points corresponding to the first annual data axis.

[0095] For subsequent detailed division of the first annual data axis, this solution will determine the quantity of pixel points corresponding to each annual data axis. For example, it is 1,000 pixel points.

[0096] S212, determine the number of days corresponding to each first annual data axis, perform normalization processing on the number of days of all first annual data axes to obtain a unified number of days, obtain the day time information corresponding to each group of pixel points based on the first quantity and the unified number of days, and establish a storage format for the collected data based on the day time information, where each day time information corresponds to a storage unit.

[0097] First, for each annual data axis, it is first necessary to determine the number of days included in that year. This may involve considering the different numbers of days in leap years and common years. For the convenience of subsequent processing, the number of days of all annual data axes is normalized. The goal of normalization is to convert the number of days in different years into a unified standard for easy analysis and comparison. Through this normalization processing, data inconsistency problems caused by year differences (such as leap years and common years) can be avoided. Based on the determined first quantity and the unified number of days, calculate the day time information corresponding to each group of pixel points. Use this day time information to establish the storage format of the data, where each day time information corresponds to a storage unit. This can ensure the logical and orderly storage of the data.

[0098] S213, according to the order of each group of pixel points, add the corresponding day time information to each group of pixel points, and store the information in the collected data corresponding to the day time information and the corresponding group of pixel points, so that each group of pixel points has a corresponding storage unit.

[0099] This solution will add corresponding daily time information to each group of pixel points in sequence. This step ensures the accurate positioning of each group of data in the time dimension. Then, the information in the collected data is stored corresponding to the daily time information and the corresponding group of pixel points. Such a correspondence ensures the accuracy and traceability of the data. Each group of pixel points (e.g., 5 pixel points) will have a corresponding storage unit, which can effectively store and manage the data associated with it. Through these steps, the standardized and systematic management of a large amount of collected data can be achieved. This method not only improves the efficiency of data storage but also lays a good foundation for subsequent data analysis and prediction. Through strict normalization processing and the design of the storage format, the integrity and consistency of the data can be ensured.

[0100] S22. Based on the first-year data axis, establish corresponding display slots, place each first-year data axis in the display slots, establish a division slot below the last display slot, and establish a division data axis corresponding to the first-year data axis in the division slot.

[0101] This solution will establish corresponding display slots for each year in combination with the first-year data axis. These display slots are used for visualizing and managing the display of annual data. Place each first-year data axis in its corresponding display slot to ensure the orderly display of the data. Such a structure helps in the intuitive analysis and comparison of the data.

[0102] It should be noted that this solution will also establish a division slot below the last display slot. This slot is used to accommodate and manage the division data. Establish a division data axis corresponding to the first-year data axis in the division slot. These division data axes are used for further data processing and analysis, especially for data segmentation and periodic analysis, which will be described in detail in subsequent processing.

[0103] S23. Based on the division data axis, divide all the first-year data axes to obtain the target period and retrieve the corresponding first sub-information to obtain the first prediction sample set.

[0104] This solution will divide all the first-year data axes based on the established division data axis. This step is to segment the data according to specific periods or conditions for subsequent analysis. Through division, different data periods can be identified, such as quarters, months, or other periodic data characteristics, so as to use specific data for accurate analysis. After the division, identify the target period. This period refers to the time range of interest in prediction or further analysis. Retrieve the first sub-information associated with the target period to form the first prediction sample set. This sample set is an important basis for subsequent prediction model training and verification.

[0105] Among them, dividing all the first-year data axes based on the divided data axis to obtain a target period and retrieving the corresponding first sub-information to obtain a first prediction sample set includes:

[0106] S231. Determine the target period based on the current time point. Each current time point has a preset target period, and extract the start time point and end time point of the target period.

[0107] This solution will determine the target period based on the current time point (such as the current date or a specific time marker). Each current time point has a preset target period, which determines the time range of data analysis. Extract the start time point (for example, January 1st) and end time point (for example, February 29th) from the target period. These two time points define the start and end boundaries of data analysis, ensuring data analysis and prediction within an appropriate time range.

[0108] S232. Determine the day time information and pixel points corresponding to the start time point and end time point respectively in the divided data axis to obtain the start pixel point and end pixel point.

[0109] On the divided data axis, find the specific day time information corresponding to the start time point and end time point. This information may include specific dates, times, and other time-related attributes. At the same time, determine the pixel points corresponding to the start time point and end time point on the data axis, that is, the start pixel point and end pixel point. These pixel points are used to accurately locate the position of the data on the time axis.

[0110] S233. Divide all the first-year data axes based on the start pixel point and end pixel point, and retrieve the first sub-information based on the storage format of the collected data to obtain the first prediction sample set.

[0111] Using the determined initial pixel point and cut-off pixel point, all the first-year data axes are divided. Through this division, the relevant data within the target period can be intercepted, facilitating subsequent analysis and processing. The division process should consider the integrity and continuity of the data to ensure that the obtained sub-datasets are valid and representative. Based on the storage format of the collected data, the first sub-information is retrieved from the divided data. These sub-information are important inputs for predictive analysis and may include various types of feature data and metrics. Finally, these first sub-information are summarized and organized to form the first prediction sample set. This sample set will be used for subsequent prediction model training and validation to help generate more accurate prediction results. Through this series of steps, the required data can be accurately located and extracted on the time axis, laying a solid foundation for constructing an effective data model and making accurate predictions. Using pixel points for data division is a refined data processing method that helps improve the accuracy and efficiency of data analysis.

[0112] Among them, the division of all the first-year data axes based on the initial pixel point and cut-off pixel point, and the retrieval of the first sub-information based on the storage format of the collected data to obtain the first prediction sample set, includes:

[0113] An initial division line perpendicular to the divided data axis is established upward starting from the initial pixel point, and a cut-off division line perpendicular to the divided data axis is established upward starting from the cut-off pixel point. In this solution, a line perpendicular to the divided data axis is established upward starting from the initial pixel point, which is called the initial division line. This division line is used to determine the starting boundary of the data axis and define the position from which data is extracted. Similarly, a line perpendicular to the divided data axis is established upward starting from the cut-off pixel point, which is called the cut-off division line. The cut-off division line is used to define the termination boundary of the data axis and identify the position where data extraction ends.

[0114] All the pixel points of each first-year data axis located within the initial division line and the cut-off division line are obtained, and the first sub-information of the corresponding storage unit in the storage format is retrieved according to the order of the pixel points to obtain the first prediction sample set. In this solution, for each first-year data axis, all the pixel points located between the initial division line and the cut-off division line are obtained. These pixel points represent the data range within the target period, ensuring the accuracy and relevance of data extraction. According to the order of the pixel points, the first sub-information of the corresponding storage unit in the storage format is retrieved in sequence. The storage format is the organization method of data and may include forms such as data tables, arrays, database records, etc. The information extracted according to the storage format needs to ensure the consistency and integrity of the data. The extracted first sub-information is summarized and organized to form a complete first prediction sample set. This prediction sample set contains all the data features and metrics required for model prediction and can support subsequent analysis and modeling work.

[0115] In some embodiments, obtaining all pixel points of each first-year data axis within the initial division line and the cut-off division line, and retrieving first sub-information of corresponding storage units in the storage format according to the order of the pixel points to obtain a first prediction sample set, includes:

[0116] If it is determined that the user selects and moves the initial division line and / or the cut-off division line based on the interaction terminal, the initial division line and / or the cut-off division line are synchronously moved in response to the user's movement. To meet the user's needs, this solution also provides an interaction mechanism. This interaction mechanism allows users to flexibly select and adjust the data range on the data visualization interface, thereby enhancing the accuracy and flexibility of data analysis. Through an intuitive graphical interface, users can better understand the distribution characteristics of the data and perform further analysis or prediction within the selected data range. Among them, the user can select and move the initial division line and / or the cut-off division line through the interaction terminal (such as a mouse or a touch screen). When the user moves these division lines, the positions of these lines are dynamically updated to respond to the user's operation. By the above method, the distance between the initial division line and the cut-off division line can be changed.

[0117] After determining the termination of the movement of the initial division line and / or the cut-off division line, count the pixel points of the initial division line and / or the cut-off division line. Once the user stops moving the division line, the program will count all the pixel points included in the division line. These pixel points are used to extract the corresponding first sub-information from the storage format, thereby forming a prediction sample set.

[0118] If it is determined that the user double-clicks at any position of the initial division line and / or the cut-off division line based on the interaction terminal, the initial division line and / or the cut-off division line are split at the double-click position to obtain multiple moving split lines. The user can perform split processing by double-clicking at any position of the division line. After double-clicking, the division line will be split into multiple moving split lines, enabling the user to move these newly generated line segments separately. Thus, the user can customize the selection of the required data segments.

[0119] If it is determined that the user selects and moves any moving split line based on the interaction terminal, the moving split line is synchronously moved in response to the user's movement, and the pixel points between the moving split line and the moving split line, between the moving split line and the cut-off division line, or between the initial division line and the moving split line are counted. It can be understood that the user can select any moving split line and move it. The program will synchronously update the positions of these line segments and count the pixel points between these line segments.

[0120] S3. If it is determined that the first prediction sample set does not meet the power generation prediction requirements, traverse the second power generation groups corresponding to the attributes of the first power generation group in the database, and perform data supplementation on the first prediction sample set based on the historical second synchronization data of the second power generation groups to obtain a second prediction sample set.

[0121] It is understandable that if the current predicted quantity of the first sub - information in the prediction sample set is less than the preset predicted threshold quantity, it is determined that the prediction sample set does not meet the power generation prediction requirements. The predicted threshold quantity is a predefined standard to ensure that the data volume in the sample set is sufficient for reliable prediction.

[0122] In some embodiments, if it is determined that the prediction sample set does not meet the power generation prediction requirements, then traverse the second power generation groups corresponding to the attributes of the first power generation group in the database, and perform data complementation on the first prediction sample set based on the historical second synchronization data of the second power generation groups, to obtain a second prediction sample set, including:

[0123] S31, if the current predicted quantity of the first sub - information in the prediction sample set is less than the predicted threshold quantity, then determine that the prediction sample set does not meet the power generation prediction requirements.

[0124] If the data volume of the data is insufficient and does not meet the requirements at this time, data complementation is required.

[0125] S32, calculate the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity, and traverse the second power generation groups corresponding to the attributes of the first power generation group in the database.

[0126] When performing data complementation, this solution first needs to calculate the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity. Then, search for the second power generation groups corresponding to the attributes of the first power generation group in the database. These second power generation groups have similar or related attributes to the first power generation group and may provide useful historical data. Use the historical second synchronization data of the second power generation groups to perform data complementation on the first prediction sample set.

[0127] Among them, the calculating the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity, and traversing the second power generation groups corresponding to the attributes of the first power generation group in the database, includes:

[0128] S321, obtain the attributes of the first power generation group, and the attributes at least include power generation type attribute, power generation specification attribute, power generation region attribute, and power generation environment attribute.

[0129] First of all, this solution needs to obtain the attributes of the first power generation group first. Power generation type attribute: For example, wind energy, solar energy, nuclear energy, etc. Power generation specification attribute: May include power generation capacity, equipment type, etc. Power generation region attribute: The geographical location where the power generation facility is located. Power generation environment attribute: For example, environmental factors such as climate conditions and altitude that affect power generation efficiency.

[0130] S322, based on the power generation type attribute, determine the screening of other power generation groups in the database, and determine other power generation groups with the same power generation type.

[0131] This solution will screen out other power generation groups with the same power generation type attribute in the database based on the power generation type attribute of the first power generation group. This step is to ensure that the selected power generation groups are consistent in the basic power generation method.

[0132] S323, calculate the fusion similarity values of other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes.

[0133] Finally, this solution will calculate the fusion similarity values of the screened power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes. The fusion similarity value is a comprehensive index, which can be obtained through weighted average or other similarity calculation methods, and is used to measure the comprehensive similarity degree of two power generation groups. Among the screened power generation groups, those with higher fusion similarity values with the first power generation group are preferentially selected for data filling. These power generation groups with higher similarity can provide more relevant historical data to improve the prediction sample set. In this way, appropriate power generation groups can be selected for data filling under specific conditions to ensure the quality and relevance of the filled data, and finally obtain a second prediction sample set that meets the power generation prediction requirements. This method makes full use of the existing data assets in the database to support more accurate power generation prediction.

[0134] In some embodiments, the calculation of the fusion similarity values of other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes includes:

[0135] Perform differential processing on the power generation specification attributes of the power generation group to obtain power generation specification coefficients of different dimensions. The power generation specification coefficients at least include rated power generation, area information, and / or volume specification information. Differentially process the power generation specifications of different dimensions to obtain corresponding power generation specification coefficients. These coefficients may include: Rated power generation: refers to the maximum output power of the power generation equipment under standard conditions. Area information: such as the power generation area of the power generation facility. Volume specification information: such as the volume or size of the equipment.

[0136] Perform quantization processing on the power generation region attributes of the power generation group to obtain region quantization coefficients. The region quantization coefficients at least include plain, hilly, mountainous, and urban. Quantize the region attributes to obtain region quantization coefficients. These coefficients can reflect geographical features, including but not limited to: Plain, hilly, mountainous, urban: These terrain features can affect climate conditions and equipment maintenance. Among them, when performing quantization, it can be combined with a preset quantization table for quantization, and different regions in the preset quantization table have pre-corresponding coefficients.

[0137] The power generation specification attributes of the power generation group are distinguished and processed to obtain power generation environment coefficients in different dimensions. The power generation environment coefficients at least include a temperature coefficient and a humidity coefficient. This solution will distinguish different dimensions of the power generation environment and calculate the power generation environment coefficients. These coefficients at least include: Temperature coefficient: reflecting the impact of temperature on power generation efficiency. Humidity coefficient: reflecting the impact of humidity on equipment operation.

[0138] Based on the power generation specification coefficient, the regional quantification coefficient, and the power generation environment coefficient, a fusion calculation is performed to obtain the fusion similarity value between the other power generation group and the first power generation group. Finally, this solution will perform a fusion calculation based on the coefficients in the above three aspects: the power generation specification coefficient, the regional quantification coefficient, and the power generation environment coefficient. Through a certain algorithm, the fusion similarity value between the other power generation group and the first power generation group is calculated. The fusion similarity value is a comprehensive index that reflects the similarity of different power generation groups in multi-dimensional attributes. Through the above method, other power generation groups with a high fusion similarity value to the first power generation group are selected to facilitate more accurate power generation prediction or data filling. This method ensures that the selected comparison power generation groups have a high degree of matching in key attributes, thereby improving the accuracy and reliability of the prediction model.

[0139] Among them, the fusion calculation based on the power generation specification coefficient, the regional quantification coefficient, and the power generation environment coefficient to obtain the fusion similarity value between the other power generation group and the first power generation group includes:

[0140] The fusion similarity value is calculated through the following formula

[0141] ;

[0142] where is the fusion similarity value between the c th power generation group and the first power generation group, is the power generation specification weight, is the power generation specification coefficient of the i th dimension, is the specification coefficient weight of the i th dimension, n is the upper limit value of the dimension of the power generation specification, is the regional specification weight, is the regional specification coefficient of the m th dimension, is the specification coefficient weight of the m th dimension, q is the upper limit value of the dimension of the regional specification, is the environment specification weight, is the environment specification coefficient of the r th dimension, is the specification coefficient weight of the r th dimension, eThe upper limit value for the dimension of the environmental specification.

[0143] It can be understood that in this formula, we calculate the fusion similarity value between different power generation groups and the first power generation group by combining the weights of power generation specifications, regional characteristics, and environmental factors. This fusion calculation method can comprehensively consider various influencing factors and help to more accurately predict power generation and compare between power generation groups. Among them, the power generation specification weight is used to adjust the influence of power generation specifications in the calculation of the overall fusion similarity value. The power generation specification coefficient of the i-th dimension represents the power generation capacity or characteristics in this dimension. The specification coefficient weight of the i-th dimension reflects the importance of this dimension in the overall specification. The upper limit value of the power generation specification dimension represents the number of dimensions covered by the power generation specification coefficient. The regional specification weight is used to adjust the influence of regional characteristics in the calculation of the overall similarity value. The regional specification coefficient of the m-th dimension represents the characteristics of this region in this dimension. The specification coefficient weight of the m-th dimension indicates the importance of the regional characteristics of this dimension. The upper limit value of the regional specification dimension represents how many regional characteristic dimensions are considered. The environmental specification weight is used to adjust the influence of environmental factors in the calculation of the overall similarity value. The environmental specification coefficient of the r-th dimension represents the environmental characteristics in this dimension. The specification coefficient weight of the r-th dimension reflects the importance of the environmental characteristics of this dimension. The upper limit value of the environmental specification dimension represents the number of dimensions covered by environmental factors. Through this method of weighted aggregation, it is possible to effectively combine the influences of different factors on the performance of power generation groups, thereby more comprehensively evaluating the similarity between different power generation groups. This is of great significance for optimizing power generation strategies, resource allocation, and improving the overall efficiency of the power system.

[0144] S324, determine the second power generation group among other power generation groups based on the fusion similarity value and the missing prediction quantity, and add a corresponding prediction offset label to the second power generation group based on each fusion similarity value.

[0145] This solution will combine the fusion similarity value and the missing prediction quantity, and select those power generation groups that meet the preset requirements as the second power generation group. For each eligible second power generation group, according to its fusion similarity value, add a prediction offset label. This label may be used to adjust the prediction model, taking into account the differences between different power generation groups, to improve the accuracy of prediction.

[0146] In some embodiments, the determining the second power generation group among other power generation groups based on the fusion similarity value and the missing prediction quantity, and adding a corresponding prediction offset label to the second power generation group based on each fusion similarity value includes:

[0147] Sort all other power generation groups based on the fusion similarity value to obtain a similarity value sequence. This solution will sort all other power generation groups based on the fused similarity value to obtain a similarity value sequence. This step is to identify which power generation group is most similar to the target power generation group in terms of data characteristics.

[0148] Successively obtain the second-year data axis of the first power generation group in the similarity value sequence and add it above the first-year data axis, and obtain the pixel points and second sub-information corresponding to the second-year data based on the initial division line, the cut-off division line, and the moving split line. This solution will, according to the similarity value sequence, start from the first power generation group, extract the data axis of the second year, and add it above the data axis of the first year. Then, this solution can extend or complete the initial division line, the cut-off division line, and the moving split line so that they span the data axis of the second year to obtain the pixel points and second sub-information corresponding to the second-year data.

[0149] If the number of the second sub-information is greater than or equal to the missing prediction quantity, it is determined that the currently obtained second sub-information meets the requirements, and the second power generation group corresponding to the currently obtained second sub-information is selected. The second sub-information belongs to the second synchronous data of the corresponding second power generation group. This solution will check whether the number of the extracted second sub-information is greater than or equal to the missing prediction quantity. If the requirements are met, the currently obtained second sub-information and its corresponding power generation group are selected and marked as the second power generation group, where the second sub-information belongs to the second synchronous data of this power generation group.

[0150] Determine the corresponding prediction offset label based on the fusion similarity value of the second power generation group. Each similarity value interval has a preset prediction offset label. This solution will determine the corresponding prediction offset label based on the fusion similarity value of the second power generation group. Each similarity value interval has a preset prediction offset label, and these labels are used to correct the prediction result to better adapt to the situation of a specific power generation group. It should be noted that the prediction offset label can be a prediction offset coefficient, and the smaller the fusion similarity value, the larger the corresponding prediction offset label needs to be.

[0151] On the basis of the above embodiments, it further includes:

[0152] If the number of the second sub-information is less than the missing prediction quantity, other power generation groups are selected again according to the similarity value sequence, and the second annual data axes of the other power generation groups are added above the previous second annual data axis. It can be understood that if, among the initially selected power generation groups, the number of the extracted second sub-information is less than the required missing prediction quantity, then it is necessary to further select the data of other power generation groups. According to the similarity value sequence, continue to select the unused power generation groups. Add the second annual data axes of these power generation groups above the previous second annual data axis. This process is similar to stacking the data of different power generation groups to accumulate sufficient data information.

[0153] Based on the initial division line, the cut-off division line, and the moving split line, obtain the pixel points and the second sub-information corresponding to the newly added second annual data until the number of the second sub-information is greater than or equal to the missing prediction quantity or there are no unselected power generation groups in the similarity value sequence. Use the initial division line, the cut-off division line, and the moving split line to process the newly added second annual data. Extract the pixel points and the second sub-information corresponding to the new data. Such operations may involve complex image processing or signal processing techniques to locate and extract in the multi-dimensional data space. This process continues until one of the following conditions is met: the number of the extracted second sub-information is greater than or equal to the missing prediction quantity; there are no unselected power generation groups in the similarity value sequence.

[0154] S33, generate the display slot and the second annual data axis corresponding to the second power generation group, and complete the data of the first prediction sample set based on the historical second synchronization data of the second power generation group to obtain the second prediction sample set.

[0155] This solution will generate a display slot and a second annual data axis corresponding to each second power generation group. Based on the historical second synchronization data of the second power generation group, complete the data of the first prediction sample set. Completing the data means filling the missing values or incomplete information in the prediction sample set through the historical data, thereby forming the second prediction sample set. This filling method helps to improve the comprehensiveness and accuracy of the prediction results, especially in the case of missing or incomplete data. After the above steps, the second prediction sample set can provide a more complete and accurate data basis for subsequent analysis and prediction. This method is particularly applicable to the data analysis of complex power generation systems and can achieve efficient data sharing and utilization among different power generation groups.

[0156] S4, perform power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain the new energy power generation prediction result.

[0157] Select to use the first prediction sample set or the second prediction sample set according to the specific situation. The selection basis may be the integrity, accuracy, and relevance of the data. The first prediction sample set may be based on the initial data, while the second prediction sample set is further processed with data filling and historical data and may be more complete.

[0158] In some embodiments, performing power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result includes:

[0159] S41. If it is determined that only the first prediction sample set exists, no prediction interval is generated, and the new energy power generation prediction result at the predicted target time is calculated based on the first prediction sample set through the following formula,

[0160] ;

[0161] ;

[0162] ;

[0163] Wherein, is the new energy power generation prediction result at the predicted target time calculated based on the first prediction sample set, is the average power generation result of all first sub-information in the first prediction sample set, is the weight value of the F dimension that affects the power generation result, is the value of the time predicted by the first power generation group in the F dimension, is the average value of the F dimension that affects the power generation result in the first prediction sample set, B is a constant value, G is the upper limit value of the dimension that affects the power generation result, is the power generation result of the th first sub-information in the first prediction sample set, is the upper limit value of the first prediction sample, is the quantity value of the first prediction sample, is the th first sub-information in the first prediction sample set in the F dimension, D is the upper limit value of the first sub-information in the first prediction sample set, X is the quantity value of the first sub-information in the first prediction sample set;

[0164] First, it can be understood that if there is only the first prediction sample set, it means that the data does not need to be adjusted, and the power generation prediction result at the predicted target time can be directly calculated through the formula. In the above formula, is the average power generation result of all the first sub-information in the first prediction sample set, represents the power generation results of all the first sub-information in the first prediction sample set, and the average power generation result is obtained after taking the average. Similarly, the sum of the values of the th first sub-information in the first prediction sample set in the F dimension can be averaged to obtain the average value of the F dimension that affects the power generation result in the first prediction sample set. Finally, the influence of different factors on the power generation result is evaluated by calculating the weighted deviation of each dimension. The result combines the historical average result of the sample and the weighted deviation under the current conditions, providing a comprehensive prediction result. The design of this formula aims to accurately estimate the new energy power generation according to the historical data and the current prediction conditions. By adjusting the weights and the constant B , the model can flexibly adapt to the prediction requirements in different scenarios and conditions.

[0165] S42, if it is determined that there is a second prediction sample set, then the second prediction sample set is segmented to obtain the first prediction sample subset of the first power generation group and the second prediction sample subset of the second power generation group.

[0166] If there is a second prediction sample set, this solution requires sample set segmentation, and then the second prediction sample set is segmented into the first prediction sample subset of the first power generation group and the second prediction sample subset of the second power generation group, and subsequent calculations are performed separately.

[0167] S43, calculate and process the first prediction sample subset according to the calculation method when there is only the first prediction sample set to obtain the first sub-power generation result, calculate the second sub-power generation result of the second prediction sample subset, and fuse the first sub-power generation result, the second sub-power generation result, and the prediction offset label to obtain the new energy power generation prediction result.

[0168] For the processing of the first predicted sample subset, this solution will be processed according to the calculation method of the first predicted sample set to obtain the first sub-generation result. For the processing of the second predicted sample subset, the second sub-generation result is calculated. Result fusion: The first sub-generation result and the second sub-generation result are combined with the predicted offset label for fusion to obtain the final new energy generation prediction result. This prediction method enhances the flexibility and accuracy of prediction through the reasonable division and calculation of different sample sets: By splitting and subset calculation of the second predicted sample set, the system can comprehensively predict using multiple types of data, not relying solely on a single data source, thereby improving the reliability of the prediction result. Result fusion mechanism: By combining the predicted offset label, the prediction result can be further adjusted and optimized, which is particularly important for the real-time changing new energy generation scenario.

[0169] Among them, the first predicted sample subset is calculated and processed according to the calculation method when only the first predicted sample set exists to obtain the first sub-generation result, the second sub-generation result of the second predicted sample subset is calculated, and the new energy generation prediction result is fused based on the first sub-generation result, the second sub-generation result, and the predicted offset label, including:

[0170] The second sub-generation result is calculated through the following formula

[0171] ;

[0172] ;

[0173] Among them, is the second sub-generation result, is the prediction result of the second electronic information based on the I th second power generation group, h is the upper limit value of the second power generation group, indicating the total number of second power generation groups, is the quantity value of the second power generation group, used to calculate the average value, is the I th average power generation value of the second power generation group, is the average value of the I th second power generation group in the T dimension, is the average value of the first power generation group in the T dimension, is the weight value of the T dimension, N is the upper limit value of the dimension when comparing the first power generation group and the second power generation group, M is the quantity value of the dimension when comparing the first power generation group and the second power generation group.

[0174] Calculate the mean of the first sub-generation result and the second sub-generation result to obtain the predicted mean. Determine the sum of the offset coefficients obtained by adding the offset coefficients corresponding to each predicted offset label. Each offset label has a preset offset coefficient.

[0175] This solution will obtain the predicted mean by calculating the mean of the first sub-generation result and the second sub-generation result. Determine the offset coefficient corresponding to each predicted offset label. After adding the coefficients of all offset labels, calculate the sum of the offset coefficients. Each offset label has a preset offset coefficient for adjusting the prediction result. It can be understood that the smaller the fusion similarity value, the larger the offset coefficient of the corresponding predicted offset label, and the greater the adjustment amplitude.

[0176] Perform upward and downward offset processing on the predicted mean based on the sum of the offset coefficients to obtain the predicted result interval, which is used as the new energy power generation prediction result.

[0177] Finally, based on the sum of the offset coefficients, perform upward and downward offset processing on the predicted mean. The obtained result interval is the final new energy power generation prediction result. This process first calculates the first sub-generation result and the second sub-generation result separately, and then combines these results and adjusts them through offset labels to finally obtain a more accurate power generation prediction. The calculation of the second sub-generation result quantifies the influence of each factor on the prediction by considering the differences between the second power generation group and the first power generation group in different dimensions and the weights of different dimensions. A more refined prediction method is provided, which enhances the prediction accuracy through grouping and dimension comparison.

[0178] The present invention also provides a new energy power generation prediction platform based on collected data, including:

[0179] A time module for the server to obtain the first power generation time point when the first power generation group starts initial power generation, and obtain the first synchronization data of the first power generation group's history based on the first power generation time point. The first synchronization data is the collected data of the first power generation group at historical moments;

[0180] A decomposition module for decomposing the data volume of the first synchronization data according to a period to obtain first sub-information under different periods, determining the target period based on the current time point, and determining the first sub-information corresponding to the target period to obtain the first prediction sample set;

[0181] A filling module for, if it is determined that the prediction sample set does not meet the power generation prediction requirements, traversing the second power generation group corresponding to the attributes of the first power generation group in the database, and filling the first prediction sample set with the second synchronization data of the second power generation group's history to obtain the second prediction sample set;

[0182] A prediction module, configured to perform power generation prediction on a first power generation group based on a first prediction sample set or a second prediction sample set, so as to obtain a new energy power generation prediction result.

[0183] The present invention also provides a storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the above various embodiments.

[0184] Among them, the storage medium may be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, the storage medium is coupled to the processor, so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may be located in an application specific integrated circuit (ASIC). In addition, the ASIC may be located in a user device. Of course, the processor and the storage medium may also exist as discrete components in a communication device. The storage medium may be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0185] The present invention also provides a program product, which includes execution instructions stored in a storage medium. At least one processor of the device can read the execution instructions from the storage medium, and the execution of the execution instructions by at least one processor causes the device to implement the methods provided by the above various embodiments.

[0186] In the above embodiments of the terminal or the server, it should be understood that the processor may be a central processing unit (CPU for short), and may also be other general-purpose processors, digital signal processors (DSP for short), application specific integrated circuits (ASIC for short), etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the present invention may be directly embodied as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A new energy power generation prediction method based on collected data, characterized in that: include: The server obtains a first power generation time point when the first power generation group performs initial power generation, and obtains first historical synchronization data of the first power generation group based on the first power generation time point, where the first synchronization data is the collected data of the first power generation group at a historical moment; Decomposing the data volume of the first synchronization data according to the period to obtain first sub-information under different periods, determining a target period based on the current time point, and determining the first sub-information corresponding to the target period to obtain a first prediction sample set; If it is determined that the prediction sample set does not meet the power generation prediction requirements, traverse the second power generation group corresponding to the attributes of the first power generation group in the database, and complete the first prediction sample set based on the historical second synchronization data of the second power generation group to obtain a second prediction sample set; Perform power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result; Based on the first prediction sample set or the second prediction sample set, power generation prediction is performed on the first power generation group to obtain a new energy power generation prediction result, including: If it is determined that only the first prediction sample set exists, the prediction interval is not generated, and the new energy power generation prediction result at the prediction target time is calculated based on the first prediction sample set; If it is determined that the second prediction sample set exists, the second prediction sample set is divided to obtain a first prediction sample subset of the first power generation group and a second prediction sample subset of the second power generation group; The first prediction sample subset is calculated and processed according to the calculation method of only the first prediction sample set to obtain a first sub-generation result, and a second sub-generation result of the second prediction sample subset is calculated. The new energy generation prediction result is obtained based on the first sub-generation result, the second sub-generation result, and the prediction offset label fusion, including: Calculate the average of the first sub-generation result and the second sub-generation result to obtain a predicted mean, determine the offset coefficients corresponding to each predicted offset label, add them together and weight them to obtain the sum of the offset coefficients, and each offset label has a preset offset coefficient; Based on the sum of the offset coefficients, the predicted mean is offset up and down to obtain the prediction result interval as the new energy power generation prediction result.

2. The new energy power generation prediction method based on collected data according to claim 1 is characterized in that: The step of decomposing the data volume of the first synchronization data according to the period to obtain first sub-information in different periods, determining a target period based on a current time point, and determining the first sub-information corresponding to the target period to obtain a first prediction sample set includes: Extracting the earliest time of the first synchronization data, determining the annual time period between the earliest time and the current time based on the earliest time, and establishing a first annual data axis corresponding to each annual time period; Establish corresponding display slots based on the first-year data axis, place each first-year data axis in the display slot, establish a division slot at the bottom of the last display slot, and establish a division data axis corresponding to the first-year data axis in the division slot; All first-year data axes are divided based on the divided data axes to obtain a target period and retrieve the corresponding first sub-information to obtain a first prediction sample set.

3. The new energy power generation prediction method based on collected data according to claim 2 is characterized in that: The extracting the earliest time of the first synchronization data, determining the annual time period between the earliest time and the current time based on the earliest time, and establishing the first annual data axis corresponding to each annual time period, includes: Determine a first number of pixel points corresponding to the first annual data axis; Determine the number of days corresponding to each first-year data axis, normalize the number of days of all first-year data axes to obtain a unified number of days, obtain the day time information corresponding to each group of pixel points based on the first number and the unified number of days, and establish a storage format for the collected data based on the day time information, where each day time information corresponds to a storage unit; According to the order of each group of pixels, corresponding time of day information is added to each group of pixels, and the information in the collected data is stored in correspondence with the corresponding group of pixels according to the time of day information, so that each group of pixels has a corresponding storage unit.

4. The new energy power generation prediction method based on collected data according to claim 3 is characterized in that: The step of dividing all first-year data axes based on the divided data axes to obtain a target period and adjusting the corresponding first sub-information to obtain a first prediction sample set includes: Determine a target period based on the current time point, each current time point has a preset target period, and extract an initial time point and a deadline time point of the target period; Determine the time information and pixel points corresponding to the initial time point and the end time point in the divided data axis, and obtain the initial pixel point and the end pixel point; All first-year data axes are divided based on the initial pixel point and the cutoff pixel point, and the first sub-information is retrieved based on the storage format of the collected data to obtain a first prediction sample set.

5. The new energy power generation prediction method based on collected data according to claim 4 is characterized in that: The step of dividing all first-year data axes based on the initial pixel point and the cutoff pixel point, and retrieving the first sub-information based on the storage format of the collected data to obtain the first prediction sample set includes: An initial dividing line perpendicular to the dividing data axis is established upward based on the initial pixel point as a starting point, and a cutoff dividing line perpendicular to the dividing data axis is established upward based on the cutoff pixel point as a starting point; All pixel points of each first-year data axis located within the initial dividing line and the cutoff dividing line are obtained, and the first sub-information of the corresponding storage unit in the storage format is retrieved according to the order of the pixel points to obtain the first prediction sample set.

6. The new energy power generation prediction method based on collected data according to claim 5 is characterized in that: The step of obtaining all pixel points of each first-year data axis located within the initial dividing line and the cutoff dividing line, and retrieving the first sub-information of the corresponding storage unit in the storage format according to the order of the pixel points to obtain the first prediction sample set includes: If it is determined that the user selects the initial dividing line and / or the cut-off dividing line based on the interactive terminal and moves, the initial dividing line and / or the cut-off dividing line are synchronously moved in response to the user's movement; After determining that the movement of the initial dividing line and / or the cut-off dividing line has terminated, counting the pixel points of the initial dividing line and / or the cut-off dividing line; If it is determined that the user double-clicks at any position of the initial dividing line and / or the cut-off dividing line selected by the interactive terminal, the initial dividing line and / or the cut-off dividing line is segmented at the double-clicked position to obtain multiple moving segmenting lines; If it is determined that the user selects and moves any moving dividing line based on the interactive terminal, the moving dividing line is moved synchronously in response to the user's movement, and the pixel points between the moving dividing line and the moving dividing line, the moving dividing line and the cutoff dividing line, or the initial dividing line and the moving dividing line are counted.

7. The method for predicting new energy power generation based on collected data according to claim 4, characterized in that: If it is determined that the prediction sample set does not meet the power generation prediction requirements, traverse the second power generation group in the database corresponding to the attributes of the first power generation group, and complete the first prediction sample set based on the historical second synchronization data of the second power generation group to obtain the second prediction sample set, including: If the current prediction quantity of the first sub-information in the prediction sample set is less than the prediction threshold quantity, it is determined that the prediction sample set does not meet the power generation prediction requirement; Calculate the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity, and traverse the second power generation group in the database corresponding to the attribute of the first power generation group; A display slot and a second annual data axis corresponding to the second power generation group are generated, and the first prediction sample set is supplemented based on the second historical synchronous data of the second power generation group to obtain a second prediction sample set.

8. The method for predicting new energy power generation based on collected data according to claim 7, characterized in that: The step of calculating the difference between the current predicted quantity and the predicted threshold quantity to obtain the missing predicted quantity and traversing the second power generation group in the database corresponding to the attribute of the first power generation group includes: Acquire attributes of the first power generation group, wherein the attributes include at least a power generation type attribute, a power generation specification attribute, a power generation region attribute, and a power generation environment attribute; Based on the power generation type attribute, determine to filter other power generation groups in the database to determine other power generation groups with the same power generation type; Calculate the fusion similarity values ​​of the other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes; A second power generation group is determined from other power generation groups based on the fused similarity values ​​and the number of missing predictions, and a corresponding prediction offset label is added to the second power generation group based on each fused similarity value.

9. The method for predicting new energy power generation based on collected data according to claim 8, characterized in that: The calculating of the fusion similarity values ​​of the other power generation groups and the first power generation group in terms of power generation specification attributes, power generation region attributes, and power generation environment attributes includes: Differentiating and processing the power generation specification attributes of the power generation group to obtain power generation specification coefficients of different dimensions, wherein the power generation specification coefficients at least include rated power generation, area information and / or volume specification information; Quantifying the power generation area attributes of the power generation group to obtain a regional quantification coefficient, wherein the regional quantification coefficient at least includes plains, hills, mountains, and cities; Differentiating and processing the power generation specification attributes of the power generation group to obtain power generation environment coefficients of different dimensions, wherein the power generation environment coefficients at least include a temperature coefficient and a humidity coefficient; Based on the power generation specification coefficient, the regional quantification coefficient and the power generation environment coefficient, a fusion calculation is performed to obtain the fusion similarity value of other power generation groups and the first power generation group.

10. The method for predicting new energy power generation based on collected data according to claim 9, characterized in that: The fusion calculation based on the power generation specification coefficient, the regional quantification coefficient and the power generation environment coefficient is performed to obtain the fusion similarity value of other power generation groups and the first power generation group, including: The fusion similarity value is calculated by the following formula: ； in, For the c The fusion similarity value of the first power generation group and the first power generation group, is the power generation specification weight, For the i The power generation specification coefficient of each dimension, For the i The specification coefficient weights of the dimensions, n is the upper limit of the dimension of power generation specification, is the regional specification weight, For the m The regional specification coefficient of the dimension, For the m The specification coefficient weights of the dimensions, q is the upper limit of the dimension of the region specification. is the environmental specification weight, For the r Environmental specification coefficients for each dimension, For the r The specification coefficient weights of the dimensions, e The upper limit value of the dimension for the environment specification.

11. The method for predicting new energy power generation based on collected data according to claim 9, characterized in that: The method of determining the second power generation group from other power generation groups based on the fused similarity value and the missing prediction quantity, and adding a corresponding prediction offset label to the second power generation group based on each fused similarity value, includes: Sort all other power generation groups based on the fused similarity value to obtain a similarity value sequence; Sequentially obtain the second year data axis of the first power generation group in the similar value sequence and add it above the first year data axis, and obtain the pixel points and second sub-information corresponding to the second year data based on the initial dividing line, the cutoff dividing line, and the moving dividing line; If the number of the second sub-information is greater than or equal to the missing prediction number, it is determined that the currently acquired second sub-information meets the requirement, and the second power generation group corresponding to the currently acquired second sub-information is selected, and the second sub-information belongs to the second synchronization data of the corresponding second power generation group; A corresponding prediction offset label is determined based on the fused similarity value of the second power generation group, and each similarity value interval has a preset prediction offset label.

12. The method for predicting new energy power generation based on collected data according to claim 11, characterized in that: Also includes: If the amount of the second sub-information is less than the missing prediction amount, selecting another power generation group again according to the similarity value sequence, and adding the second year data axis of the other power generation group to the top of the previous second year data axis; Based on the initial dividing line, the cutoff dividing line and the moving dividing line, the pixel points and the second sub-information corresponding to the newly added second year data are obtained until the number of second sub-information is greater than or equal to the missing prediction number or there are no more unselected power generation groups in the similar value sequence.

13. The method for predicting new energy power generation based on collected data according to claim 11, characterized in that: The step of performing power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result includes: If it is determined that only the first prediction sample set exists, the prediction interval is not generated, and the prediction result of new energy power generation at the prediction target time is calculated based on the first prediction sample set by the following formula: ； ； ； in, To calculate the new energy power generation prediction result at the prediction target time based on the first prediction sample set, is the average power generation result of all first sub-information in the first prediction sample set, The first prediction sample set affects the power generation results. F The weight value of the dimension, The predicted time for the first generating unit is F The value of the dimension, The first prediction sample set affects the power generation results. F The average value of the dimension, B is a constant value, G is the upper limit of the dimension that affects the power generation results, The first prediction sample set The generation result of the first sub-information, is the upper limit value of the first prediction sample, is the number of the first prediction sample, The first prediction sample set The first sub-information is in F The value of the dimension, D is the upper limit value of the first sub-information in the first prediction sample set, X is the quantity value of the first sub-information in the first prediction sample set.

14. The method for predicting new energy power generation based on collected data according to claim 13, characterized in that: The first prediction sample subset is calculated and processed according to the calculation method of only the first prediction sample set to obtain a first sub-generation result, a second sub-generation result of the second prediction sample subset is calculated, and a new energy generation prediction result is obtained based on the first sub-generation result, the second sub-generation result, and the prediction offset label fusion, including: The second sub-generation result is calculated by the following formula: ； ； in, The result of generating electricity for the second child is, Based on I The predicted result of the second electronic information of the second power generation group, h is the upper limit value of the second generating set, is the quantity value of the second power generation group, For the I The average power generation value of the second power generation group, For the I The average value of the second generation group in the Tth dimension, is the average value of the first power generation group in the Tth dimension, is the weight value of the Tth dimension, N is the upper limit value of the dimension when comparing the first power generation group and the second power generation group, M It is the quantitative value of the dimension when comparing the first power generation group and the second power generation group.

15. The new energy power generation prediction platform based on collected data according to any one of claims 1 to 14, characterized in that: include: A time module, used for the server to obtain a first power generation time point when the first power generation group performs initial power generation, and to obtain first historical synchronization data of the first power generation group based on the first power generation time point, wherein the first synchronization data is the collected data of the first power generation group at a historical moment; A decomposition module, configured to decompose the data volume of the first synchronization data according to the period to obtain first sub-information under different periods, determine a target period based on a current time point, and determine the first sub-information corresponding to the target period to obtain a first prediction sample set; A completion module, for traversing a second power generation group in the database corresponding to the attribute of the first power generation group, and completing the first prediction sample set based on the historical second synchronization data of the second power generation group to obtain a second prediction sample set if it is determined that the prediction sample set does not meet the power generation prediction requirement; The prediction module is used to perform power generation prediction on the first power generation group based on the first prediction sample set or the second prediction sample set to obtain a new energy power generation prediction result.

Images