Method, device, equipment and medium for predicting power generation

Abstract

The application discloses a power generation power prediction method, device, equipment and medium, which is applied to the field of data processing. The method comprises the following steps: acquiring target meteorological prediction data corresponding to a power generation device, wherein the target meteorological prediction data is provided by a target meteorological source set; generating a power influence feature of the target meteorological prediction data according to at least one of the target meteorological prediction data and associated meteorological data corresponding to the target meteorological prediction data, wherein the associated meteorological data comprises complementary information which is not possessed by the target meteorological prediction data and is used to improve the power prediction accuracy, and the power influence feature is a feature for enhancing and / or correcting the target meteorological prediction data; and performing power prediction by using the power influence feature and the target meteorological prediction data to obtain the predicted power of the power generation device. By extracting and strengthening useful information, eliminating redundant and error information and adding complementary information, the accuracy of the predicted power is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, and in particular to a power generation power prediction method, device, equipment and medium. BACKGROUND

[0002] NWP (Numerical Weather Prediction) is one of the important data sources for power generation power prediction of power generation equipment. Since NWP data itself has errors, it will be transmitted to the power generation equipment power prediction error caused by NWP error.

[0003] The related technology obtains meteorological prediction data provided by multiple meteorological prediction sources, eliminates abnormal data in the meteorological prediction data, and then fuses the meteorological prediction data provided by the multiple meteorological sources to obtain fusion data, which is input into a power prediction model to obtain the predicted power of the power generation equipment.

[0004] The related technology does not fully utilize the historical comparison of meteorological prediction data itself, the advantages and disadvantages of different meteorological predictions, and useful information such as meteorological prediction data of adjacent geographical positions, and the accuracy of the predicted power is low. SUMMARY

[0005] The embodiments of the present application provide a power generation power prediction method, device, equipment and medium, which can generate power influence features according to the associated meteorological data corresponding to the target meteorological prediction data, and predict power using the power influence features and the target meteorological prediction data, so that the predicted power is more accurate. The technical solution is as follows:

[0006] According to one aspect of the present application, a power generation power prediction method is provided, which comprises:

[0007] Obtaining target meteorological prediction data corresponding to a power generation equipment, the target meteorological prediction data being provided by a target meteorological source set;

[0008] Generating power influence features of the target meteorological prediction data according to at least one of the target meteorological prediction data and associated meteorological data corresponding to the target meteorological prediction data, the associated meteorological data including complementary information that the target meteorological prediction data does not have and is used to improve the power prediction accuracy of the target meteorological prediction data, and the power influence features being features that enhance and / or correct the target meteorological prediction data;

[0009] Using the power influence features and the target meteorological prediction data to perform power prediction to obtain the predicted power of the power generation equipment.

[0010] According to one aspect of the present application, a power generation power prediction device is provided, which comprises:

[0011] an obtaining module, configured to obtain target meteorological prediction data corresponding to the power generation device, the target meteorological prediction data being provided by a target meteorological source set;

[0012] a processing module, configured to generate a power impact feature of the target meteorological prediction data according to associated meteorological data corresponding to the target meteorological prediction data, the associated meteorological data including complementary information that is not possessed by the target meteorological prediction data and is used to improve power prediction accuracy of the target meteorological prediction data, the power impact feature being a feature for enhancing and / or correcting the target meteorological prediction data;

[0013] the processing module is further configured to perform power prediction using the power impact feature and the target meteorological prediction data to obtain predicted power of the power generation device.

[0014] According to another aspect of the present application, a computer device is provided, which includes a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the power generation power prediction method according to the above aspect.

[0015] According to another aspect of the present application, a computer storage medium is provided, the computer readable storage medium storing at least one program code, the program code being loaded and executed by a processor to implement the power generation power prediction method according to the above aspect.

[0016] According to another aspect of the present application, a computer program product or a computer program is provided, the computer program product or the computer program including computer instructions, the computer instructions being stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device performs the power generation power prediction method according to the above aspect.

[0017] The technical scheme provided by the embodiments of the present application has at least the following beneficial effects:

[0018] According to the associated meteorological data corresponding to the target meteorological prediction data, the power impact feature of the target meteorological prediction data is generated, and the power impact feature and the target meteorological prediction data are used to predict the predicted power of the power generation device. Since the power impact feature is introduced into the process of power prediction, and the power impact feature is a feature for enhancing and / or correcting the target meteorological prediction data, the power impact feature not only provides additional information for the target meteorological prediction data, but also makes the target meteorological prediction data more accurate, and thus makes the predicted power more accurate. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0020] Figure 1 is a schematic diagram of a computer system provided by an exemplary embodiment of the present application;

[0021] Figure 2 is a flowchart of a method for predicting power generation provided by an exemplary embodiment of the present application;

[0022] Figure 3 is a flowchart of a method for predicting power generation provided by an exemplary embodiment of the present application;

[0023] Figure 4 is a schematic diagram of a target grid point and a neighboring grid point provided by an exemplary embodiment of the present application;

[0024] Figure 5 is a schematic diagram of a target grid point and a neighboring grid point provided by an exemplary embodiment of the present application;

[0025] Figure 6 is a flowchart of a method for obtaining a target meteorological source set provided by an exemplary embodiment of the present application;

[0026] Figure 7 is a schematic diagram of a method for obtaining a target meteorological source set provided by an exemplary embodiment of the present application;

[0027] Figure 8 is a flowchart of a method for generating a power generation strategy provided by an exemplary embodiment of the present application;

[0028] Figure 9 is a structural schematic diagram of a power generation prediction device provided by an exemplary embodiment of the present application;

[0029] Figure 10 is a structural block diagram of a computer device provided by another exemplary embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the present application more clear, the embodiments of the present application will be further described in detail with reference to the drawings.

[0031] First, the terms involved in the embodiments of the present application will be introduced:

[0032] Artificial Intelligence (AI): is to use digital computer or digital computer controlled machine simulation, extension and expansion of human intelligence, perception of environment, acquisition of knowledge and use of knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which tries to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machine has the functions of perception, reasoning and decision-making.

[0033] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software level technology. Artificial intelligence basic technology generally includes such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics and other technologies. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning and other several major directions.

[0034] NWP (Numerical Weather Prediction): refers to the method of predicting the state of atmospheric motion and weather phenomena in a certain period of time according to the actual situation of the atmosphere under certain initial value and boundary value conditions, solving the equation set describing the evolution process of weather by numerical calculation of fluid mechanics and thermodynamics.

[0035] Wind power prediction: refers to the prediction of the power of wind turbines in a wind farm. Wind power prediction technology predicts the power of a wind farm or wind turbine through algorithms based on meteorological information and other relevant data of the wind farm.

[0036] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.) and signals involved in the present application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data need to comply with relevant national and regional laws, regulations and standards. For example, the meteorological prediction data involved in the present application is obtained under full authorization.

[0037] Figure 1 The structure schematic diagram of the computer system provided by an example embodiment of the present application is shown. The computer system 100 includes a terminal 120, a server 140 and a power generation device 160.

[0038] A power prediction related application program is installed on the terminal 120. The application program can be an applet in an app (application), a dedicated application program, or a web client. The terminal 120 is at least one of a smartphone, a tablet computer, an e-book reader, an MP3 player, an MP4 player, a laptop computer, and a desktop computer.

[0039] The terminal 120 is connected to the server 140 through a wireless network or a wired network.

[0040] The server 140 can be a standalone physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and basic cloud computing services such as big data and artificial intelligence platforms. The server 140 is used to provide background services for the power prediction application program and send power prediction related information to the terminal 120. Alternatively, the server 140 undertakes the main computing work, and the terminal 120 undertakes the secondary computing work; or the server 140 undertakes the secondary computing work, and the terminal 120 undertakes the main computing work; or the server 140 and the terminal 120 both use a distributed computing architecture for collaborative computing.

[0041] The power generation equipment 160 refers to equipment that generates power through natural weather. The server 140 predicts the power of the power generation equipment 160. Exemplarily, the power generation equipment includes at least one of a wind power generation equipment, a solar power generation equipment, and a tidal power generation equipment.

[0042] Figure 2 A flowchart of a power prediction method provided by an exemplary embodiment of the present application is shown. The method can be executed by the computer system 100 shown, and the method includes the following steps. Figure 1 The method includes the following steps.

[0043] Step 202: Obtain target weather prediction data corresponding to the power generation equipment, the target weather prediction data being provided by a target weather source set, and the target weather prediction data being used to predict the power of the power generation equipment.

[0044] Alternatively, the power generation equipment refers to equipment that generates power through natural weather. Exemplarily, the power generation equipment includes at least one of a wind power generation equipment, a solar power generation equipment, and a tidal power generation equipment. Alternatively, the power generation equipment refers to one equipment, or the power generation equipment refers to at least two equipments.

[0045] For example, if the power generation device refers to a wind power generation device, the power generation device refers to devices in the same wind farm area, and the target meteorological prediction data refers to data of the wind farm.

[0046] Optionally, the target meteorological prediction data is meteorological prediction data of the power generation device. The meteorological prediction data is meteorological data at a future time relative to a current time, and the current time refers to a time when the power prediction is started.

[0047] For example, the target meteorological prediction data is meteorological data in a future period relative to a current time. For example, the current time is January 1st, and the target meteorological prediction data is meteorological data from January 2nd to January 8th.

[0048] For example, the target meteorological prediction data is meteorological data in a future period relative to a current time. For example, the current time is January 1st, and the target meteorological prediction data is meteorological data from January 2nd to January 8th.

[0049] For example, the target meteorological prediction data is meteorological data at n future time points relative to a current time, and n is a positive integer. For example, the current time is January 1st, and the target meteorological prediction data is meteorological data at 12:00 on January 2nd, 19:00 on January 2nd, and 14:00 on January 3rd.

[0050] For example, the target meteorological prediction data is meteorological data in a future period relative to a current time. For example, the current time is January 1st, and the target meteorological prediction data is meteorological data from January 2nd to January 8th.

[0051] Optionally, the target meteorological prediction data of the area where the power generation device is located is obtained from the meteorological sources in the target meteorological source set. For example, the computer device sends a first data acquisition request to the server corresponding to the meteorological source in the target meteorological source set, and the first data acquisition request includes at least one of the longitude and latitude of the power generation device; after the server receives the first data acquisition request, the target meteorological prediction data is returned to the computer device.

[0052] Optionally, the target meteorological prediction data of the target time period corresponding to the power generation device is obtained from a meteorological source in the target meteorological source set. Illustratively, the computer device sends a second data acquisition request to a server corresponding to the meteorological source in the target meteorological source set, the second data acquisition request including at least one of the longitude and the latitude of the power generation device, and the second data acquisition request further including at least one of the time span, the start and end time, the start time, and the end time of the target time period; after the server receives the second data acquisition request, the target meteorological prediction data is returned to the computer device.

[0053] In a specific embodiment, the target meteorological prediction data is meteorological prediction data of a location of the power generation device. Illustratively, the power generation device is located at point A, and the meteorological prediction data of point A is the target meteorological prediction data. In another specific embodiment, the target meteorological prediction data is an average of meteorological prediction data of a region where the power generation device is located. Illustratively, the power generation device is located in region B, which includes point C and point D, and the average of the meteorological prediction data of point C and point D is the target meteorological prediction data.

[0054] Optionally, the target meteorological prediction data belongs to a time series. Illustratively, the target meteorological prediction data is (60, 80, 70, 55, 45), and the data in the target meteorological prediction data is arranged in chronological order from early to late.

[0055] Optionally, in the case of a wind power generation device, the target meteorological prediction data includes at least one of wind speed, wind direction, temperature, pressure, and density. Optionally, in the case of a solar power generation device, the target meteorological prediction data includes at least one of light intensity, light duration, light incidence angle, and temperature. Optionally, in the case of a tidal power generation device, the target meteorological prediction data includes at least one of flow rate, temperature, pressure, and density. Illustratively, taking the wind power generation device as an example, if the power generation of the wind power generation device on May 2 is to be predicted, the predicted wind speed and wind direction of the region where the wind power generation device is located on May 2 are determined as the target meteorological prediction data.

[0056] The target meteorological source set includes at least one meteorological source, and the meteorological source is used to provide meteorological data, which includes at least one of future meteorological data, historical meteorological data, and current meteorological data of the power generation device, the future meteorological data is meteorological data of a future time relative to a current time, the historical meteorological data is meteorological data of a past time relative to the current time, and the current meteorological data is meteorological data of the current time. Illustratively, Table 1 shows an optional meteorological source:

[0057] Table 1 Optional meteorological source

[0058]

[0059] Step 204: generating a power impact feature of the target meteorological prediction data according to associated meteorological data corresponding to the target meteorological prediction data, the associated meteorological data including complementary information that is not included in the meteorological prediction data and is used to improve the power prediction accuracy of the target meteorological prediction data, the power impact feature being a feature that enhances and / or corrects the target meteorological prediction data.

[0060] Optionally, the associated meteorological data is provided by the target meteorological source described above.

[0061] Optionally, the power impact feature is a feature obtained after useful information of the target meteorological prediction data is extracted and strengthened, redundant information and error information are removed, and complementary information is added.

[0062] Illustratively, the power impact feature includes at least one of an error prediction feature, a spatial prediction feature, and a time prediction feature.

[0063] The error prediction feature is used to correct statistical errors of the target meteorological prediction data. The error prediction feature is related to at least one of a target meteorological source, a time, a location, and a meteorological type of the target meteorological prediction data. Illustratively, data provided by the target meteorological source A has a higher accuracy in windy weather, and data provided by the target meteorological source B has a lower accuracy but a higher recall. Illustratively, meteorological prediction data of the region C has a higher accuracy in winter.

[0064] The spatial prediction feature is used to correct spatial errors of the target meteorological prediction data. Meteorological prediction data has continuity in space, and meteorological prediction data of adjacent regions will affect each other. For example, taking wind speed data as an example, at a preset time, the wind speed of region A is 10 m / s, and the wind speed of region B is 5 m / s, and the direction of the wind is from region B to region A, so the wind speed of region B will affect the wind speed of region A at the next time.

[0065] The time prediction feature is used to correct time errors of the target meteorological prediction data. Meteorological prediction data has continuity in time. For example, taking temperature data as an example, at the current time, the temperature data is 20 degrees, at the next time, the temperature data is 22 degrees, and at the previous time, the temperature data is 18 degrees, and the temperature data of adjacent times is within a reasonable temperature change range.

[0066] Optionally, the associated meteorological data corresponds to the power impact feature. Optionally, in the case where the power impact feature includes the error prediction feature, the associated meteorological data includes historical meteorological prediction data and historical meteorological actual data of a target meteorological source set.

[0067] Optionally, in the case that the power influence feature comprises a spatial prediction feature, the associated meteorological data comprises adjacent meteorological prediction data. The adjacent meteorological prediction data refers to meteorological prediction data of a region adjacent to the region where the power generation device is located. For example, if the region where the power generation device is located is a rectangle, the region adjacent to the region where the power generation device is located is a rectangle adjacent to the aforementioned rectangle. For example, if the region where the power generation device is located is a circle with a radius of 500 m, the region adjacent to the region where the power generation device is located is an annular region formed by a circle with a radius of 500 m and a circle with a radius of 800 m.

[0068] Optionally, in the case that the power influence feature comprises a time prediction feature, the associated meteorological data comprises a set of time meteorological prediction data corresponding to the target meteorological prediction data, the set of time meteorological prediction data comprising at least one of future meteorological data of the target meteorological prediction data and historical meteorological data of the target meteorological prediction data, the future meteorological data of the target meteorological prediction data being meteorological data relative to a future time of the target meteorological prediction data, and the historical meteorological data of the target meteorological prediction data being meteorological data relative to a past time of the target meteorological prediction data.

[0069] For example, the historical meteorological data of the target meteorological prediction data is meteorological data within a historical time period of the target meteorological prediction data. For example, the target meteorological prediction data is data at 14:00, and the historical meteorological data of the target meteorological prediction data is prediction data from 11:00 to 13:00. For example, the historical meteorological data of the target meteorological prediction data is meteorological data within t time units in the past of the target meteorological prediction data, t being a positive integer, and the time unit comprising at least one of an hour, a day, a week, a month, a quarter, and a year. For example, the target meteorological prediction data is data on February 1, and the historical meteorological data of the target meteorological prediction data is data of the past 3 days on February 1.

[0070] Step 206: performing power prediction using the power influence feature and the target meteorological prediction data to obtain predicted power of the power generation device.

[0071] Optionally, the target meteorological prediction data and the power influence feature are jointly input into a power prediction model; the power prediction model is called to perform data processing on the target meteorological prediction data and the power influence feature to obtain the predicted power. The power prediction model comprises, but is not limited to, at least one of a ResNet (Residual Network) model, a CNN (Convolutional Neural Networks) model, an RNN (Recurrent Neural Network) model, and a Transformer model. The type of the power prediction model is not limited in the present application.

[0072] The training of the power prediction model comprises the following steps: obtaining sample meteorological prediction data corresponding to a sample power generation device, the sample meteorological prediction data being provided by a sample meteorological source set; obtaining actual power corresponding to the sample meteorological prediction data; generating a sample power influence feature of the sample meteorological prediction data according to at least one of the sample meteorological prediction data and sample associated meteorological data corresponding to the sample meteorological prediction data; inputting the target meteorological prediction data and the power influence feature into the power prediction model to obtain a sample predicted power of the sample power generation device; and training the power prediction model according to a difference between the sample predicted power and the actual power.

[0073] In the embodiments of the present application, the process of obtaining the power influence feature can be implemented by the power prediction model. Alternatively, the target meteorological prediction data is input into the power prediction model; a data correction layer in the power prediction model is called to perform data processing on the target meteorological prediction data by using the power influence feature to obtain the power influence feature; and a power prediction layer in the power prediction model is called to perform data processing on the target meteorological prediction data and the power influence feature to obtain the predicted power.

[0074] In summary, the power influence feature of the target meteorological prediction data is generated according to the associated meteorological data corresponding to the target meteorological prediction data, and the power influence feature and the target meteorological prediction data are used to predict the predicted power of the power generation device. Since the power influence feature is introduced into the process of power prediction, and the power influence feature is a feature that enhances and / or corrects the target meteorological prediction data, the power influence feature not only provides additional information for the target meteorological prediction data, but also makes the target meteorological prediction data more accurate, and thus makes the predicted power more accurate.

[0075] In the following embodiments, the power influence feature of the target meteorological prediction data is required to predict the power of the power generation device, and the power influence feature comprises at least one of an error prediction feature, a spatial prediction feature and a time prediction feature. At this time, the power influence feature can be used to improve the accuracy of the predicted power of the power generation device.

[0076] Figure 3 A flowchart of a method for predicting power generation is shown, which is provided by an example embodiment of the present application. The method can be executed by the computer system 100 shown, and the method comprises the following steps: Figure 1 The computer system 100 shown executes the method, and the method comprises the following steps:

[0077] Step 301: obtaining target meteorological prediction data corresponding to a power generation device.

[0078] Alternatively, the power generation device refers to a device that generates power through natural meteorological conditions. Alternatively, the power generation device refers to one device, or the power generation device refers to at least two devices. For example, the power generation device refers to devices in the same wind field area.

[0079] In one specific implementation, the target meteorological prediction data is meteorological prediction data of a location of the power generation device. In another specific implementation, the target meteorological prediction data is an average of meteorological prediction data of a region where the power generation device is located.

[0080] Optionally, in the case that the power generation device is a wind power generation device, the target meteorological prediction data includes at least one of wind speed, wind direction, temperature, pressure, and density. Optionally, in the case that the power generation device is a solar power generation device, the target meteorological prediction data includes at least one of light intensity, light duration, light incident angle, and temperature. Optionally, in the case that the power generation device is a tidal power generation device, the target meteorological prediction data includes at least one of flow speed, temperature, pressure, and density.

[0081] Step 302: Obtain historical meteorological prediction data provided by the target meteorological source set.

[0082] Optionally, the historical meteorological prediction data is historical data provided by the target meteorological source set.

[0083] Optionally, the historical meteorological prediction data is meteorological data within p time units in the past with respect to the target meteorological source set, p is a positive integer, and the time unit includes at least one of minute, hour, day, week, month, quarter, and year. For example, the target meteorological prediction data is data on February 20, and the historical meteorological prediction data is data on the past 3 days of February 20.

[0084] Optionally, the historical meteorological prediction data is meteorological data within a historical period with respect to the target meteorological source set. For example, the target meteorological prediction data is data on February 20, and the historical meteorological prediction data is data from February 1 to February 10.

[0085] Optionally, the historical meteorological prediction data is meteorological data of m future time points with respect to the target meteorological prediction data, m is a positive integer. For example, the target meteorological prediction data is data on February 20, and the historical meteorological prediction data is data on February 15 at 14:22 and February 18 at 15:58.

[0086] Step 303: Obtain historical meteorological actual data corresponding to the historical meteorological prediction data.

[0087] Optionally, the terminal or the server stores a correspondence between the historical meteorological prediction data and the historical meteorological actual data. The historical meteorological actual data is determined according to the historical meteorological prediction data and the correspondence.

[0088] In other embodiments of the present application, step 302 can be replaced by "obtaining historical weather actual data of the target weather source set", and step 303 can be replaced by "obtaining historical weather prediction data corresponding to the historical weather actual data".

[0089] Step 304: determining an error prediction feature of the target weather prediction data according to the error between the historical weather prediction data and the historical weather actual data.

[0090] Optionally, the error prediction feature is determined according to the difference between the historical weather prediction data and the historical weather actual data. For example, the error prediction feature can be represented as wherein, y(t) is the historical weather prediction data, y(t) is the historical weather actual data, and t represents time. For example, the error prediction feature can be represented as a time series.

[0091] Optionally, the error prediction feature is determined according to the root mean square error (RMSE) between the historical weather prediction data and the historical weather actual data.

[0092] Optionally, the error prediction feature is determined according to the weighted quantile loss (WQL) between the historical weather prediction data and the historical weather actual data.

[0093] Optionally, the error prediction feature is determined according to the mean absolute percentage error (MAPE) between the historical weather prediction data and the historical weather actual data.

[0094] Step 305: determining a neighboring grid of the target grid.

[0095] The power generation device is located at a target grid point in the target grid, and the target grid is obtained by dividing the area where the power generation device is located.

[0096] The neighboring grid is a grid located in the peripheral area of the target grid. Optionally, the distance between the neighboring grid and the target grid is less than a distance threshold.

[0097] Optionally, the neighboring grid and the target grid have an overlapping area.

[0098] For example, as shown in FIG. 4, the wind power generation device 401 is located in the target grid, which is the center grid of a nine-grid, and the neighboring grids are the remaining eight grids of the nine-grid. In order to distinguish the target grid and the neighboring grid, Figure 4 the target grid is in gray, and the neighboring grid is in white. Figure 4 ​​

[0099] Step 306: Obtain the adjacent meteorological forecast data of adjacent grid points within the adjacent grid.

[0100] Optionally, adjacent meteorological forecast data for adjacent grid points within an adjacent grid can be obtained from the target meteorological source set. Adjacent meteorological forecast data corresponds to meteorological forecast data for adjacent grid points.

[0101] For example, there are adjacent grid points A and B in an adjacent grid. The predicted wind speed at adjacent grid point A is 5 m / s, while the predicted wind speed at adjacent grid point B is 7 m / s.

[0102] Step 307: Based on the relative positions between adjacent grids and the target grid and adjacent meteorological forecast data, obtain the spatial forecast characteristics of the target meteorological forecast data.

[0103] In the embodiments of this application, the meteorological forecast data is a scalar or a vector. For example, when the meteorological forecast data is a vector, it includes, but is not limited to, at least one of wind speed and flow velocity. When the meteorological forecast data is a scalar, it includes, but is not limited to, at least one of temperature, density, and light intensity.

[0104] For example, when adjacent meteorological forecast data are vectors, a first connection is determined between the adjacent grid and the target grid; spatial forecast features are obtained based on the projected physical quantities of the adjacent meteorological forecast data along the first connection. Here, the first connection between the adjacent grid and the target grid refers to the connection between the center point of the adjacent grid and the center point of the target grid.

[0105] For example, such as Figure 5 As shown, the gray microgrid containing point A represents the target grid. Power generation equipment D is located within target grid A. Vector BC represents the adjacent meteorological forecast data of neighboring grid B. AB is the first line connecting neighboring grid B and target grid A. Here, we assume the horizontal direction to the right is the reference direction, then θ... 邻 θ represents the angle between adjacent meteorological forecast data and the reference direction. 连线 Let y represent the angle between the first line AB and the reference direction. To calculate the projected physical quantity of adjacent meteorological forecast data onto the first line AB, we need the spatial prediction characteristic y of the target meteorological forecast data. 融合 =y 邻 ·cos(θ 邻 -θ 连线 ), y 邻 Represents adjacent weather forecast data. Optionally, θ 邻 θ 连线 and y 邻 Since it is a variable related to time t, the spatial prediction feature can be represented as a time series.

[0106] Exemplarily, in the case that the adjacent weather prediction data belongs to a scalar, the embodiment of the application converts the adjacent weather prediction data into a vector through the gradient of the adjacent weather prediction data. The gradient of the adjacent weather prediction data at the adjacent grid point is calculated; a second connecting line between the adjacent grid and the target grid is taken; and the spatial prediction feature is obtained according to the projection physical quantity of the gradient on the second connecting line.

[0107] Step 308: determining a set of time weather prediction data corresponding to the target weather prediction data.

[0108] Optionally, the set of time weather prediction data includes at least one of future weather data of the target weather prediction data and historical weather data of the target weather prediction data.

[0109] Exemplarily, the future weather data of the target weather prediction data is relative to weather data in a future period of the target weather prediction data. For example, the target weather prediction data is data on February 1, and the future weather data of the target weather prediction data is data from February 4 to February 6. Exemplarily, the future weather data of the target weather prediction data is relative to weather data in a future t time units of the target weather prediction data, t being a positive integer, and the time unit including at least one of hour, day, week, month, quarter and year. For example, the target weather prediction data is data on February 1, and the future weather data of the target weather prediction data is data in the next 3 days on February 1. Exemplarily, the historical weather data of the target weather prediction data is relative to weather data at a future a time point of the target weather prediction data, a being a positive integer. For example, the target weather prediction data is data on February 1, and the future weather data of the target weather prediction data is data at 14:54 on February 3 and 4:05 on February 4.

[0110] Exemplarily, the historical weather data of the target weather prediction data is relative to weather data in a historical period of the target weather prediction data. For example, the target weather prediction data is data on February 1, and the historical weather data of the target weather prediction data is data from January 29 to January 31. Exemplarily, the historical weather data of the target weather prediction data is relative to weather data in a historical t time units of the target weather prediction data, t being a positive integer, and the time unit including at least one of hour, day, week, month, quarter and year. For example, the target weather prediction data is data on February 1, and the historical weather data of the target weather prediction data is data in the past 3 days on February 1. Exemplarily, the future weather data of the target weather prediction data is relative to weather data at a future a time point of the target weather prediction data, a being a positive integer. For example, the target weather prediction data is data on February 1, and the historical weather data of the target weather prediction data is data at 18:54 on January 31 and 9:53 on January 29.

[0111] Optionally, the time meteorological prediction data set corresponding to the target meteorological prediction data is determined through the target meteorological source set.

[0112] Step 309: obtaining the time prediction feature according to the target meteorological prediction data, the time meteorological prediction data set, the error prediction feature and the space prediction feature, the time prediction feature corresponding to at least one of the target meteorological prediction data, the error prediction feature and the error prediction feature.

[0113] Optionally, the time prediction feature of the target meteorological prediction data is determined according to the target meteorological prediction data and the time meteorological prediction data set. Optionally, the time prediction feature of the error prediction feature is determined according to the error prediction feature. Optionally, the time prediction feature of the space prediction feature is determined according to the space prediction feature.

[0114] The time meteorological prediction data set includes at least one of future meteorological data of the target meteorological prediction data and historical meteorological data of the target meteorological prediction data, the future meteorological data of the target meteorological prediction data is meteorological data relative to future time of the target meteorological prediction data, and the historical meteorological data of the target meteorological prediction data is meteorological data relative to past time of the target meteorological prediction data.

[0115] Optionally, the historical time corresponding to the target meteorological prediction data is determined; and the historical meteorological prediction data corresponding to the historical time in the time meteorological prediction data set is taken as the first time prediction feature of the target meteorological prediction data. The first time prediction feature is a feature including previous time step data of the target meteorological prediction data. The historical time is a time before the current time.

[0116] For example, the target meteorological prediction data is data at time t, and (t-N) is taken as the historical time, N is a lag constant, and the value of N can be adjusted by the technician according to the actual demand. Then the first time prediction feature is:

[0117] y lagN (t)=y(t-N);

[0118] Wherein, y represents the physical quantity of the meteorological prediction data.

[0119] Optionally, the historical time corresponding to the target meteorological prediction data is determined; the historical meteorological prediction data corresponding to the historical time is determined from the time meteorological prediction data set; and the second time prediction feature of the target meteorological prediction data is obtained according to the difference between the target meteorological prediction data and the historical meteorological prediction data. The second time prediction feature is a feature including the difference between the historical meteorological prediction data and the target meteorological prediction data.

[0120] For example, the target meteorological prediction data is data at time t, (t-N) is taken as the historical time, and N is a difference constant. The value of N can be adjusted by the technical personnel according to actual needs. Then the second time prediction feature of the target meteorological prediction data is:

[0121] y diffNN (t) = y(t) - y(t-N);

[0122] where y represents a physical quantity of the meteorological prediction data.

[0123] Optionally, a time period corresponding to the time meteorological prediction data set is determined, the time period is divided into n sub-time periods, n meteorological prediction data corresponding to the n sub-time periods are determined from the time meteorological prediction data set, and a third time prediction feature of the target meteorological prediction data is obtained according to the mean of the n meteorological prediction data, where n is a positive integer greater than 1.

[0124] For example, the target meteorological prediction data is data at time t, and N is a rolling window size constant. The value of N can be adjusted by the technical personnel according to actual needs. Then the third time prediction feature of the target meteorological prediction data is:

[0125] y rollN (t) = [y(t) + y(t-1) + … + y(t-N+1)] / N;

[0126] where y represents a physical quantity of the meteorological prediction data.

[0127] It should be noted that since the error prediction feature and the space prediction feature can both be represented as time series, based on the processing method of the target meteorological prediction data, the time prediction feature in the error prediction feature and the time prediction feature in the space prediction feature can also be obtained.

[0128] For example, in the case where the error prediction feature is an error feature time series, a time meteorological prediction data set corresponding to the error feature time series is determined, and a time prediction feature of the error prediction feature is obtained according to the error feature time series and the time meteorological prediction data set corresponding to the error feature time series.

[0129] Optionally, a historical time corresponding to the error prediction feature is determined; historical meteorological prediction data corresponding to the historical time in the time meteorological prediction data set is determined; and a second time prediction feature of the error prediction feature is obtained according to a difference between the error prediction feature and the historical meteorological prediction data. For example, the error prediction feature is [y(t1), y(t2), …, y(t6)], the error prediction feature is a feature in the time period from t1 to t6, and then the historical meteorological prediction data of the error prediction feature is [y(t1-N), y(t2-N), …, y(t6-N)] determined from the time meteorological prediction data set; and the second time prediction feature of the error prediction feature is [y(t1)-y(t1-N), y(t2)-y(t2-N), …, y(t6)-y(t6-N)] obtained according to the difference between the error prediction feature and the historical meteorological prediction data.

[0130] Optionally, a historical time corresponding to the error prediction feature is determined; historical meteorological prediction data corresponding to the historical time in the time meteorological prediction data set is determined; and a second time prediction feature of the error prediction feature is obtained according to a difference between the error prediction feature and the historical meteorological prediction data. For example, the error prediction feature is [y(t1), y(t2), …, y(t6)], the error prediction feature is a feature in the time period from t1 to t6, and then the historical meteorological prediction data of the error prediction feature is [y(t1-N), y(t2-N), …, y(t6-N)] determined from the time meteorological prediction data set; and the second time prediction feature of the error prediction feature is [y(t1)-y(t1-N), y(t2)-y(t2-N), …, y(t6)-y(t6-N)] obtained according to the difference between the error prediction feature and the historical meteorological prediction data.

[0131] Optionally, a time period corresponding to the time meteorological prediction data set is determined; the time period is divided into q sub-time periods; q meteorological prediction data corresponding to the q sub-time periods are determined from the time meteorological prediction data set corresponding to the error feature time series; and a third time prediction feature of the error prediction feature is obtained according to a mean value of the q meteorological prediction data, q being a positive integer greater than 1.

[0132] For example, in the case of the spatial prediction feature being a spatial feature time series, a time meteorological prediction data set corresponding to the spatial feature time series is determined; and a time prediction feature of the spatial prediction feature is obtained according to the spatial feature time series and the time meteorological prediction data set corresponding to the spatial feature time series.

[0133] Optionally, a historical time corresponding to the error prediction feature is determined; historical meteorological prediction data corresponding to the historical time in the time meteorological prediction data set is determined; and a second time prediction feature of the error prediction feature is obtained according to a difference between the error prediction feature and the historical meteorological prediction data. For example, the error prediction feature is [y(t1), y(t2), …, y(t6)], the error prediction feature is a feature in the time period from t1 to t6, and then the historical meteorological prediction data of the error prediction feature is [y(t1-N), y(t2-N), …, y(t6-N)] determined from the time meteorological prediction data set; and the second time prediction feature of the error prediction feature is [y(t1)-y(t1-N), y(t2)-y(t2-N), …, y(t6)-y(t6-N)] obtained according to the difference between the error prediction feature and the historical meteorological prediction data.

[0134] Optionally, a historical time corresponding to the spatial prediction feature is determined; historical meteorological prediction data corresponding to the historical time is determined from the set of time meteorological prediction data; and a second time prediction feature of the spatial prediction feature is obtained according to a difference between the spatial prediction feature and the historical meteorological prediction data.

[0135] Optionally, a time period corresponding to the set of time meteorological prediction data is determined; the time period is divided into a number of a sub-time periods; a number of a meteorological prediction data corresponding to the number of a sub-time periods is determined from the set of time meteorological prediction data corresponding to the spatial feature time series; and a third time prediction feature of the spatial prediction feature is obtained according to a mean value of the number of a meteorological prediction data, where a is a positive integer greater than 1.

[0136] Step 310: power prediction is performed using the error prediction feature, the spatial prediction feature, the time prediction feature, and the target meteorological prediction data to obtain the predicted power of the power generation device.

[0137] Optionally, the power prediction model is used to perform power prediction on the error prediction feature, the spatial prediction feature, the time prediction feature, and the target meteorological prediction data to obtain the predicted power of the power generation device.

[0138] The training of the power prediction model includes the following steps: obtaining sample meteorological prediction data corresponding to a sample power generation device, the sample meteorological prediction data being provided by a set of sample meteorological sources; obtaining actual power corresponding to the sample meteorological prediction data; generating a sample power influence feature of the sample meteorological prediction data according to at least one of the sample meteorological prediction data and sample associated meteorological data corresponding to the sample meteorological prediction data; inputting the target meteorological prediction data and the power influence feature into the power prediction model to obtain a sample predicted power of the sample power generation device; and training the power prediction model according to a difference between the sample predicted power and the actual power.

[0139] In the embodiments of the present application, the acquisition processes of the error prediction feature, the spatial prediction feature, and the time prediction feature can be implemented by the power prediction model. Optionally, the target meteorological prediction data is input into the power prediction model; a data correction layer in the power prediction model is called to perform data processing on the target meteorological prediction data through the power influence feature to obtain the error prediction feature, the spatial prediction feature, and the time prediction feature; and a power prediction layer in the power prediction model is called to perform data processing on the target meteorological prediction data, the error prediction feature, the spatial prediction feature, and the time prediction feature to obtain the predicted power.

[0140] In summary, the embodiment generates a power impact feature of the target meteorological prediction data according to the associated meteorological data corresponding to the target meteorological prediction data, and uses the power impact feature and the target meteorological prediction data to predict the predicted power of the power generation equipment. Since the power impact feature is introduced in the process of power prediction, and the power impact feature is a feature that enhances and / or corrects the target meteorological prediction data, the power impact feature not only provides additional information for the target meteorological prediction data, but also makes the target meteorological prediction data more accurate, and thus makes the predicted power more accurate. Moreover, the target prediction meteorological data is corrected from the time, space and its own characteristics, respectively correcting the errors from different angles, and improving the accuracy of the target prediction meteorological data.

[0141] Figure 6 A flowchart of a method for obtaining a target meteorological source set provided by an example embodiment of the present application is shown. The method can be executed by the computer system 100 shown, and the method comprises: Figure 1

[0142] Step 601: Perform power prediction by using the meteorological prediction data provided by the first meteorological source set to obtain a first prediction accuracy.

[0143] The first meteorological source set comprises at least one meteorological source. Optionally, the historical prediction accuracy of the meteorological source set is calculated, and the meteorological source with the maximum historical prediction accuracy is taken as the first meteorological source set. The meteorological source set refers to the meteorological sources that can be obtained.

[0144] Obtain a first power impact feature corresponding to the meteorological prediction data provided by the first meteorological source set; in the process of data processing of the meteorological prediction data provided by the first meteorological source set by using the power prediction model, the first power impact feature is used for correction to obtain a first predicted power; and the first prediction accuracy is obtained according to the difference between the first predicted power and the first actual power. The specific process of generating the first predicted power can refer to the above-mentioned embodiments, which will not be described here.

[0145] The first prediction accuracy comprises at least one of root mean square error, weighted quantile loss, mean absolute percentage error, mean absolute scaled error and weighted absolute percentage error.

[0146] Step 602: Perform power prediction by using the meteorological prediction data provided by the i-th meteorological source in the first meteorological source set and the second meteorological source set to obtain a second prediction accuracy.

[0147] Wherein, i is an integer with an initial value of 1.

[0148] ​A second power influence feature is obtained corresponding to the meteorological forecast data provided by the i-th meteorological source in the first and second meteorological source sets. During the data processing of the meteorological forecast data provided by the i-th meteorological source in the first and second meteorological source sets using a power prediction model, the second power influence feature is used for correction to obtain the second predicted power. The second prediction accuracy is obtained based on the difference between the second predicted power and the second actual power. The specific process for generating the second predicted power can be referred to the above embodiment, and will not be repeated here.

[0149] The second prediction accuracy includes at least one of the following: root mean square error, weighted quantile loss, mean absolute percentage error, mean absolute scaling error, and weighted absolute percentage error.

[0150] For example, such as Figure 7 As shown, when calculating the second prediction accuracy, the first meteorological source set 701, the i-th meteorological source 702, and other input data 703 are substituted into the power prediction model 704 to obtain the second prediction accuracy. The first meteorological source set 701 includes m first meteorological sources, where m is a positive integer. Calculating the second prediction accuracy requires the meteorological prediction data, error prediction features, adjacent meteorological prediction data, time series of meteorological prediction data, time prediction features, and spatial prediction features of each first meteorological source. Similarly, the i-th meteorological source 702 requires the meteorological prediction data, error prediction features, adjacent meteorological prediction data, time series of meteorological prediction data, time prediction features of error prediction features, and time prediction features of spatial prediction features. Furthermore, the other input data are optional input data defined by the technician; for example, other input data include natural errors, calculation errors, etc.

[0151] Step 603: If the second prediction accuracy is greater than the first prediction accuracy, merge the i-th meteorological source into the first meteorological source set and update the first meteorological source set; and update i to i+1.

[0152] If the accuracy of the second prediction is no greater than that of the first prediction, the i-th meteorological source is discarded.

[0153] For example, such as Figure 7 As shown, it is determined whether the second prediction accuracy is greater than the first prediction accuracy. If the second prediction accuracy is greater than the first prediction accuracy, the i-th meteorological source is merged into the first meteorological source set and the first meteorological source set is updated; if the second prediction accuracy is not greater than the first prediction accuracy, the i-th meteorological source is discarded.

[0154] Step 604: Repeat the above three steps until the second meteorological source set is traversed, and take the first meteorological source set as the target meteorological source set.

[0155] As shown in Figure 7 If the second meteorological source set has been traversed, the first meteorological source set is taken as the target meteorological source set; if the second meteorological source set has not been traversed, i is updated as i+1, the meteorological prediction data provided by the i-th meteorological source is taken, and the above three steps are repeated.

[0156] In summary, the embodiment provides a method for generating a target meteorological source set, which selects a target meteorological source set suitable for a power generation device, and improves the accuracy of power prediction when using the meteorological prediction data provided by the target meteorological source set.

[0157] In the following embodiment, the power generation device is taken as a wind power generation device. Since wind power generation is unstable, the output power of the wind power generation device needs to be predicted, which has practical significance for power grid dispatching, improving the access capacity of wind power, and reducing system operation cost. In this embodiment, the power generation strategy of the wind power generation device can be formulated according to the predicted power obtained by prediction.

[0158] Figure 8 A flowchart of a method for generating a power generation strategy provided by an example embodiment of the present application is shown. The method can be executed by Figure 1 The method includes the following steps:

[0159] Step 801: Obtain target meteorological prediction data corresponding to the wind power generation device, the target meteorological prediction data being provided by a target meteorological source set.

[0160] Optionally, the target meteorological prediction data is the meteorological prediction data of the wind power generation device. The meteorological prediction data refers to the prediction data of future meteorological conditions. The target meteorological prediction data includes at least one of wind speed, wind direction, temperature, pressure, and density. In a specific embodiment, the target meteorological prediction data is the meteorological prediction data of the wind field where the wind power generation device is located.

[0161] The target meteorological source set includes at least one meteorological source, and the meteorological source is used to provide prediction data of future meteorological conditions.

[0162] Step 802: Generate a power impact feature of the target meteorological prediction data according to at least one of the target meteorological prediction data and the associated meteorological data corresponding to the target meteorological prediction data.

[0163] The power impact feature includes at least one of an error prediction feature, a spatial prediction feature, and a time prediction feature.

[0164] The error prediction feature is used to correct statistical errors of the target meteorological prediction data, the space prediction feature is used to correct spatial errors of the target meteorological prediction data, and the time prediction feature is used to correct temporal errors of the target meteorological prediction data.

[0165] Step 803: performing power prediction using the power influence feature and the target meteorological prediction data to obtain the predicted power of the power generation device.

[0166] Optionally, the target meteorological prediction data and the power influence feature are jointly input into the power prediction model; the power prediction model is called to perform data processing on the target meteorological prediction data and the power influence feature to obtain the predicted power.

[0167] The training of the power prediction model includes the following steps: obtaining sample meteorological prediction data corresponding to the sample power generation device, the sample meteorological prediction data being provided by a sample meteorological source set; obtaining actual power corresponding to the sample meteorological prediction data; generating a sample power influence feature of the sample meteorological prediction data according to at least one of the sample meteorological prediction data and sample associated meteorological data corresponding to the sample meteorological prediction data; jointly inputting the target meteorological prediction data and the power influence feature into the power prediction model to obtain sample predicted power of the sample power generation device; and training the power prediction model according to a difference between the sample predicted power and the actual power.

[0168] In the embodiments of the present application, the process of obtaining the power influence feature can be implemented by the power prediction model. Optionally, the target meteorological prediction data is input into the power prediction model; a data correction layer in the power prediction model is called to perform data processing on the target meteorological prediction data by the power influence feature to obtain the power influence feature; and a power prediction layer in the power prediction model is called to perform data processing on the target meteorological prediction data and the power influence feature to obtain the predicted power.

[0169] Step 804: determining a power generation strategy of the wind power generation device according to the predicted power.

[0170] Optionally, the power generation strategy of the wind power generation device is determined according to a relationship between the predicted power and a power threshold. For example, when the predicted power is greater than the power threshold, power generation strategy A is adopted; and when the predicted power is greater than the power threshold, power generation strategy B is adopted.

[0171] For example, the predicted power of the wind power generation device A on March 4 is greater than the power threshold, and therefore a power generation strategy is formulated to connect the wind power generation device A to the power grid B on March 4.

[0172] In conclusion, the embodiment of the present application has high accuracy of the obtained predicted power, and the generated power generation strategy of the wind power generation device can provide more accurate information for economic scheduling of the power grid, optimize the output of conventional units, reduce the operation cost of the power system, reduce the system cost of wind power renewable energy access to the power grid, and enhance the safety, reliability and controllability of the system.

[0173] The following is an apparatus embodiment of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the apparatus embodiments of the present application, please refer to the method embodiments of the present application.

[0174] Please refer to Figure 9 which shows a block diagram of a power generation power prediction device provided by an embodiment of the present application. The above functions can be realized by hardware or corresponding software executed by hardware. The device 900 includes:

[0175] An acquisition module 901 is configured to acquire target meteorological prediction data corresponding to a power generation device, wherein the target meteorological prediction data is provided by a target meteorological source set, and the target meteorological prediction data is used to predict the power of the power generation device.

[0176] A processing module 902 is configured to generate a power influence feature of the target meteorological prediction data according to associated meteorological data corresponding to the target meteorological prediction data, wherein the associated meteorological data includes complementary information that is not possessed by the target meteorological prediction data and is used to improve the power prediction accuracy of the target meteorological prediction data, and the power influence feature is a feature that enhances and / or corrects the target meteorological prediction data.

[0177] The processing module 902 is further configured to perform power prediction using the power influence feature and the target meteorological prediction data, and obtain a predicted power of the power generation device.

[0178] In an optional design, the power influence feature of the target meteorological prediction data includes at least one of an error prediction feature, a spatial prediction feature and a time prediction feature.

[0179] The error prediction feature is used to correct statistical errors of the target meteorological prediction data.

[0180] The spatial prediction feature is used to correct spatial errors of the target meteorological prediction data.

[0181] The time prediction feature is used to correct time errors of the target meteorological prediction data.

[0182] In an optional design, the processing module 902 is further configured to acquire historical weather prediction data provided by a target weather source set, acquire historical weather actual data corresponding to the historical weather prediction data, and determine the error prediction feature of the target weather prediction data according to an error between the historical weather prediction data and the historical weather actual data.

[0183] In an optional design, the processing module 902 is further configured to determine the error prediction feature according to a difference between the historical weather prediction data and the historical weather actual data, or according to a root mean square error between the historical weather prediction data and the historical weather actual data, or according to a weighted quantile loss between the historical weather prediction data and the historical weather actual data, or according to a mean absolute percentage error between the historical weather prediction data and the historical weather actual data.

[0184] In an optional design, the power generation device is located in a target grid, the target grid is obtained by dividing a region where the power generation device is located, the target weather prediction data is weather prediction data of the target grid, and the processing module 902 is further configured to determine a neighboring grid of the target grid, acquire neighboring weather prediction data of the neighboring grid, and obtain the spatial prediction feature of the target weather prediction data according to a relative position between the neighboring grid and the target grid and the neighboring weather prediction data.

[0185] In an optional design, the neighboring weather prediction data belongs to a vector, and the processing module 902 is further configured to determine a first connecting line between the neighboring grid and the target grid, and obtain the spatial prediction feature according to a projection physical quantity of the neighboring weather prediction data on the first connecting line.

[0186] In an optional design, the neighboring weather prediction data belongs to a scalar, and the processing module 902 is further configured to calculate a gradient of the neighboring weather prediction data in the neighboring grid, take a second connecting line between the neighboring grid and the target grid, and obtain the spatial prediction feature according to a projection physical quantity of the gradient on the second connecting line.

[0187] In an optional design, the processing module 902 is further configured to determine a set of time weather prediction data corresponding to the target weather prediction data, and obtain the time prediction feature of the target weather prediction data according to the target weather prediction data and the set of time weather prediction data.

[0188] In an optional design, the processing module 902 is further configured to determine a historical time corresponding to the target meteorological prediction data; determine historical meteorological prediction data corresponding to the historical time in the set of time meteorological prediction data as a first time prediction feature of the target meteorological prediction data; or determine the historical time corresponding to the target meteorological prediction data; determine historical meteorological prediction data corresponding to the historical time from the set of time meteorological prediction data; obtain a second time prediction feature of the target meteorological prediction data according to a difference between the target meteorological prediction data and the historical meteorological prediction data; or determine a time period corresponding to the set of time meteorological prediction data; divide the time period into n sub-time periods; determine n meteorological prediction data corresponding to the n sub-time periods from the set of time meteorological prediction data; and obtain a third time prediction feature of the target meteorological prediction data according to a mean value of the n meteorological prediction data, where n is a positive integer greater than 1.

[0189] In an optional design, the obtaining module 901 is further configured to perform power prediction by using meteorological prediction data provided by a first set of meteorological sources to obtain a first prediction accuracy; perform power prediction by using meteorological prediction data provided by an i th meteorological source in the first set of meteorological sources and a second set of meteorological sources to obtain a second prediction accuracy, where i is an integer with an initial value of 1; in a case where the second prediction accuracy is greater than the first prediction accuracy, merge the i th meteorological source into the first set of meteorological sources, and update the first set of meteorological sources; and update i to i+1; repeat the above three steps until the first set of meteorological sources is obtained by traversing the second set of meteorological sources.

[0190] In an optional design, the obtaining module 901 is further configured to obtain a first power influence feature corresponding to meteorological prediction data provided by the first set of meteorological sources; use the first power influence feature to correct data processing of the meteorological prediction data provided by the first set of meteorological sources by using a power prediction model to obtain a first predicted power; and obtain the first prediction accuracy according to a difference between the first predicted power and a first actual power.

[0191] In an optional design, the obtaining module 901 is further configured to obtain a second power influence feature corresponding to meteorological prediction data provided by the i th meteorological source in the first set of meteorological sources and the second set of meteorological sources; use the second power influence feature to correct data processing of the meteorological prediction data provided by the i th meteorological source in the first set of meteorological sources and the second set of meteorological sources by using a power prediction model to obtain a second predicted power; and obtain the second prediction accuracy according to a difference between the second predicted power and a second actual power.

[0192] In an optional design, the prediction module 903 is configured to input the target meteorological prediction data and the power influence feature into a power prediction model, and call the power prediction model to process the target meteorological prediction data and the power influence feature to obtain the predicted power.

[0193] In an optional design, the acquisition module 901 is further configured to acquire sample meteorological prediction data corresponding to a sample power generation device, where the sample meteorological prediction data is provided by a sample meteorological source set; and acquire actual power corresponding to the sample meteorological prediction data. The processing module 902 is further configured to generate a sample power influence feature of the sample meteorological prediction data according to at least one of the sample meteorological prediction data and sample associated meteorological data corresponding to the sample meteorological prediction data; input the target meteorological prediction data and the power influence feature into the power prediction model to obtain a sample predicted power of the sample power generation device; and train the power prediction model according to a difference between the sample predicted power and the actual power.

[0194] In summary, the embodiment generates a power influence feature of target meteorological prediction data according to associated meteorological data corresponding to the target meteorological prediction data, corrects the target meteorological prediction data using the power influence feature, and obtains a device predicted power of a meteorological device. Since the power influence feature is introduced into the process of power prediction, the power influence feature is a feature of enhancing and / or correcting the target meteorological prediction data, and the power influence feature not only provides additional information for the target meteorological prediction data, but also makes the target meteorological prediction data more accurate, and further makes the predicted power more accurate.

[0195] Figure 10 FIG. 1 is a structural schematic diagram of a computer device according to an example embodiment. The computer device 1000 includes a central processing unit (CPU) 1001, a system memory 1004 including a random access memory (RAM) 1002 and a read-only memory (ROM) 1003, and a system bus 1005 connecting the system memory 1004 and the central processing unit 1001. The computer device 1000 further includes a basic input / output system (I / O) 1006 that helps transfer information between various devices in the computer device, and a mass storage device 1007 for storing an operating system 1013, application programs 1014, and other program modules 1015.

[0196] The basic input / output system 1006 includes the various components expected to be found within the basic input / output system, including a display 1008 for displaying information and input devices 1009, such as a mouse, keyboard, or the like, for inputting information. Both the display 1008 and input devices 1009 are connected to the central processing unit 1001 through an input / output controller 1010 connected to the system bus 1005. The basic input / output system 1006 can also include the input / output controller 1010 for receiving and processing input from a number of other devices, including a keyboard, mouse, or electronic stylus. Similarly, the input / output controller 1010 provides output to a display screen, printer, or other type of output device.

[0197] The mass storage device 1007 is connected to the central processing unit 1001 through a mass storage controller (not shown) connected to the system bus 1005. The mass storage device 1007 and its associated computer device readable medium provide non-volatile storage for the computer device 1000. That is, the mass storage device 1007 can comprise a computer device readable medium (not shown) such as a hard disk or a Compact Disc Read-Only Memory (CD-ROM) drive.

[0198] Without loss of generality, the computer device readable medium can include computer device storage media and communication media. Computer device storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer device readable instructions, data structures, program modules or other data. Computer device storage media include RAM, ROM, Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM), digital video disc (DVD), or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices. Of course, those skilled in the art will recognize that the computer device storage media described above can be embodied in any computer device readable medium known in the art. The above-described system memory 1004 and mass storage device 1007 can be collectively referred to as memory.

[0199] According to various embodiments of the present disclosure, the computer device 1000 can further operate connected to a remote computer device on a network, such as the Internet, through a network connection. That is, the computer device 1000 can be connected to a network 1011 through a network interface unit 1012 connected to the system bus 1005, or can be connected to other types of networks or remote computer device systems (not shown) using the network interface unit 1012.

[0200] The memory further includes one or more programs stored in the memory, and the central processing unit 1001 implements all or part of the steps of the power generation power prediction method by executing the one or more programs.

[0201] In an exemplary embodiment, a computer readable storage medium is also provided, and the computer readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by a processor to implement the power generation power prediction method provided by each method embodiment.

[0202] The present application also provides a computer readable storage medium, and the storage medium stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by the processor to implement the power generation power prediction method provided by the above method embodiment.

[0203] The present application also provides a computer program product or computer program, which includes computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to make the computer device execute the power generation power prediction method provided by the above aspect embodiment.

[0204] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0205] Those of ordinary skill in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or by programs instructing relevant hardware, and the programs can be stored in a computer readable storage medium, and the storage medium can be a read-only memory, a magnetic disk or an optical disk.

[0206] The above-mentioned only for optional embodiments of the present application, and do not limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application, should be included in the protection scope of the present application.

Claims

1. A method of predicting power generation, characterized by, The method comprises: acquiring target meteorological prediction data corresponding to a power generation device, the target meteorological prediction data being provided by a target meteorological source set, and the target meteorological prediction data being used to predict power of the power generation device; generating a power influence feature of the target meteorological prediction data according to at least one of the target meteorological prediction data and associated meteorological data corresponding to the target meteorological prediction data, the associated meteorological data comprising complementary information that is not possessed by the target meteorological prediction data and is used to improve power prediction accuracy of the target meteorological prediction data, and the power influence feature being a feature for enhancing and / or correcting the target meteorological prediction data; performing power prediction using the power influence feature and the target meteorological prediction data to obtain predicted power of the power generation device; performing power prediction by using meteorological prediction data provided by a first meteorological source set to obtain a first prediction accuracy, performing power prediction by using meteorological prediction data provided by an i-th meteorological source in the first meteorological source set and a second meteorological source set to obtain a second prediction accuracy, i being an integer with an initial value of 1, in a case where the second prediction accuracy is greater than the first prediction accuracy, merging the i-th meteorological source into the first meteorological source set to update the first meteorological source set, and updating i to i+1; and re-entering the step of performing power prediction by using meteorological prediction data provided by the first meteorological source set to obtain the first prediction accuracy, until the first meteorological source set is taken as the target meteorological source set after traversing the second meteorological source set.

2. The method of claim 1, wherein, The power influence feature of the target meteorological prediction data comprises at least one of an error prediction feature, a spatial prediction feature, and a time prediction feature; The error prediction feature is used to correct statistical errors of the target meteorological prediction data; The spatial prediction feature is used to correct spatial errors of the target meteorological prediction data; The time prediction feature is used to correct time errors of the target meteorological prediction data.

3. The method of claim 2, wherein, The generating of the power influence feature of the target meteorological prediction data according to the target meteorological prediction data and the associated meteorological data corresponding to the target meteorological prediction data comprises: acquiring historical meteorological prediction data in the target meteorological source set; acquiring historical meteorological actual data corresponding to the historical meteorological prediction data; determining the error prediction feature of the target meteorological prediction data according to errors between the historical meteorological prediction data and the historical meteorological actual data.

4. The method of claim 3, wherein, The determining of the error prediction feature of the target meteorological prediction data according to errors between the historical meteorological prediction data and the historical meteorological actual data comprises: determining the error prediction feature according to differences between the historical meteorological prediction data and the historical meteorological actual data; or, determining the error prediction feature according to root mean square errors between the historical meteorological prediction data and the historical meteorological actual data; or, determining the error prediction feature according to weighted quantile losses between the historical meteorological prediction data and the historical meteorological actual data; Or, according to the average absolute percentage error between the historical meteorological prediction data and the historical meteorological actual data, the error prediction feature is determined.

5. The method of claim 2, wherein, The power generation equipment is located in a target grid; the target grid is obtained by dividing a region where the power generation equipment is located; and the target meteorological prediction data include meteorological prediction data of the target grid. The power influence feature of the target meteorological prediction data is generated according to the associated meteorological data corresponding to the target meteorological prediction data, and includes: A neighboring grid of the target grid is determined. Neighboring meteorological prediction data of the neighboring grid is obtained. The spatial prediction feature of the target meteorological prediction data is obtained according to the relative position between the neighboring grid and the target grid and the neighboring meteorological prediction data.

6. The method of claim 5, wherein, The neighboring meteorological prediction data belong to a vector. The spatial prediction feature of the target meteorological prediction data is obtained according to the relative position between the neighboring grid and the target grid and the neighboring meteorological prediction data, and includes: A first connecting line between the neighboring grid and the target grid is determined. The spatial prediction feature is obtained according to a projection physical quantity of the neighboring meteorological prediction data on the first connecting line.

7. The method of claim 5, wherein, The neighboring meteorological prediction data belong to a scalar. The spatial prediction feature of the target meteorological prediction data is obtained according to the relative position between the neighboring grid and the target grid and the neighboring meteorological prediction data, and includes: A gradient of the neighboring meteorological prediction data in the neighboring grid is calculated. A second connecting line between the neighboring grid and the target grid is determined. The spatial prediction feature is obtained according to a projection physical quantity of the gradient on the second connecting line.

8. The method of claim 2, wherein, The power influence feature of the target meteorological prediction data is generated according to the target meteorological prediction data, and includes: A time meteorological prediction data set corresponding to the target meteorological prediction data is determined. The time prediction feature of the target meteorological prediction data is obtained according to the target meteorological prediction data and the time meteorological prediction data set.

9. The method of claim 8, wherein, The time prediction feature of the target meteorological prediction data is obtained according to the target meteorological prediction data and the time meteorological prediction data set, and includes: A historical time corresponding to the target meteorological prediction data is determined; and historical meteorological prediction data corresponding to the historical time in the time meteorological prediction data set is taken as a first time prediction feature of the target meteorological prediction data; Or, the historical time corresponding to the target meteorological prediction data is determined; historical meteorological prediction data corresponding to the historical time is determined from the time meteorological prediction data set; and a second time prediction feature of the target meteorological prediction data is obtained according to a difference between the target meteorological prediction data and the historical meteorological prediction data. Or, determine the time period corresponding to the time weather forecast data set; divide the time period into n sub-time periods; determine n weather forecast data corresponding to the n sub-time periods from the time weather forecast data set; according to the mean of the n weather forecast data, obtain the third time prediction feature of the target weather forecast data, and n is a positive integer greater than 1.

10. The method of claim 1, wherein, The power prediction is performed by using the weather forecast data provided by the first weather source set, and a first prediction accuracy is obtained, including: obtaining a first power influence feature corresponding to the weather forecast data provided by the first weather source set; In the process of data processing of the weather forecast data provided by the first weather source set by the power prediction model, the first power influence feature is used for correction to obtain a first predicted power; According to the difference between the first predicted power and the first actual power, the first prediction accuracy is obtained.

11. The method of claim 1, wherein, The power prediction is performed by using the weather forecast data provided by the first weather source set and the i-th weather source in the second weather source set, and a second prediction accuracy is obtained, including: obtaining a second power influence feature corresponding to the weather forecast data provided by the first weather source set and the i-th weather source in the second weather source set; In the process of data processing of the weather forecast data provided by the first weather source set and the i-th weather source in the second weather source set by the power prediction model, the second power influence feature is used for correction to obtain a second predicted power; According to the difference between the second predicted power and the second actual power, the second prediction accuracy is obtained.

12. The method according to any one of claims 1 to 9, characterized in that, The power prediction is performed by using the power influence feature and the target weather forecast data, and the predicted power of the power generation equipment is obtained, including: The target weather forecast data and the power influence feature are input into a power prediction model; The power prediction model is called to process the target weather forecast data and the power influence feature to obtain the predicted power.

13. The method of claim 12, wherein, The method further includes: obtaining sample weather forecast data corresponding to a sample power generation equipment, the sample weather forecast data being provided by a sample weather source set; obtaining an actual power corresponding to the sample weather forecast data; generating a sample power influence feature of the sample weather forecast data according to at least one of the sample weather forecast data and sample associated weather data corresponding to the sample weather forecast data; The target weather forecast data and the power influence feature are input into the power prediction model to obtain a sample predicted power of the sample power generation equipment; The power prediction model is trained according to the difference between the sample predicted power and the actual power.

14. A power generation power prediction device characterized by comprising: The device includes: an acquisition module configured to acquire target weather forecast data corresponding to a power generation equipment, the target weather forecast data being provided by a target weather source set, and the target weather forecast data being used to predict a power of the power generation equipment; The processing module is configured to generate a power impact feature of the target meteorological prediction data according to at least one of the target meteorological prediction data and associated meteorological data corresponding to the target meteorological prediction data, the associated meteorological data including complementary information that is not included in the target meteorological prediction data and is used to improve the power prediction accuracy of the target meteorological prediction data, and the power impact feature being a feature for enhancing and / or correcting the target meteorological prediction data. The processing module is further configured to perform power prediction using the power impact feature and the target meteorological prediction data to obtain a predicted power of the power generation device. The processing module is further configured to perform power prediction using meteorological prediction data provided by a first set of meteorological sources to obtain a first prediction accuracy, perform power prediction using meteorological prediction data provided by an i th meteorological source in a second set of meteorological sources to obtain a second prediction accuracy, where i is an integer with an initial value of 1, merge the i th meteorological source into the first set of meteorological sources and update the first set of meteorological sources when the second prediction accuracy is greater than the first prediction accuracy, and update i to i+1, and re-enter the step of performing power prediction using meteorological prediction data provided by the first set of meteorological sources to obtain the first prediction accuracy until the second set of meteorological sources is traversed, and the first set of meteorological sources is used as the target set of meteorological sources.

15. A computer device, comprising: The computer device includes a processor and a memory, and the memory stores at least one instruction, at least one program, a code set, or an instruction set, which are loaded and executed by the processor to implement the power generation power prediction method according to any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that, The computer readable storage medium stores at least one program code, which is loaded and executed by the processor to implement the power generation power prediction method according to any one of claims 1 to 13.

17. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instruction is executed by the processor to implement the power generation power prediction method according to any one of claims 1 to 13.

Images