Meteorological data processing method, device, computer equipment and readable storage medium

Abstract

The present application relates to a meteorological data processing method, device, computer equipment and readable storage medium. The method includes: pre-processing the initial meteorological data at each sampling time within a preset time period corresponding to all grid points in the target area, obtaining the direct irradiance data of all grid points at each sampling time, and the index meteorological data related to the direct irradiance, and screening out the direct irradiance data and index meteorological data corresponding to the preset number of times before and after the specified time of the known points adjacent to the unknown points in the grid points from the direct irradiance data and the index meteorological data as meteorological characteristic data, based on the target meteorological data inference model, obtaining the target meteorological inference data of the unknown point at the specified time through the meteorological characteristic data; the above method can process the initial meteorological data of different qualities, so as to further infer the meteorological data of the unknown point through the initial meteorological data, thereby improving the accuracy of the unknown point inference result.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic power prediction, and in particular to a meteorological data processing method, device, computer equipment and readable storage medium. Background Art

[0002] Numerical Weather Prediction (NWP) meteorological data such as solar irradiance are very important input features for photovoltaic power prediction. In order to achieve distributed photovoltaic prediction, it is very uneconomical to install distributed photovoltaic meteorological collection devices for full coverage. Therefore, how to obtain high-precision NWP meteorological data such as solar irradiance at low cost is one of the difficulties in distributed photovoltaic prediction research.

[0003] In traditional technology, distributed photovoltaic meteorological collection devices are installed in the target area to collect NWP meteorological data of known points in all grid points; at the same time, bilinear interpolation is used to perform bilinear interpolation on the NWP irradiation data of known points to deduce the NWP irradiation data of unknown points in all grids. However, traditional deduction methods are highly sensitive to the NWP meteorological data of unknown points. When the quality of the NWP meteorological data of known points is poor, the deduction results of unknown points will be less accurate. Summary of the invention

[0004] Based on this, it is necessary to provide a meteorological data processing method, device, computer equipment and readable storage medium that can improve the accuracy of the estimated results in response to the above technical problems.

[0005] A meteorological data processing method, the method comprising:

[0006] Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, obtaining the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance;

[0007] Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0008] Based on the target meteorological data extrapolation model, the target meteorological extrapolation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data extrapolation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0009] In one embodiment, the initial meteorological data at each sampling time within a preset time period corresponding to the grid points in the target area are preprocessed to obtain the direct irradiance data of all grid points at each sampling time, as well as the index meteorological data related to the direct irradiance, including:

[0010] An outlier processing method is adopted to perform exception processing on the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, and the direct irradiance data of the grid point at each sampling moment after exception processing and the index meteorological data related to the direct irradiance are obtained.

[0011] In one embodiment, the method further comprises:

[0012] The interpolation method is used to interpolate the initial meteorological data at each sampling moment within the preset time period corresponding to the grid points in the target area, and obtain the direct irradiance data of the grid points at each sampling moment after interpolation, as well as the index meteorological data related to the direct irradiance.

[0013] In one embodiment, the direct irradiance data and the index meteorological data are selected from the direct irradiance data and the index meteorological data, and the direct irradiance data and the index meteorological data corresponding to the known points adjacent to the unknown point in the grid points at a preset number of times before and after the specified time are selected as meteorological characteristic data, including:

[0014] Performing calculation processing on the direct irradiance data and the index meteorological data to obtain a correlation coefficient between the direct irradiance data and the index meteorological data on each designated day within the preset time period;

[0015] The direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point are selected as meteorological characteristic data through the correlation coefficient.

[0016] In one embodiment, the direct irradiance data and the index meteorological data are processed to obtain the correlation coefficient between the direct irradiance data and the index meteorological data on each specified day within the preset time period, including:

[0017] The direct irradiance data and the index meteorological data are subjected to operation processing, the Pearson correlation coefficient between the direct irradiance data and the index meteorological data is calculated, and the Pearson correlation coefficient is used as the correlation coefficient; the operation processing includes a combination of addition, subtraction, multiplication and division.

[0018] In one embodiment, the direct irradiance data and the index meteorological data corresponding to a preset number of time periods before and after a specified time period of a known point adjacent to an unknown point in the grid point are selected through the correlation coefficient as meteorological characteristic data, including:

[0019] For each specified day, the correlation coefficients corresponding to each type of index meteorological data are averaged and the absolute value is taken to obtain the absolute average correlation coefficient;

[0020] Determining a strong correlation coefficient and a weak correlation coefficient according to the absolute average correlation coefficient;

[0021] Acquire the direct irradiance data and the index meteorological data corresponding to the strong correlation coefficient and / or the weak correlation coefficient;

[0022] From the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid are selected as the meteorological characteristic data.

[0023] In one embodiment, determining the strong correlation coefficient and the weak correlation coefficient according to the absolute average correlation coefficient includes:

[0024] Each absolute average correlation coefficient is compared with a preset correlation threshold value, and the strong correlation coefficient and the weak correlation coefficient are determined according to the comparison result.

[0025] A meteorological data processing device, comprising:

[0026] A preprocessing module is used to preprocess the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, and obtain the direct irradiance data of the grid point at each sampling time, as well as the index meteorological data related to the direct irradiance;

[0027] A data screening module, used to screen out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0028] The calculation module is used to calculate the target meteorological data based on the target meteorological data, and obtain the target meteorological calculation data of the unknown point at the specified time through the meteorological characteristic data; the target meteorological data calculation model is a model obtained by training with meteorological characteristic data extracted from historical meteorological data as sample data.

[0029] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0030] Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, obtaining the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance;

[0031] Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0032] Based on the target meteorological data extrapolation model, the target meteorological extrapolation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data extrapolation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0033] A readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0034] Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, obtaining the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance;

[0035] Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0036] Based on the target meteorological data extrapolation model, the target meteorological extrapolation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data extrapolation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0037] The above-mentioned meteorological data processing method, device, computer equipment and readable storage medium, the computer equipment pre-processes the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, obtains the direct irradiance data of the grid point at each sampling moment, and the index meteorological data related to the direct irradiance, and selects the direct irradiance data and index meteorological data corresponding to the preset number of moments before and after the specified moment of the known points adjacent to the unknown points in the grid points from the direct irradiance data and the index meteorological data as meteorological characteristic data, and obtains the target meteorological calculation data of the unknown points at the specified moment through the meteorological characteristic data based on the target meteorological data calculation model; the above-mentioned method can process initial meteorological data of different qualities, so as to further calculate the meteorological data of the unknown points in the grid points through the initial meteorological data, thereby improving the accuracy of the calculation results of the unknown points. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of a flow chart of a meteorological data processing method in one embodiment;

[0039] Figure 2 is an internal structure diagram of a computer device in one embodiment;

[0040] Figure 3 A schematic diagram of a process for selecting meteorological characteristic data in one embodiment;

[0041] Figure 4 A schematic diagram of a specific process of selecting meteorological characteristic data in another embodiment;

[0042] Figure 5 1 is a structural block diagram of a meteorological data processing device in one embodiment. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0044] In one embodiment, Figure 1 The figure provides a flow chart of a method for processing meteorological data, and the method is applied to Figure 2 The computer device in the example is used to illustrate ( Figure 2 The internal structure of the computer device is shown), comprising the following steps:

[0045] Step S1000: pre-process the initial meteorological data at each sampling time within a preset time period corresponding to the grid points in the target area, and obtain the direct irradiance data of the grid points at each sampling time, as well as the index meteorological data related to the direct irradiance.

[0046] Specifically, the target area can be divided into grid areas, and the intersection of the grids can be understood as grid points; the preset time period can be a historical time period, which can be greater than or equal to 1 year. The above-mentioned initial meteorological data can be numerical weather forecast meteorological data collected by the distributed photovoltaic meteorological collection device, that is, NWP meteorological data; it can include NWP meteorological data at each sampling time within the preset time period corresponding to all grid points in the target area; the initial meteorological data can include meteorological data such as wind, precipitation, temperature, air pressure, relative humidity, etc., and the initial meteorological data can contain abnormal or missing data; the time resolution of the sampling data can be at the hourly level. The above-mentioned target area range can be understood as a spatial range. Within the spatial range, there can be a certain correlation between different NWP meteorological data in the initial meteorological data. It should be noted that the above-mentioned preprocessing can be at least one of feature encoding processing, feature selection processing, normalization processing, interpolation processing, average filling processing, and outlier processing.

[0047] In this embodiment, the time resolution of different NWP meteorological data can be 1 hour, the spatial range can be the target area range of about 100km*100km, and the spatial resolution of each NWP meteorological data within the spatial range can be 15km~20km; that is, within the spatial range of 100km*100km, there can be 5~7 rows and 5~7 columns with a total of 25~49 known NWP meteorological data points, and these known NWP meteorological data points have corresponding known NWP meteorological data. Each NWP meteorological data point can be called a grid point; the known NWP meteorological data can be called the NWP meteorological data of the known point in the grid point; the grid point can be understood as the spatial position corresponding to the NWP meteorological data point; the known points in the target area can form a matrix sequence. In addition, in addition to the known NWP meteorological data points, the spatial range can also include unknown NWP meteorological data points, that is, unknown points in the grid points.

[0048] It can be understood that the meteorological factors that have a strong correlation with direct irradiance may include four factors other than diffuse irradiance, temperature, wind speed and humidity, as well as other meteorological factors other than these four factors. That is, the initial meteorological data may also include other meteorological data other than the five meteorological data of direct irradiance, diffuse irradiance, temperature, wind speed and humidity; if other meteorological data are to be acquired and the acquisition cost is low, this embodiment may also acquire other meteorological data other than the five meteorological data as initial meteorological data. However, since the main meteorological factors that have a strong correlation with direct irradiance are diffuse irradiance, temperature, wind speed and humidity, in this embodiment, the meteorological data processing process is described by taking the initial meteorological data including the five meteorological data of direct irradiance data, diffuse irradiance data, temperature data, wind speed data and humidity data as an example.

[0049] The direct irradiance data can represent the irradiance data of the sun directly irradiating the ground; the scattered irradiance data can represent the irradiance data of the sun reaching the ground after being reflected by atmospheric molecules, water vapor, dust and other particles. The computer device can pre-process the initial meteorological data to obtain the direct irradiance data of all grid points at each sampling time, as well as the index meteorological data related to the direct irradiance. In this embodiment, the index meteorological data can include scattered irradiance data, temperature data, wind speed data and humidity data.

[0050] Step S2000, from the direct irradiance data and the index meteorological data, filter out the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data.

[0051] Specifically, the computer device can filter out the direct irradiance data and index meteorological data at each sampling time within the preset time period corresponding to all grid points in the target area, and the direct irradiance data and index meteorological data corresponding to the preset number of times before and after the specified time of the known points adjacent to the unknown points in the grid points required for inferring the direct irradiance data and index meteorological data of the unknown points in the grid points. Optionally, the grid points include known points and unknown points, and the known points can be understood as grid points with a device for collecting meteorological data, and the unknown points can be understood as grid points without a device for collecting meteorological data.

[0052] If all grid points in the target area form a 5*5 rectangular array, where the point at the 3rd row and 3rd column is an unknown point, the known points adjacent to the unknown point in the grid points include: the points at the 2nd row and 2nd column, the 2nd row and 3rd column, the 2nd row and 4th column, the 3rd row and 2nd column, the 3rd row and 4th column, the 4th row and 2nd column, the 4th row and 3rd column, and the 4th row and 4th column.

[0053] In this embodiment, the above-mentioned designated time can be understood as any time of a day within the preset time period; the above-mentioned preset number can be any number. n The meteorological data of N moments before and after can be respectively n-2 , time t n-1 , time t n+1 , time t n+2 ; where the specified time t n Multiple moments before and a specified time t n The multiple moments after can be equal or unequal, as long as the specified time t n The filtered meteorological data can be used before and after. Optionally, the filtered direct irradiance data and index meteorological data can be used as meteorological characteristic data.

[0054] Step S3000: based on the target meteorological data estimation model, the target meteorological estimation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data estimation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0055] Specifically, the computer device can input the meteorological characteristic data into the target meteorological data estimation model, and accurately estimate the NWP meteorological data of the unknown point at the specified time through the target meteorological data estimation model to output the target meteorological estimation data; the target meteorological estimation data may include the estimated direct irradiance data and index meteorological data of the unknown point at the specified time. The above-mentioned target meteorological data estimation model may be a three-dimensional convolutional neural network model, which is a network model obtained by training the initial meteorological data estimation model through the meteorological characteristic data extracted from the historical meteorological data. Among them, the above-mentioned meteorological characteristic data and the target meteorological estimation data may be the corresponding meteorological data at the same date and different times in the preset time period. For example, the preset time period is 1 year, that is, 365 days. If the present embodiment needs to estimate the meteorological data at time t within the 108th day, at this time, the same date may be the 108th day; the meteorological characteristic data may be the meteorological data at multiple different times before and after time t within the 108th day within the preset time period; the target meteorological estimation data may be the meteorological data at time t within the 108th day within the preset time period, that is, the meteorological data at the specified time t in the target area.

[0056] It should be noted that the initial meteorological data estimation model may include a three-layer structure of a convolutional layer, a pooling layer, and a fully connected layer. Before training, NWP meteorological data within a preset time period can be collected and divided into training set data and test set data in a ratio of 8:2. The training set data can be used to train the network model, and the test set data can be used to verify the fitting effect of the network model. During training, the time dimension L of the initialized meteorological characteristic data can be used as the training set data. in A set of network parameters (including learning factors, number of hidden layer neurons, convolution kernel size, etc.) are input into the initial meteorological data estimation model. The network model is trained by continuously adjusting the network structure to obtain an optimal network model. Then, the traversal method is used to make the network model's fitting accuracy and training speed optimal, and the network parameters at this time are determined to be the optimal network parameters. Finally, after fixing the optimal network model and optimal network parameters, the time dimension L of the meteorological characteristic data is calculated. in Optimize and take the time dimension L of the meteorological characteristic data when the test set data fitting accuracy is the largest in is the optimal time dimension L e The time dimension L of the above meteorological characteristic data in The number of categories of NWP meteorological data contained in the meteorological characteristic data can be characterized.

[0057] In addition, in order to make the meteorological characteristic data and the target meteorological estimated data in the same distribution interval, before inputting the meteorological characteristic data into the target meteorological data estimation model, the computer device can first normalize the meteorological characteristic data, and further input the normalized meteorological characteristic data into the target meteorological data estimation model for further processing. The specific calculation formula of the normalization processing is shown in the following formula (1), that is:

[0058]

[0059] Among them, y represents the normalized meteorological characteristic data, x represents the meteorological characteristic data, and the variables with max and min subscripts are the maximum and minimum values ​​in the meteorological characteristic data respectively. In this embodiment, the meteorological characteristic data needs to be normalized to the interval [-1, 1], so y max can be equal to 1, y min Can be equal to -1.

[0060] It should be noted that the meteorological characteristic data input into the target meteorological estimation model each time can be composed of the meteorological data of all known points in the grid points at a preset number of N moments before and after the specified time t, including G k ·T NC meaningful elements. The target meteorological calculation data output by the target meteorological calculation model is the direct irradiance data of all unknown points to be fitted at the specified time t. The output target meteorological calculation data includes G k Elements. N The value cannot exceed Γ, T N The target meteorological estimation model uses T before and after time t N The meteorological data at time a is used to fit the meteorological data at time t.

[0061] For example, the target area contains 5 rows and 5 columns of NWP grid points, there are 15 known points in the grid points, 10 unknown points to be fitted, the time resolution when collecting meteorological data is 1 hour, and the daily irradiance data in a specified day contains 12 points (that is, one point in 1 hour in 12 hours of daytime in a specified day). The meteorological characteristic data input into the target meteorological estimation model includes three types of direct irradiance, diffuse irradiance and temperature data. For example, the single-point fitting uses the meteorological data of 7 times around time t (i.e., direct irradiance, diffuse irradiance and temperature data) as meteorological characteristic data. The values ​​of the above symbols in this example are shown in Table 1, and C in Table 1 represents the number of meteorological data categories in the meteorological characteristic data.

[0062] Table 1

[0063]

[0064] Furthermore, in order to adapt to the input form of the three-dimensional convolutional neural network model, it is necessary to adjust the meaningful elements contained in the above-mentioned input matrix (i.e., meteorological characteristic data).

[0065] Among them, the three-dimensional convolutional neural network requires (C in ,W in ,H in ,L in ) form, where C in is the number of channels, which has the same meaning as C above. in ,H in Represents the spatial characteristics of the input matrix, and its values ​​are the number of rows and columns of grid points in the target area. in Indicates the time characteristics of the input matrix, which has the same meaning as T above. In single-point fitting, it is the same as T above. N The meaning is the same. In addition, (C in ,W in ,H in ,L in ) can be a sparse matrix containing many zero elements, where the only meaningful elements are G k ·L in ·C in indivual.

[0066] If G k , C in , C, T, T N The value of is the same as that in Table 1. The target area contains 5 rows and 5 columns of NWP grid points. The meteorological characteristic data of the target meteorological estimation model based on the three-dimensional convolutional neural network can be a four-dimensional matrix of (3, 5, 5, 7), and the output target meteorological estimation data can be a vector of (10, 1).

[0067] The above-mentioned meteorological data processing method pre-processes the initial meteorological data at each sampling moment within a preset time period corresponding to all grid points in the target area, obtains the direct irradiance data of all grid points at each sampling moment, and the index meteorological data related to the direct irradiance, and selects the direct irradiance data and index meteorological data corresponding to a preset number of moments before and after the specified moment of the known points adjacent to the unknown points in the grid points from the direct irradiance data and the index meteorological data as meteorological characteristic data, and obtains the target meteorological estimated data of the unknown points at the specified moment through the meteorological characteristic data based on the target meteorological data extrapolation model; the above-mentioned method can process initial meteorological data of different qualities, so as to further extrapolate the meteorological data of the unknown points in the grid points through the initial meteorological data, thereby improving the accuracy of the unknown points extrapolation results; at the same time, the above-mentioned method can be applied to initial meteorological data of different qualities, thereby improving the robustness of the extrapolation method.

[0068] In one of the embodiments, the process of pre-processing the initial meteorological data at each sampling moment within the preset time period corresponding to the grid points in the target area in the above step S1000, obtaining the direct irradiance data of all grid points at each sampling moment, and the index meteorological data related to the direct irradiance, may include: using an outlier processing method to perform outlier processing on the initial meteorological data at each sampling moment within the preset time period corresponding to the grid points in the target area, and obtaining the direct irradiance data of the grid points at each sampling moment after outlier processing, and the index meteorological data related to the direct irradiance.

[0069] Specifically, the computer equipment can use a detection method such as a 3-sigma method or a correlation factor analysis method to detect whether there is abnormal data in the initial meteorological data; if abnormal data is detected in the initial meteorological data, the initial meteorological data can be processed by an abnormal value processing method to obtain the direct irradiance data of the grid point at each sampling time in the initial meteorological data after abnormal processing, and the index meteorological data related to the direct irradiance. Usually, there will be abnormal data in the initial meteorological data.

[0070] It should be noted that if there are fewer abnormal data in the initial meteorological data, the computer equipment can directly eliminate the abnormal values, and obtain the direct irradiance data of the grid point at each sampling time in the initial meteorological data after the abnormality processing, as well as the index meteorological data related to the direct irradiance, thereby reducing the amount of calculation; if there are more abnormal data in the initial meteorological data, the computer equipment can use interpolation method or average filling method and other processing methods to replace the abnormal data in the initial meteorological data, and obtain the direct irradiance data of the grid point at each sampling time in the initial meteorological data after the abnormality processing, as well as the index meteorological data related to the direct irradiance.

[0071] Among them, the process of pre-processing the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area in the above step S1000, obtaining the direct irradiance data of the grid point at each sampling moment, and the index meteorological data related to the direct irradiance, may also include: using the interpolation method to interpolate the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, and obtaining the direct irradiance data of the grid point at each sampling moment after the interpolation processing, and the index meteorological data related to the direct irradiance.

[0072] At the same time, the computer equipment can also use the query method to detect whether there is missing data in the initial meteorological data; if it is detected that there is missing data in the initial meteorological data, the initial meteorological data can be interpolated by using interpolation methods and other processing methods to obtain the direct irradiance data of the grid point at each sampling time after interpolation, as well as the index meteorological data related to the direct irradiance. Usually, there will be missing data in the initial meteorological data.

[0073] In this embodiment, if there are only abnormal data in the initial meteorological data, the computer device can only perform abnormal processing on the initial meteorological data to obtain the direct irradiance data of the grid point at each sampling time after abnormal processing, and the index meteorological data related to the direct irradiance; if there are only missing data in the initial meteorological data, the computer device can only perform missing value processing (i.e., interpolation processing) on ​​the initial meteorological data to obtain the direct irradiance data of the grid point at each sampling time after interpolation processing, and the index meteorological data related to the direct irradiance; if there are both abnormal data and missing data in the initial meteorological data, at this time, the computer device can perform abnormal processing and missing value processing on the initial meteorological data to obtain the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance. Among them, the execution order of abnormal processing and missing value processing can be interchanged, and there is no limitation on this execution order. In this embodiment, the missing values ​​in the initial meteorological data are the meteorological data missing due to the failure of the communication process of the meteorological data collection device at the grid point where the meteorological data collection device has been set in the target area, and most of them belong to small segments of data missing.

[0074] The above-mentioned meteorological data processing method can pre-process the initial meteorological data to obtain direct irradiance data and index meteorological data, and then process the direct irradiance data and index meteorological data to infer the target meteorological calculation data; the above-mentioned method can pre-process the initial meteorological data of different qualities to improve the data quality, thereby improving the accuracy of the calculation results of unknown points in the grid; at the same time, the above-mentioned method can be applied to initial meteorological data of different qualities, thereby improving the robustness of the calculation method.

[0075] As one of the embodiments, Figure 3 As shown, the process of selecting the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of the known points adjacent to the unknown point in the grid points in the above step S2000 as meteorological characteristic data may include the following steps:

[0076] Step S2100: performing calculation processing on the direct irradiance data and the index meteorological data to obtain a correlation coefficient between the direct irradiance data and the index meteorological data on each designated day within the preset time period.

[0077] Specifically, the above-mentioned operation processing can be at least one of addition, subtraction, multiplication and division. The correlation coefficient between the above-mentioned direct irradiance data and the index meteorological data can reflect the correlation between the direct irradiance data and the index meteorological data. Optionally, the specified day can be understood as one day, that is, 24 hours, so the preset time period can include multiple specified days. Each specified day has corresponding direct irradiance data and index meteorological data, so that the correlation coefficient between the direct irradiance data and the index meteorological data on each specified day in the preset time period can be calculated.

[0078] It should be noted that the direct irradiance data and index meteorological data processed by calculation can be the direct irradiance data and index meteorological data at the same time and the same NWP meteorological data point for each specified day; they can also be the direct irradiance data and index meteorological data at the same time and different NWP meteorological data points for each specified day; of course, they can also be the direct irradiance data and index meteorological data at different times and different NWP meteorological data points for each specified day.

[0079] Among them, the above-mentioned step S2100 performs operation processing on the direct irradiance data and the index meteorological data to obtain the correlation coefficient between the direct irradiance data and the index meteorological data on each specified day within the preset time period, which may specifically include: performing operation processing on the direct irradiance data and the index meteorological data, calculating the Pearson correlation coefficient between the direct irradiance data and the index meteorological data, and using the Pearson correlation coefficient as the correlation coefficient; the operation processing includes a combination of addition, subtraction, multiplication and division operations.

[0080] Specifically, the computer device can perform calculation processing on the direct irradiance data and the scattered irradiance data to obtain the correlation coefficient between the two. In this embodiment, the computer device can use the correlation coefficient calculation method to perform calculation processing on the direct irradiance data and the scattered irradiance data to obtain the correlation coefficient between the two. Generally, the computer device can calculate the correlation coefficient between two different meteorological data on the same date.

[0081] It should be noted that the above correlation coefficient calculation method can be a calculation method such as the Jaccard similarity coefficient method, the Picasso correlation coefficient method or the Pearson correlation coefficient method; the above operation processing can include a combination of addition, subtraction, multiplication and division operations.

[0082] Understandably, X ij,d represents the direct irradiance data of the dth day at the spatial position of the i-th row and j-th column in the target area, Y ij,drepresents the index meteorological data (i.e., diffuse irradiance data, temperature data, wind speed data, and humidity data) of the dth day in the historical time period at the spatial location of the i-th row and the j-th column in the target area. The Pearson correlation coefficient ρ can be expressed by the following formula (2), namely:

[0083]

[0084] Wherein, Cov(·) represents covariance, var(·) represents variance; the above correlation coefficient can reflect the correlation between direct irradiance data and index meteorological data on the same date (i.e., the dth day in the preset time period) and the same spatial position (i.e., the spatial position of the i-th row and the j-th column); d can represent the estimated date in the historical time period. For example, the preset time period is 1 year, i.e., 365 days. In this embodiment, the meteorological data within the 108th day can be estimated. At this time, d can be equal to 108.

[0085] Step S2200, selecting, through the correlation coefficient, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to the unknown point in the grid point as meteorological characteristic data.

[0086] Specifically, the computer device can obtain the correlation coefficient between the direct irradiance data and the index meteorological data on each specified day, and select the direct irradiance data and the index meteorological data corresponding to the known points adjacent to the unknown points in the grid points at a preset number of times before and after the specified time as meteorological characteristic data, so as to serve as the input data of the target meteorological data extrapolation model. Among them, the meteorological data (i.e., direct irradiance data and index meteorological data) corresponding to the maximum correlation coefficient and the minimum correlation coefficient in the correlation coefficient corresponding to each specified day are combined, and the combined result can be determined as meteorological characteristic data; the direct irradiance data and the index meteorological data corresponding to multiple interval correlation coefficients in the correlation coefficient corresponding to each specified day are combined, and the combined result can also be determined as meteorological characteristic data; of course, the combination of direct irradiance data and index meteorological data corresponding to the correlation coefficient corresponding to each specified day can also be determined as meteorological characteristic data.

[0087] The above-mentioned meteorological data processing method can calculate the correlation coefficient between the direct irradiance data and the index meteorological data, and through the correlation coefficient, select the direct irradiance data and index meteorological data corresponding to a preset number of time periods before and after the specified time period of the known points adjacent to the unknown points in the grid points as meteorological characteristic data, and then calculate the target meteorological calculation data based on the target meteorological calculation model; the above-mentioned method can pre-process the initial meteorological data of different qualities to improve the data quality, thereby improving the accuracy of the calculation results of the unknown points in the grid; at the same time, the above-mentioned method can be applied to initial meteorological data of different qualities, thereby improving the robustness of the calculation method.

[0088] In one embodiment, Figure 4 As shown, in the above step S2200, the process of selecting the direct irradiance data and the index meteorological data corresponding to a preset number of time periods before and after a specified time period of the known points adjacent to the unknown point in the grid points through the correlation coefficient as the meteorological characteristic data can be achieved by the following steps:

[0089] Step S2210: For each specified day, the correlation coefficients corresponding to each type of indicator meteorological data are averaged and the absolute values ​​are taken to obtain the absolute average correlation coefficient.

[0090] Specifically, the computer device can perform combined operation processing on all correlation coefficients between each type of indicator meteorological data and direct irradiance data on each specified day to obtain an absolute average correlation coefficient. The above combined operation processing can be a combined operation of arithmetic operation processing and absolute value processing; however, in this embodiment, the above combined operation processing can include averaging and taking absolute value processing.

[0091] In this embodiment, the computer device can average all the correlation coefficients between each type of index meteorological data and direct irradiance data on each specified day, and then take the absolute value of the average value to obtain the absolute average correlation coefficient. Taking the index meteorological data as diffuse irradiance data for illustration, the computer device can average all the correlation coefficients between direct irradiance data and diffuse irradiance data on each specified day, and then take the absolute value of the average value to obtain the absolute average correlation coefficient. At the same time, the computer device can also calculate the absolute average correlation coefficients corresponding to the temperature data, wind speed data and humidity data respectively. Generally, the value range of the absolute average correlation coefficient can be between 0 and 1.

[0092] Step S2220: Determine a strong correlation coefficient and a weak correlation coefficient according to the absolute average correlation coefficient.

[0093] Specifically, the computer device may determine the strong correlation coefficient and the weak correlation coefficient among all the absolute average correlation coefficients.

[0094] Among them, the process of determining the strong correlation coefficient and the weak correlation coefficient according to the absolute average correlation coefficient in the above step S2220 may specifically include: comparing each absolute average correlation coefficient with a preset correlation threshold, and determining the strong correlation coefficient and the weak correlation coefficient according to the comparison result.

[0095] In this embodiment, the preset correlation threshold may be 0.8 and / or 0.3. The computer device may compare the absolute average correlation coefficient with 0.8, and determine the absolute average correlation coefficient greater than 0.8 in the comparison result as a strong correlation coefficient; at the same time, the computer device may also compare the absolute average correlation coefficient with 0.8, and then compare the absolute average correlation coefficient less than 0.8 in the comparison result with 0.3, and determine the absolute average correlation coefficient less than 0.8 and greater than 0.3 in the comparison result as a weak correlation coefficient. In this embodiment, the absolute average correlation coefficient less than 0.3 can be characterized as no correlation.

[0096] Step S2230: Obtain the direct irradiance data and the index meteorological data corresponding to the strong correlation coefficient and / or the weak correlation coefficient.

[0097] Specifically, the computer device can use all the direct irradiance data and index meteorological data corresponding to the strong correlation coefficient as meteorological characteristic data; and / or according to actual conditions, the direct irradiance data and index meteorological data corresponding to the weak correlation coefficient can also be used as meteorological characteristic data. The above-mentioned actual conditions may include whether the initial meteorological data corresponding to the weak correlation coefficient is sufficient, whether it is easy to obtain, and whether the acquisition cost is low; if all three conditions are yes, the direct irradiance data and index meteorological data corresponding to the weak correlation coefficient can be used as meteorological characteristic data; if one of the three conditions is no, the direct irradiance data and index meteorological data corresponding to the weak correlation coefficient cannot be used as meteorological characteristic data.

[0098] Step S2240, from the direct irradiance data and the index meteorological data, select the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to the unknown point in the grid point as the meteorological characteristic data.

[0099] Specifically, the computer device can select the direct irradiance data and the index meteorological data corresponding to the known points adjacent to the unknown point in the grid at a preset number of times before and after the specified time from the direct irradiance data and the index meteorological data corresponding to the strong correlation coefficient and / or the weak correlation coefficient as the meteorological characteristic data.

[0100] If all grid points in the target area form a 5*5 rectangular array, where the point at the 3rd row and 3rd column is an unknown point, the known points adjacent to the unknown point in the grid points include: the points at the 2nd row and 2nd column, the 2nd row and 3rd column, the 2nd row and 4th column, the 3rd row and 2nd column, the 3rd row and 4th column, the 4th row and 2nd column, the 4th row and 3rd column, and the 4th row and 4th column.

[0101] In this embodiment, the above-mentioned designated time can be understood as any time of a day within the preset time period; the above-mentioned preset number can be any number. n The meteorological data of N moments before and after can be respectively n-2 , time t n-1 , time t n+1 , time t n+2 ; where the specified time t n Multiple moments before and a specified time t n The multiple moments after can be equal or unequal, as long as the specified time t n The screened meteorological data can be used before and after. Optionally, the screened direct irradiance data and the index meteorological data can be used as meteorological characteristic data. Optionally, the meteorological characteristic data and the target meteorological estimated data can be meteorological data corresponding to the same specified day.

[0102] The above-mentioned meteorological data processing method can obtain the correlation coefficient between direct irradiance data and meteorological data of different indicators, and then select the meteorological characteristic data of the target meteorological data extrapolation model through the correlation coefficient, and obtain the target meteorological extrapolation data through the meteorological characteristic data based on the target meteorological data extrapolation model; the above-mentioned method can process initial meteorological data of different qualities, further extrapolate the meteorological data of unknown grid points, and improve the accuracy of the extrapolation results of unknown grid points.

[0103] It should be understood that although Figure 1 , 3 The steps in the flowcharts of and 4 are shown in sequence as indicated by arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 , 3 At least part of the steps in 4 may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0104] In one embodiment, Figure 5 As shown, a meteorological data processing device is provided, which includes: a preprocessing module 11, a data screening module 12 and a calculation module 13.

[0105] The preprocessing module 11 is used to preprocess the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, and obtain the direct irradiance data of the grid point at each sampling time, as well as the index meteorological data related to the direct irradiance;

[0106] The data screening module 12 is used to screen out the direct irradiance data and the index meteorological data from the direct irradiance data and the index meteorological data, as meteorological characteristic data, the direct irradiance data and the index meteorological data corresponding to the known points adjacent to the unknown points in the grid points at a preset number of time moments before and after the specified time moment;

[0107] The calculation module 13 is used to calculate the target meteorological data based on the target meteorological data calculation model, and obtain the target meteorological calculation data of the unknown point at the specified time through the meteorological characteristic data; the target meteorological data calculation model is a model obtained by training with meteorological characteristic data extracted from historical meteorological data as sample data.

[0108] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0109] In one embodiment, the pre-processing module 11 includes: an exception handling unit.

[0110] Among them, the exception processing unit is used to adopt an outlier processing method to perform exception processing on the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, and obtain the direct irradiance data of the grid point at each sampling moment after exception processing, as well as the index meteorological data related to the direct irradiance.

[0111] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0112] In one embodiment, the preprocessing module 11 further includes: an interpolation processing unit.

[0113] Among them, the interpolation processing unit is used to adopt the interpolation method to interpolate the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, and obtain the direct irradiance data of the grid point at each sampling moment after interpolation processing, as well as the index meteorological data related to the direct irradiance.

[0114] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0115] In one embodiment, the data screening module 12 includes: a calculation unit and a selection unit.

[0116] The computing unit is used to perform computing processing on the direct irradiance data and the index meteorological data to obtain a correlation coefficient between the direct irradiance data and the index meteorological data on each designated day within the preset time period;

[0117] The selection unit is used to select, through the correlation coefficient, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid points as meteorological characteristic data.

[0118] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0119] In one of the embodiments, the operation unit is specifically used to perform operation processing on the direct irradiance data and the index meteorological data, calculate the Pearson correlation coefficient between the direct irradiance data and the index meteorological data, and use the Pearson correlation coefficient as the correlation coefficient; the operation processing includes a combination of addition, subtraction, multiplication and division operations.

[0120] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0121] In one of the embodiments, the selection unit includes: a calculation subunit, a determination subunit, an acquisition subunit and a selection subunit.

[0122] The calculation subunit is used to average and take absolute value of the correlation coefficients corresponding to each type of index meteorological data on each designated day to obtain an absolute average correlation coefficient;

[0123] The determination subunit is used to determine a strong correlation coefficient and a weak correlation coefficient according to the absolute average correlation coefficient;

[0124] The acquisition subunit is used to acquire the direct irradiance data and the index meteorological data corresponding to the strong correlation coefficient and / or the weak correlation coefficient;

[0125] The selection subunit is used to select, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as the meteorological characteristic data.

[0126] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0127] In one of the embodiments, the calculation subunit is specifically used to compare each absolute average correlation coefficient with a preset correlation threshold value, and determine the strong correlation coefficient and the weak correlation coefficient according to the comparison result.

[0128] The meteorological data processing device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0129] For the specific definition of the meteorological data processing device, please refer to the definition of the meteorological data processing method above, which will not be repeated here. Each module in the above-mentioned meteorological data processing device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0130] In one embodiment, a computer device is provided. The computer device may be a server. The internal structure diagram thereof may be further described in detail in FIG. Figure 2 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the collected meteorological data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a meteorological data processing method is implemented.

[0131] Those skilled in the art will understand that Figure 2 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0132] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0133] Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, obtaining the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance;

[0134] Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0135] Based on the target meteorological data extrapolation model, the target meteorological extrapolation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data extrapolation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0136] In one embodiment, a readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0137] Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, obtaining the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance;

[0138] Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data;

[0139] Based on the target meteorological data extrapolation model, the target meteorological extrapolation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data extrapolation model is a model trained by using the meteorological characteristic data extracted from the historical meteorological data as sample data.

[0140] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0141] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0142] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A meteorological data processing method, characterized in that: The method comprises: Preprocessing the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area to obtain the direct irradiance data of the grid point at each sampling time, and the index meteorological data related to the direct irradiance, wherein the preprocessing includes at least one of feature encoding processing, feature selection processing, normalization processing, interpolation processing, average filling processing, and outlier processing; Filtering out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data; Based on the target meteorological data estimation model, the target meteorological estimation data of the unknown point at the specified time is obtained through the meteorological characteristic data; the target meteorological data estimation model is a model obtained by training with the meteorological characteristic data extracted from the historical meteorological data as sample data, and the training process of the initial meteorological data estimation model corresponding to the target meteorological data estimation model includes: The time dimension of the initialized meteorological characteristic data and a set of network parameters are input into the initial meteorological data extrapolation model, and the network model is trained by continuously adjusting the network structure to obtain the optimal network model; the traversal method is used to determine the optimal fitting accuracy and the optimal training speed of the network model, and the network parameters corresponding to the optimal fitting accuracy and the optimal training speed are determined as the optimal network parameters; the time dimension of the meteorological characteristic data is optimized, and when the fitting accuracy of the test set data is the largest, the time dimension of the current meteorological characteristic data is the optimal time dimension; the time dimension of the meteorological characteristic data represents the number of classes of numerical weather forecast meteorological data contained in the meteorological characteristic data.

2. The method according to claim 1, characterized in that The initial meteorological data at each sampling time in the preset time period corresponding to the grid points in the target area are preprocessed to obtain the direct irradiance data of all grid points at each sampling time, as well as the index meteorological data related to the direct irradiance, including: An outlier processing method is adopted to perform exception processing on the initial meteorological data at each sampling moment within the preset time period corresponding to the grid point in the target area, and the direct irradiance data of the grid point at each sampling moment after exception processing and the index meteorological data related to the direct irradiance are obtained.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: The interpolation method is used to interpolate the initial meteorological data at each sampling moment within the preset time period corresponding to the grid points in the target area, and obtain the direct irradiance data of the grid points at each sampling moment after interpolation, as well as the index meteorological data related to the direct irradiance.

4. The method according to claim 1, characterized in that: The direct irradiance data and the index meteorological data are selected from the direct irradiance data and the index meteorological data, and the direct irradiance data and the index meteorological data corresponding to the known points adjacent to the unknown point in the grid points at a preset number of time points before and after the specified time point are selected as meteorological characteristic data, including: Performing calculation processing on the direct irradiance data and the index meteorological data to obtain a correlation coefficient between the direct irradiance data and the index meteorological data on each designated day within the preset time period; The direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point are selected as meteorological characteristic data through the correlation coefficient.

5. The method according to claim 4, characterized in that The performing of calculation processing on the direct irradiance data and the index meteorological data to obtain a correlation coefficient between the direct irradiance data and the index meteorological data on each designated day within the preset time period includes: The direct irradiance data and the index meteorological data are subjected to operation processing, the Pearson correlation coefficient between the direct irradiance data and the index meteorological data is calculated, and the Pearson correlation coefficient is used as the correlation coefficient; the operation processing includes a combination of addition, subtraction, multiplication and division.

6. The method according to claim 4, characterized in that The direct irradiance data and the index meteorological data corresponding to a preset number of time periods before and after a specified time period of a known point adjacent to the unknown point in the grid points are selected through the correlation coefficient as meteorological characteristic data, including: For each specified day, the correlation coefficients corresponding to each type of index meteorological data are averaged and the absolute value is taken to obtain the absolute average correlation coefficient; Determining a strong correlation coefficient and a weak correlation coefficient according to the absolute average correlation coefficient; Acquire the direct irradiance data and the index meteorological data corresponding to the strong correlation coefficient and / or the weak correlation coefficient; From the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid are selected as the meteorological characteristic data.

7. The method according to claim 6, characterized in that Determining a strong correlation coefficient and a weak correlation coefficient according to the absolute average correlation coefficient includes: Each absolute average correlation coefficient is compared with a preset correlation threshold value, and the strong correlation coefficient and the weak correlation coefficient are determined according to the comparison result.

8. A meteorological data processing device, characterized in that: The device comprises: A preprocessing module is used to preprocess the initial meteorological data at each sampling time within a preset time period corresponding to the grid point in the target area, and obtain the direct irradiance data of the grid point at each sampling time, as well as the index meteorological data related to the direct irradiance, wherein the preprocessing includes at least one of feature coding processing, feature selection processing, normalization processing, interpolation processing, average filling processing, and outlier processing; A data screening module, used to screen out, from the direct irradiance data and the index meteorological data, the direct irradiance data and the index meteorological data corresponding to a preset number of time moments before and after a specified time moment of a known point adjacent to an unknown point in the grid point as meteorological characteristic data; The calculation module is used to obtain the target meteorological calculation data of the unknown point at the specified time through the meteorological characteristic data based on the target meteorological data calculation model; the target meteorological data calculation model is a model obtained by training with the meteorological characteristic data extracted from the historical meteorological data as sample data, and the training process of the initial meteorological data calculation model corresponding to the target meteorological data calculation model includes: The time dimension of the initialized meteorological characteristic data and a set of network parameters are input into the initial meteorological data extrapolation model, and the network model is trained by continuously adjusting the network structure to obtain the optimal network model; the traversal method is used to determine the optimal fitting accuracy and the optimal training speed of the network model, and the network parameters corresponding to the optimal fitting accuracy and the optimal training speed are determined as the optimal network parameters; the time dimension of the meteorological characteristic data is optimized, and when the fitting accuracy of the test set data is the largest, the time dimension of the current meteorological characteristic data is the optimal time dimension; the time dimension of the meteorological characteristic data represents the number of classes of numerical weather forecast meteorological data contained in the meteorological characteristic data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images