Photovoltaic power generation power prediction method and system

Abstract

The invention provides a photovoltaic power generation power prediction method and system, and relates to the technical field of power prediction, and the method comprises the steps: obtaining photovoltaic power generation power prediction data; extracting data features, determining prediction influence data features based on photovoltaic power generation power prediction influence factors, calibrating prediction main data features, and determining prediction influence values of the prediction influence data features on the prediction main data features; obtaining prediction parameters, constructing a photovoltaic power generation power prediction model, inputting current photovoltaic power generation power prediction data, and obtaining a photovoltaic power generation power prediction value; setting a photovoltaic power generation power prediction evaluation index, determining an evaluation value of a photovoltaic power generation power prediction value in combination with the prediction influence value, and when the evaluation value of the photovoltaic power generation power prediction value is not lower than a preset value, determining that the photovoltaic power generation power prediction value is accurate; and accurate prediction of the photovoltaic power generation power is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power prediction, and in particular to a photovoltaic power generation power prediction method and system. Background Art

[0002] With the development of new energy, the scale of photovoltaic power generation continues to expand. The output of photovoltaic power generation has the characteristic of unstable output, which seriously affects the stability and safety of power grid operation. It is necessary to accurately predict the photovoltaic power generation and then effectively utilize it. However, there are many factors affecting photovoltaic power generation, and data collection is prone to a large amount of abnormal data and missing data, which seriously affects the prediction accuracy and efficiency of photovoltaic power generation.

[0003] Therefore, the present invention provides a photovoltaic power generation power prediction method and system. Summary of the invention

[0004] The present invention provides a photovoltaic power prediction method and system, which obtain photovoltaic power prediction data; extract data features, determine prediction influencing data features based on photovoltaic power prediction influencing factors, calibrate prediction main data features and determine the prediction influence value of the prediction influencing data features on the prediction main data features; obtain prediction parameters, build a photovoltaic power prediction model, input current photovoltaic power prediction data, and obtain photovoltaic power prediction values; set photovoltaic power prediction evaluation indicators and determine the evaluation value of the photovoltaic power prediction value in combination with the prediction influence value, and when the evaluation value of the photovoltaic power prediction value is not lower than a preset value, determine that the photovoltaic power prediction value is accurate; and achieve accurate prediction of photovoltaic power.

[0005] The present invention provides a photovoltaic power generation power prediction method, comprising:

[0006] Step 1: Obtain historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to obtain photovoltaic power generation prediction data;

[0007] Step 2: Extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features;

[0008] Step 3: Obtain prediction parameters, build a photovoltaic power generation prediction model, input current photovoltaic power generation prediction data, and obtain photovoltaic power generation prediction value;

[0009] Step 4: Set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, the photovoltaic power generation prediction value is determined to be accurate.

[0010] According to a photovoltaic power prediction method provided by the present invention, historical photovoltaic power data, meteorological data, photovoltaic power generation equipment data and location data are obtained and preprocessed to obtain photovoltaic power prediction data, including:

[0011] Obtain historical photovoltaic power data, meteorological data, photovoltaic equipment data and location data, perform data preprocessing according to the data type, divide variable data and fixed data and store the data;

[0012] The variable data is subjected to secondary data preprocessing according to the preset data preprocessing strategy to obtain calibrated variable data, which are then combined with the fixed data to determine the photovoltaic power generation power prediction data.

[0013] According to a photovoltaic power prediction method provided by the present invention, secondary data preprocessing is performed on variable data according to a preset data preprocessing strategy, including:

[0014] Mark abnormal data and missing data of variable data, and perform secondary data preprocessing on the abnormal data and missing data according to a preset data preprocessing strategy, wherein the preset data preprocessing strategy includes: an abnormal value detection data preprocessing strategy and an interpolation data preprocessing strategy;

[0015] Data normalization is performed on variable data of different dimensions.

[0016] According to a photovoltaic power generation prediction method provided by the present invention, data features are extracted according to photovoltaic power generation prediction data, and prediction influencing data features are determined based on photovoltaic power generation prediction influencing factors, including:

[0017] Extract data features of photovoltaic power generation prediction data according to time series or space series to obtain photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features and location data features;

[0018] Obtaining factors influencing photovoltaic power generation prediction, performing data feature screening on the photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features, and location data features, and determining prediction-influencing data features and prediction-irrelevant data features;

[0019] When a type of data feature has data correlation with a photovoltaic power generation data feature, the type of data feature is determined to be a prediction-influencing data feature;

[0020] When there is no data correlation between a type of data feature and the photovoltaic power generation data feature, the type of data feature is determined to be a prediction irrelevant data feature.

[0021] According to a photovoltaic power generation prediction method provided by the present invention, the prediction main data feature is calibrated and the prediction influence value of the prediction influence data feature on the prediction main data feature is determined, including:

[0022] The data features are sorted according to the data correlation quantification values ​​of various data features and photovoltaic power generation data features, and the data feature with the largest data correlation quantification value is marked as the prediction main data feature;

[0023] Perform one-way data correlation analysis on the prediction influencing data features and the prediction master data features to determine the prediction impact value of the prediction influencing data features on the prediction master data features.

[0024] According to a photovoltaic power generation prediction method provided by the present invention, prediction parameters are obtained and a photovoltaic power generation prediction model is constructed, including:

[0025] Select a preset number of prediction influencing data features, and obtain prediction parameters in combination with the photovoltaic power generation power prediction demand, wherein the prediction parameters include: prediction time parameters, prediction space parameters, and prediction influencing factor parameters;

[0026] Construct a photovoltaic power generation prediction model based on the prediction parameters and the preset machine learning model;

[0027] A data training set is established based on the photovoltaic power generation prediction data, and a photovoltaic power generation prediction model is trained.

[0028] According to a photovoltaic power generation prediction method provided by the present invention, current photovoltaic power generation prediction data is input to obtain a photovoltaic power generation prediction value, including:

[0029] Obtain the current photovoltaic power generation prediction data, input it into the photovoltaic power generation prediction model after model training, and obtain the photovoltaic power generation prediction value;

[0030] When the photovoltaic power generation power prediction value is determined to be accurate, the photovoltaic power generation power prediction value is visualized according to a preset display strategy.

[0031] According to a photovoltaic power generation power prediction method provided by the present invention, setting a photovoltaic power generation power prediction evaluation index includes:

[0032] Photovoltaic power generation prediction evaluation indicators are set, and the influence weight of the photovoltaic power generation prediction evaluation indicators on the evaluation value of the photovoltaic power generation prediction value is set according to the photovoltaic power generation power prediction demand, wherein the photovoltaic power generation power prediction evaluation indicators include: photovoltaic power generation power prediction accuracy evaluation indicators, photovoltaic power generation power prediction timeliness evaluation indicators and photovoltaic power generation power prediction stability evaluation indicators.

[0033] According to a photovoltaic power prediction method provided by the present invention, an evaluation value of the photovoltaic power prediction value is determined in combination with the prediction influence value, and when the evaluation value of the photovoltaic power prediction value is not lower than a preset value, it is determined that the photovoltaic power prediction value is accurate, including:

[0034] Determine an evaluation value of the photovoltaic power generation prediction value according to the photovoltaic power generation prediction evaluation index combined with the prediction impact value;

[0035]

[0036] Among them, F represents the evaluation value of the photovoltaic power generation prediction value; xi1 represents the i1th prediction influencing data feature, x1 represents the average value of the prediction influencing data feature; yi1 represents the i1th prediction main data feature; y1 represents the average value of the prediction main data feature; represents the predicted influence value of the i1th prediction influencing data feature on the predicted main data feature; zj1 represents the j1th photovoltaic power prediction value; z1j1 represents the actual photovoltaic power value corresponding to the j1th photovoltaic power prediction value; z1 represents the average value of the photovoltaic power prediction value; tj1 represents the prediction time of the j1th photovoltaic power prediction value; t1 represents the standard photovoltaic power prediction time; n1 represents the number of data features; m1 represents the number of photovoltaic power prediction values; a1 represents the influence weight of the photovoltaic power prediction accuracy evaluation index on the evaluation value; a2 represents the influence weight of the photovoltaic power prediction timeliness evaluation index on the evaluation value; a3 represents the influence weight of the photovoltaic power prediction stability evaluation index on the evaluation value;

[0037] When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate;

[0038] When the evaluation value of the photovoltaic power generation prediction value is lower than the preset value, the photovoltaic power generation prediction value is determined to be inaccurate, and the prediction impact data features are replaced one by one according to the prediction impact value, the photovoltaic power generation prediction model is rebuilt and the photovoltaic power generation prediction value is obtained.

[0039] The present invention provides a photovoltaic power generation power prediction system, comprising:

[0040] A data acquisition module is used to acquire historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to acquire photovoltaic power generation prediction data;

[0041] A data analysis module is used to extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features;

[0042] The power prediction module is used to obtain prediction parameters, build a photovoltaic power prediction model, input prediction main data features and prediction influencing data features, and obtain the photovoltaic power prediction value;

[0043] The prediction verification module is used to set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] By acquiring photovoltaic power prediction data; extracting data features, determining prediction influencing data features based on photovoltaic power prediction influencing factors, calibrating prediction main data features and determining the prediction influence values ​​of prediction influencing data features on prediction main data features; acquiring prediction parameters, building a photovoltaic power prediction model, inputting current photovoltaic power prediction data, and acquiring photovoltaic power prediction values; setting photovoltaic power prediction evaluation indicators and determining the evaluation value of photovoltaic power prediction values ​​in combination with the prediction influence values, when the evaluation value of photovoltaic power prediction values ​​is not lower than a preset value, determining that the photovoltaic power prediction values ​​are accurate; and achieving accurate prediction of photovoltaic power.

[0046] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0047] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0049] Figure 1It is a schematic diagram of a process of a photovoltaic power generation prediction method provided by an embodiment of the present invention;

[0050] Figure 2 It is a structural schematic diagram of a photovoltaic power generation prediction system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0052] Embodiment 1:

[0053] The embodiment of the present invention provides a photovoltaic power generation power prediction method, such as Figure 1 As shown, including:

[0054] Step 1: Obtain historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to obtain photovoltaic power generation prediction data;

[0055] Step 2: Extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features;

[0056] Step 3: Obtain prediction parameters, build a photovoltaic power generation prediction model, input current photovoltaic power generation prediction data, and obtain photovoltaic power generation prediction value;

[0057] Step 4: Set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, the photovoltaic power generation prediction value is determined to be accurate.

[0058] In this embodiment, the photovoltaic power generation power prediction influencing factors refer to factors that affect the photovoltaic power generation power prediction.

[0059] In this embodiment, the prediction main data feature refers to the main data feature for photovoltaic power generation prediction.

[0060] In this embodiment, the prediction influencing data features refer to the influencing data features for photovoltaic power generation prediction.

[0061] In this embodiment, the prediction parameters include: prediction time parameters, prediction space parameters and prediction influencing factor parameters.

[0062] In this embodiment, the photovoltaic power generation prediction model is used to realize photovoltaic power generation prediction.

[0063] In this embodiment, the photovoltaic power generation prediction evaluation index is used to determine whether the photovoltaic power generation power prediction value is accurate, including: a photovoltaic power generation power prediction accuracy evaluation index, a photovoltaic power generation power prediction timeliness evaluation index, and a photovoltaic power generation power prediction stability evaluation index.

[0064] The working principle and beneficial effects of the above technical solution are: by acquiring photovoltaic power prediction data; extracting data features, determining prediction influencing data features based on photovoltaic power prediction influencing factors, calibrating prediction main data features and determining the prediction influence value of prediction influencing data features on prediction main data features; acquiring prediction parameters, constructing a photovoltaic power prediction model, inputting current photovoltaic power prediction data, and obtaining photovoltaic power prediction values; setting photovoltaic power prediction evaluation indicators and determining the evaluation value of the photovoltaic power prediction value in combination with the prediction influence value, and when the evaluation value of the photovoltaic power prediction value is not lower than the preset value, determining that the photovoltaic power prediction value is accurate; and realizing accurate prediction of photovoltaic power.

[0065] Embodiment 2:

[0066] The embodiment of the present invention provides a photovoltaic power generation prediction method, which obtains historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data and location data and performs data preprocessing to obtain photovoltaic power generation prediction data, including:

[0067] Obtain historical photovoltaic power data, meteorological data, photovoltaic equipment data and location data, perform data preprocessing according to the data type, divide variable data and fixed data and store the data;

[0068] The variable data is subjected to secondary data preprocessing according to the preset data preprocessing strategy to obtain calibrated variable data, which are then combined with the fixed data to determine the photovoltaic power generation power prediction data.

[0069] In this embodiment, primary data preprocessing refers to data division according to data type, for example, dividing historical photovoltaic power generation data and meteorological data into variable data, and dividing photovoltaic power generation equipment data and location data into fixed data.

[0070] In this embodiment, variable data refers to data that can change, such as historical photovoltaic power generation data, where the power generation power is variable.

[0071] In this embodiment, fixed data refers to data that cannot be changed, for example, position data cannot be changed.

[0072] In this embodiment, data preprocessing strategies are preset, such as an outlier detection data preprocessing strategy and an interpolation data preprocessing strategy.

[0073] In this embodiment, the secondary data preprocessing refers to data processing of abnormal data and missing data.

[0074] In this embodiment, the photovoltaic power generation power prediction data refers to historical data used for photovoltaic power generation power prediction.

[0075] The working principle and beneficial effects of the above technical solution are: obtaining historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data and location data and performing data preprocessing, obtaining photovoltaic power generation prediction data, and laying a data foundation for subsequent photovoltaic power generation prediction.

[0076] Embodiment 3:

[0077] The embodiment of the present invention provides a photovoltaic power generation prediction method, which performs secondary data preprocessing on variable data according to a preset data preprocessing strategy, including:

[0078] Mark abnormal data and missing data of variable data, and perform secondary data preprocessing on the abnormal data and missing data according to a preset data preprocessing strategy, wherein the preset data preprocessing strategy includes: an abnormal value detection data preprocessing strategy and an interpolation data preprocessing strategy;

[0079] Data normalization is performed on variable data of different dimensions.

[0080] In this embodiment, abnormal data, for example, variable data a1 far exceeds other variable data, then variable data a1 is marked as abnormal data, and an abnormal value detection data preprocessing strategy is executed to remove the abnormal data.

[0081] In this embodiment, data normalization is performed on variable data of different dimensions. For example, variable data of type b1 and variable data of type b2 are converted into data of the same dimension to achieve data comparability and reduce data anomalies.

[0082] The working principle and beneficial effect of the above technical solution are: secondary data preprocessing is performed on variable data according to a preset data preprocessing strategy, which effectively ensures the accuracy of the variable data and further ensures the accuracy of subsequent photovoltaic power generation prediction.

[0083] Embodiment 4:

[0084] The embodiment of the present invention provides a photovoltaic power generation prediction method, which extracts data features according to photovoltaic power generation prediction data, and determines prediction influencing data features based on photovoltaic power generation prediction influencing factors, including:

[0085] Extract data features of photovoltaic power generation prediction data according to time series or space series to obtain photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features and location data features;

[0086] Obtaining factors influencing photovoltaic power generation prediction, performing data feature screening on the photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features, and location data features, and determining prediction-influencing data features and prediction-irrelevant data features;

[0087] When a type of data feature has data correlation with a photovoltaic power generation data feature, the type of data feature is determined to be a prediction-influencing data feature;

[0088] When there is no data correlation between a type of data feature and the photovoltaic power generation data feature, the type of data feature is determined to be a prediction irrelevant data feature.

[0089] In this embodiment, the photovoltaic power generation power prediction influencing factors include, for example, weather influencing factors, geographical influencing factors, time influencing factors, and equipment influencing factors.

[0090] In this embodiment, the prediction impact data feature refers to a data feature that has data correlation with photovoltaic power generation power. For example, weather factors will affect photovoltaic power generation power, and the data feature corresponding to the weather factors is the prediction impact data feature.

[0091] In this embodiment, the prediction irrelevant data feature refers to a data feature that has no data correlation with photovoltaic power generation power. For example, geographical influencing factors do not affect photovoltaic power generation power, and the data feature corresponding to the geographical influencing factors is the prediction irrelevant data feature.

[0092] The working principle and beneficial effects of the above technical solution are: extracting data features based on photovoltaic power generation prediction data, determining prediction influencing data features based on photovoltaic power generation prediction influencing factors, simplifying data, and improving photovoltaic power generation prediction efficiency.

[0093] Embodiment 5:

[0094] The embodiment of the present invention provides a photovoltaic power generation prediction method, which calibrates the prediction main data feature and determines the prediction influence value of the prediction influence data feature on the prediction main data feature, including:

[0095] The data features are sorted according to the data correlation quantification values ​​of various data features and photovoltaic power generation data features, and the data feature with the largest data correlation quantification value is marked as the prediction main data feature;

[0096] Perform one-way data correlation analysis on the prediction influencing data features and the prediction master data features to determine the prediction impact value of the prediction influencing data features on the prediction master data features.

[0097] In this embodiment, the one-way data correlation analysis refers to the analysis process performed to determine the prediction influence of the prediction influence data on the prediction master data feature.

[0098] The working principle and beneficial effects of the above technical solution are: calibrating the prediction main data features and determining the prediction influence value of the prediction influencing data features on the prediction main data features, which is conducive to the subsequent photovoltaic power generation prediction.

[0099] Embodiment 6:

[0100] The embodiment of the present invention provides a photovoltaic power generation prediction method, which obtains prediction parameters and constructs a photovoltaic power generation prediction model, including:

[0101] Select a preset number of prediction influencing data features, and obtain prediction parameters in combination with the photovoltaic power generation power prediction demand, wherein the prediction parameters include: prediction time parameters, prediction space parameters, and prediction influencing factor parameters;

[0102] Construct a photovoltaic power generation prediction model based on the prediction parameters and the preset machine learning model;

[0103] A data training set is established based on the photovoltaic power generation prediction data, and a photovoltaic power generation prediction model is trained.

[0104] In this embodiment, a preset number of prediction influencing data features are selected to simplify the data features and improve the efficiency of photovoltaic power generation prediction.

[0105] In this embodiment, a machine learning model is preset, such as a neural network learning model, a clustering learning model, and a regression learning model.

[0106] The working principle and beneficial effects of the above technical solution are: obtaining prediction parameters and constructing a photovoltaic power generation prediction model, which is conducive to realizing the prediction of photovoltaic power generation.

[0107] Embodiment 7:

[0108] The embodiment of the present invention provides a photovoltaic power generation power prediction method, which inputs current photovoltaic power generation power prediction data to obtain a photovoltaic power generation power prediction value, including:

[0109] Obtain the current photovoltaic power generation prediction data, input it into the photovoltaic power generation prediction model after model training, and obtain the photovoltaic power generation prediction value;

[0110] When the photovoltaic power generation power prediction value is determined to be accurate, the photovoltaic power generation power prediction value is visualized according to a preset display strategy.

[0111] In this embodiment, a display strategy is preset, for example, a chart display strategy.

[0112] The working principle and beneficial effects of the above technical solution are: inputting the current photovoltaic power prediction data, obtaining the photovoltaic power prediction value and visualizing it, which is conducive to meeting the photovoltaic power prediction needs and then realizing the utilization of photovoltaic power.

[0113] Embodiment 8:

[0114] The embodiment of the present invention provides a photovoltaic power generation power prediction method, which sets a photovoltaic power generation power prediction evaluation index, including:

[0115] Photovoltaic power generation prediction evaluation indicators are set, and the influence weight of the photovoltaic power generation prediction evaluation indicators on the evaluation value of the photovoltaic power generation prediction value is set according to the photovoltaic power generation power prediction demand, wherein the photovoltaic power generation power prediction evaluation indicators include: photovoltaic power generation power prediction accuracy evaluation indicators, photovoltaic power generation power prediction timeliness evaluation indicators and photovoltaic power generation power prediction stability evaluation indicators.

[0116] In this embodiment, the influence weight of the photovoltaic power generation power prediction evaluation index on the evaluation value of the photovoltaic power generation power prediction value can be dynamically adjusted according to the photovoltaic power generation power prediction demand.

[0117] In this embodiment, the photovoltaic power generation power prediction requirement includes the photovoltaic power generation power output prediction requirement and the photovoltaic power generation power efficiency requirement.

[0118] The working principle and beneficial effect of the above technical solution are: setting the photovoltaic power generation power prediction evaluation index is conducive to ensuring the prediction accuracy and prediction efficiency of photovoltaic power generation.

[0119] Embodiment 9:

[0120] The embodiment of the present invention provides a photovoltaic power generation prediction method, which determines an evaluation value of a photovoltaic power generation prediction value in combination with the prediction influence value, and when the evaluation value of the photovoltaic power generation prediction value is not lower than a preset value, determines that the photovoltaic power generation prediction value is accurate, including:

[0121] Determine an evaluation value of the photovoltaic power generation prediction value according to the photovoltaic power generation prediction evaluation index combined with the prediction impact value;

[0122]

[0123] Among them, F represents the evaluation value of the photovoltaic power generation prediction value; xi1 represents the i1th prediction influencing data feature, x1 represents the average value of the prediction influencing data feature; yi1 represents the i1th prediction main data feature; y1 represents the average value of the prediction main data feature; represents the predicted influence value of the i1th prediction influencing data feature on the predicted main data feature; zj1 represents the j1th photovoltaic power prediction value; z1j1 represents the actual photovoltaic power value corresponding to the j1th photovoltaic power prediction value; z1 represents the average value of the photovoltaic power prediction value; tj1 represents the prediction time of the j1th photovoltaic power prediction value; t1 represents the standard photovoltaic power prediction time; n1 represents the number of data features; m1 represents the number of photovoltaic power prediction values; a1 represents the influence weight of the photovoltaic power prediction accuracy evaluation index on the evaluation value; a2 represents the influence weight of the photovoltaic power prediction timeliness evaluation index on the evaluation value; a3 represents the influence weight of the photovoltaic power prediction stability evaluation index on the evaluation value;

[0124] When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate;

[0125] When the evaluation value of the photovoltaic power generation prediction value is lower than the preset value, the photovoltaic power generation prediction value is determined to be inaccurate, and the prediction impact data features are replaced one by one according to the prediction impact value, the photovoltaic power generation prediction model is rebuilt and the photovoltaic power generation prediction value is obtained.

[0126] In this embodiment, the predicted impact data features are replaced one by one according to the predicted impact values, that is, the predicted impact data features with large predicted impact values ​​are replaced first.

[0127] In this embodiment, if the evaluation value of the photovoltaic power generation power prediction value is still lower than the preset value after the prediction influence data features are replaced one by one according to the prediction influence values, the prediction master data features are recalibrated.

[0128] The working principle and beneficial effect of the above technical solution are: by determining the evaluation value of the photovoltaic power generation power prediction value in combination with the predicted influence value, when the evaluation value of the photovoltaic power generation power prediction value is not lower than the preset value, the photovoltaic power generation power prediction value is judged to be accurate, thereby ensuring the prediction accuracy and prediction efficiency of the photovoltaic power generation power prediction value.

[0129] Embodiment 10:

[0130] like Figure 2 As shown, an embodiment of the present invention provides a photovoltaic power generation power prediction system, which specifically includes:

[0131] A data acquisition module is used to acquire historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to acquire photovoltaic power generation prediction data;

[0132] A data analysis module is used to extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features;

[0133] The power prediction module is used to obtain prediction parameters, build a photovoltaic power prediction model, input prediction main data features and prediction influencing data features, and obtain the photovoltaic power prediction value;

[0134] The prediction verification module is used to set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate.

[0135] The working principle and beneficial effects of the above technical solution are: by acquiring photovoltaic power prediction data; extracting data features, determining prediction influencing data features based on photovoltaic power prediction influencing factors, calibrating prediction main data features and determining the prediction influence value of prediction influencing data features on prediction main data features; acquiring prediction parameters, constructing a photovoltaic power prediction model, inputting current photovoltaic power prediction data, and obtaining photovoltaic power prediction values; setting photovoltaic power prediction evaluation indicators and determining the evaluation value of the photovoltaic power prediction value in combination with the prediction influence value, and when the evaluation value of the photovoltaic power prediction value is not lower than the preset value, determining that the photovoltaic power prediction value is accurate; and realizing accurate prediction of photovoltaic power.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A photovoltaic power generation power prediction method, characterized in that: include: Step 1: Obtain historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to obtain photovoltaic power generation prediction data; Step 2: Extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features; Step 3: Obtain prediction parameters, build a photovoltaic power generation prediction model, input current photovoltaic power generation prediction data, and obtain photovoltaic power generation prediction value; Step 4: Set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, the photovoltaic power generation prediction value is determined to be accurate.

2. A photovoltaic power generation prediction method according to claim 1, characterized in that: Obtain historical photovoltaic power data, meteorological data, photovoltaic equipment data, and location data and perform data preprocessing to obtain photovoltaic power forecast data, including: Obtain historical photovoltaic power data, meteorological data, photovoltaic equipment data and location data, perform data preprocessing according to the data type, divide variable data and fixed data and store the data; The variable data is subjected to secondary data preprocessing according to the preset data preprocessing strategy to obtain calibrated variable data, which are then combined with the fixed data to determine the photovoltaic power generation power prediction data.

3. A photovoltaic power generation prediction method according to claim 2, characterized in that: Perform secondary data preprocessing on variable data according to the preset data preprocessing strategy, including: Mark abnormal data and missing data of variable data, and perform secondary data preprocessing on the abnormal data and missing data according to a preset data preprocessing strategy, wherein the preset data preprocessing strategy includes: an abnormal value detection data preprocessing strategy and an interpolation data preprocessing strategy; Data normalization is performed on variable data of different dimensions.

4. A photovoltaic power generation prediction method according to claim 1, characterized in that: Data features are extracted from the photovoltaic power generation prediction data, and prediction influencing data features are determined based on the photovoltaic power generation prediction influencing factors, including: Extract data features of photovoltaic power generation prediction data according to time series or space series to obtain photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features and location data features; Obtaining factors influencing photovoltaic power generation prediction, performing data feature screening on the photovoltaic power generation data features, meteorological data features, photovoltaic power generation equipment data features, and location data features, and determining prediction-influencing data features and prediction-irrelevant data features; When a type of data feature has data correlation with a photovoltaic power generation data feature, the type of data feature is determined to be a prediction-influencing data feature; When there is no data correlation between a type of data feature and the photovoltaic power generation data feature, the type of data feature is determined to be a prediction irrelevant data feature.

5. A photovoltaic power generation prediction method according to claim 4, characterized in that: Calibrating the forecast master data features and determining the forecast impact values ​​of the forecast influencing data features on the forecast master data features, including: The data features are sorted according to the data correlation quantification values ​​of various data features and photovoltaic power generation data features, and the data feature with the largest data correlation quantification value is marked as the prediction main data feature; Perform one-way data correlation analysis on the prediction influencing data features and the prediction master data features to determine the prediction impact value of the prediction influencing data features on the prediction master data features.

6. A photovoltaic power generation power prediction method according to claim 1, characterized in that: Obtain prediction parameters and build a photovoltaic power generation prediction model, including: Select a preset number of prediction influencing data features, and obtain prediction parameters in combination with the photovoltaic power generation power prediction demand, wherein the prediction parameters include: prediction time parameters, prediction space parameters, and prediction influencing factor parameters; Construct a photovoltaic power generation prediction model based on the prediction parameters and the preset machine learning model; A data training set is established based on the photovoltaic power generation prediction data, and a photovoltaic power generation prediction model is trained.

7. A photovoltaic power generation prediction method according to claim 1, characterized in that: Input the current photovoltaic power generation prediction data to obtain the photovoltaic power generation prediction value, including: Obtain the current photovoltaic power generation prediction data, input it into the photovoltaic power generation prediction model after model training, and obtain the photovoltaic power generation prediction value; When the photovoltaic power generation power prediction value is determined to be accurate, the photovoltaic power generation power prediction value is visualized according to a preset display strategy.

8. A photovoltaic power generation prediction method according to claim 1, characterized in that: Set photovoltaic power generation prediction evaluation indicators, including: Photovoltaic power generation prediction evaluation indicators are set, and the influence weight of the photovoltaic power generation prediction evaluation indicators on the evaluation value of the photovoltaic power generation prediction value is set according to the photovoltaic power generation power prediction demand, wherein the photovoltaic power generation power prediction evaluation indicators include: photovoltaic power generation power prediction accuracy evaluation indicators, photovoltaic power generation power prediction timeliness evaluation indicators and photovoltaic power generation power prediction stability evaluation indicators.

9. A photovoltaic power generation prediction method according to claim 8, characterized in that: Determining an evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value, and when the evaluation value of the photovoltaic power generation prediction value is not lower than a preset value, determining that the photovoltaic power generation prediction value is accurate, including: Determine an evaluation value of the photovoltaic power generation prediction value according to the photovoltaic power generation prediction evaluation index combined with the prediction impact value; Among them, F represents the evaluation value of the photovoltaic power generation prediction value; xi1 represents the i1th prediction influencing data feature, x1 represents the average value of the prediction influencing data feature; yi1 represents the i1th prediction main data feature; y1 represents the average value of the prediction main data feature; represents the predicted influence value of the i1th prediction influencing data feature on the predicted main data feature; zj1 represents the j1th photovoltaic power prediction value; z1j1 represents the actual photovoltaic power value corresponding to the j1th photovoltaic power prediction value; z1 represents the average value of the photovoltaic power prediction value; tj1 represents the prediction time of the j1th photovoltaic power prediction value; t1 represents the standard photovoltaic power prediction time; n1 represents the number of data features; m1 represents the number of photovoltaic power prediction values; a1 represents the influence weight of the photovoltaic power prediction accuracy evaluation index on the evaluation value; a2 represents the influence weight of the photovoltaic power prediction timeliness evaluation index on the evaluation value; a3 represents the influence weight of the photovoltaic power prediction stability evaluation index on the evaluation value; When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate; When the evaluation value of the photovoltaic power generation prediction value is lower than the preset value, the photovoltaic power generation prediction value is determined to be inaccurate, and the prediction impact data features are replaced one by one according to the prediction impact value, the photovoltaic power generation prediction model is rebuilt and the photovoltaic power generation prediction value is obtained.

10. A photovoltaic power generation power prediction system, characterized in that: include: A data acquisition module is used to acquire historical photovoltaic power generation data, meteorological data, photovoltaic power generation equipment data, and location data and perform data preprocessing to acquire photovoltaic power generation prediction data; A data analysis module is used to extract data features based on photovoltaic power generation prediction data, determine prediction influencing data features based on photovoltaic power generation prediction influencing factors, calibrate prediction master data features and determine the prediction influence value of prediction influencing data features on prediction master data features; The power prediction module is used to obtain prediction parameters, build a photovoltaic power prediction model, input prediction main data features and prediction influencing data features, and obtain the photovoltaic power prediction value; The prediction verification module is used to set the photovoltaic power generation prediction evaluation index and determine the evaluation value of the photovoltaic power generation prediction value in combination with the prediction impact value. When the evaluation value of the photovoltaic power generation prediction value is not lower than the preset value, it is determined that the photovoltaic power generation prediction value is accurate.

Images