Method and device for predicting total power generation of a power grid, electronic equipment, storage medium and computer program product

Abstract

The present disclosure relates to a method and device for predicting total power generation of a power grid, electronic equipment, a storage medium and a computer program product. The method comprises: calculating a first correlation coefficient between the total power generation of the power grid and the sum of power generations of a plurality of known power fields; in response to the first correlation coefficient being greater than or equal to a first preset value, predicting a ratio of a predicted value of the total power generation of the power grid and a sum of predicted values of power generations of at least one known power field, and obtaining the predicted value of the total power generation of the power grid based on a product of the sum of the predicted values of the power generations of the at least one known power field and the ratio; and in response to the first correlation coefficient being less than the first preset value, predicting a predicted value of power generation of a to-be-determined power field, and obtaining the predicted value of the total power generation of the power grid based on a sum of the predicted value of the power generation of the to-be-determined power field and the sum of the predicted values of the power generations of the at least one known power field. The accuracy of the prediction of the total power generation of the power grid can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of electric field power generation prediction, specifically to a method, apparatus, electronic device, storage medium, and computer program product for predicting the total power generation of the entire power grid. Background Technology

[0002] The nationwide renewable energy output forecasting model, based on a province or power grid, predicts the total output of all wind and solar power plants within a province or a specific power grid in the future. Nationwide renewable energy output forecasting can assist the power grid in making control decisions and ensuring power system balance. Furthermore, in electricity trading scenarios, electricity price forecasting also relies on nationwide renewable energy output forecasting; therefore, improving the accuracy of nationwide renewable energy output forecasting has a significant impact on power grid control and electricity trading.

[0003] Currently, the forecasting of renewable energy output across the entire power grid mainly adopts a single-field modeling method, which involves summing up the predicted power reported to the grid daily by each electric field. This method relies on the power prediction results of each electric field. However, considering the confidentiality of grid information, forecasting companies find it difficult to obtain information on all electric fields in the province. How to establish a nationwide renewable energy output forecasting model based on a limited number of electric fields and improve its accuracy is an important issue that needs to be addressed. Summary of the Invention

[0004] The method, apparatus, electronic device, storage medium, and computer program product for predicting total power generation of the entire network provided in the exemplary embodiments of this disclosure can at least solve the above-mentioned technical problems and other technical problems not mentioned above.

[0005] According to one aspect of this disclosure, a method for predicting the total power generation of a power grid is provided. The method includes: obtaining the total power generation of the entire power grid and the power generation of multiple known electric fields, wherein all electric fields in the entire grid other than at least one of the multiple known electric fields are considered undetermined electric fields; calculating a first correlation coefficient between the total power generation of the entire power grid and the sum of the power generation of the multiple known electric fields; in response to the first correlation coefficient being greater than or equal to a first preset value, predicting the ratio of the predicted value of the total power generation of the entire power grid to the sum of the predicted values ​​of the power generation of the at least one known electric field, and obtaining a predicted value of the total power generation of the entire power grid based on the product of the sum of the predicted values ​​of the power generation of the at least one known electric field and the ratio; in response to the first correlation coefficient being less than the first preset value, predicting the predicted value of the power generation of the undetermined electric fields, and obtaining a predicted value of the total power generation of the entire power grid based on the sum of the predicted values ​​of the power generation of the undetermined electric fields and the sum of the predicted values ​​of the power generation of the at least one known electric field.

[0006] Optionally, the ratio of the predicted total power generation of the entire grid to the sum of the predicted power generation of the at least one known electric field includes: obtaining the temporal and meteorological characteristics of the region where the at least one known electric field is located; establishing a first model between the temporal and meteorological characteristics of the region where the at least one known electric field is located, the sum of the power generation of the at least one known electric field, and the ratio of the total power generation of the entire grid to the sum of the power generation of the at least one known electric field; and obtaining the ratio of the predicted total power generation of the entire grid to the sum of the predicted power generation of the at least one known electric field based on the first model.

[0007] Optionally, the predicted value of the power generation of the undetermined electric field includes: obtaining the power generation of the undetermined electric field based on the difference between the total power generation of the entire grid and the sum of the power generation of the at least one known electric field; calculating a second correlation coefficient between the sum of the power generation of the at least one known electric field and the power generation of the undetermined electric field; identifying a number of known electric fields whose second correlation coefficient is greater than a second preset value; obtaining the meteorological characteristics of the area where the undetermined electric field is located; establishing a second model between the sum of the power generation of the number of known electric fields, the meteorological characteristics of the area where the undetermined electric field is located, and the power generation of the undetermined electric field; and obtaining the predicted value of the power generation of the undetermined electric field based on the second model.

[0008] Optionally, obtaining the power generation of multiple known electric fields includes: obtaining the power generation data of each known electric field; performing quality detection on the power generation data of each known electric field to determine the presence of abnormal electric fields; filling the power generation data of the abnormal electric fields with data; and obtaining the power generation of the multiple known electric fields based on the power generation data of each known electric field after data filling.

[0009] Optionally, data filling is performed on the power generation data of the abnormal electric field, including: normalizing the power generation data of each known electric field based on the installed capacity of each of the plurality of known electric fields; calculating the third correlation coefficient between the normalized power generation data of each known electric field; determining the reference electric field with the largest third correlation coefficient with the abnormal electric field from among the plurality of known electric fields; performing inverse normalization on the power generation data of the reference electric field based on the installed capacity of the abnormal electric field to obtain filled data; and filling the power generation data of the abnormal electric field based on the filled data.

[0010] Optionally, the method further includes: determining at least one known electric field among the plurality of known electric fields based on the accuracy of historical power predictions of the plurality of known electric fields.

[0011] Optionally, the undetermined electric field includes multiple undetermined electric fields, wherein obtaining the meteorological characteristics of the area where the undetermined electric field is located includes: obtaining the first meteorological characteristics of the area where the multiple undetermined electric fields are located; dividing the multiple undetermined electric fields into multiple predetermined areas based on the first meteorological characteristics of the area where the multiple undetermined electric fields are located; and obtaining the second meteorological characteristics of the multiple predetermined areas as the meteorological characteristics of the area where the undetermined electric field is located.

[0012] Optionally, both the first meteorological feature and the second meteorological feature can be wind speed features or both can be irradiance features.

[0013] According to another aspect of this disclosure, a device for predicting the total power generation of an entire power grid is also provided. The device includes: a data acquisition unit configured to acquire the total power generation of the entire power grid and the power generation of multiple known electric fields, wherein all electric fields in the entire power grid except for at least one of the known electric fields are undetermined electric fields; a coefficient calculation unit configured to calculate a first correlation coefficient between the total power generation of the entire power grid and the sum of the power generation of the multiple known electric fields; and a power prediction unit configured to: predict power in response to the first correlation coefficient being greater than or equal to a first prediction value. The predicted value of the total power generation of the entire network is calculated by setting a value and then calculating the ratio of the predicted power generation of the at least one known electric field to the sum of the predicted power generation of the at least one known electric field. Based on the product of the predicted power generation of the at least one known electric field and the ratio, the predicted power generation of the entire network is obtained. In response to the first correlation coefficient being less than the first preset value, the predicted power generation of the undetermined electric field is predicted, and based on the sum of the predicted power generation of the undetermined electric field and the predicted power generation of the at least one known electric field, the predicted power generation of the entire network is obtained.

[0014] According to another aspect of the present disclosure, an electronic device is also provided, comprising: at least one processor; and at least one memory storing computer-executable instructions, wherein the computer-executable instructions, when executed by the at least one processor, cause the at least one processor to perform the prediction method for total power generation of the entire grid as described above.

[0015] According to another aspect of the embodiments of this disclosure, a computer-readable storage medium storing instructions is also provided, which, when executed by at least one processor, cause the at least one processor to perform the prediction method for total power generation of the entire network as described above.

[0016] According to another aspect of the embodiments of this disclosure, a system is also provided that includes at least one computing device and at least one storage device for storing instructions, wherein the instructions, when executed by the at least one computing device, cause the at least one computing device to perform the prediction method for total power generation of the entire network as described above.

[0017] According to another aspect of the embodiments of this disclosure, a computer program product is also provided, including a computer program / instructions, which, when executed by a processor, implement the method for predicting the total power generation of the entire network as described in any of the above embodiments.

[0018] The technical solutions provided in this disclosure offer at least the following beneficial effects:

[0019] According to the prediction method, apparatus, electronic equipment, storage medium, and computer program product of the present disclosure, it is possible to predict the total power generation of the entire grid by constructing intermediate targets and mining missing information about undetermined electric fields, based on some known electric field data and the actual total power output of the entire grid, thus predicting the power output of new energy sources in the entire grid. Furthermore, the present disclosure can mine the correlation between electric fields while modeling single electric fields, thereby reducing the error accumulation caused by the summation method of single electric field prediction and improving the accuracy of predicting the power generation of new energy sources in the entire grid. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure, and are not intended to unduly limit this disclosure.

[0021] Figure 1 A flowchart illustrating a method for predicting total power generation of the entire grid in an exemplary embodiment of this disclosure;

[0022] Figure 2 This diagram illustrates a process for preprocessing the acquired actual power generation in an exemplary embodiment of the present disclosure.

[0023] Figure 3 This illustration shows a flowchart of an exemplary embodiment of the present disclosure, in which a difference-based modeling method and a ratio-based modeling method are distinguished based on a first correlation coefficient.

[0024] Figure 4 This illustrates an exemplary embodiment of the present disclosure, showing p in a certain province during different time periods. A′ A comparison plot between the R scatter distribution and the R-point distribution;

[0025] Figure 5 A flowchart illustrating the total power generation prediction method for the entire power grid based on the ratio method in an exemplary embodiment of this disclosure is shown.

[0026] Figure 6 A flowchart illustrating the total power generation prediction method for the entire power grid based on the difference method in an exemplary embodiment of this disclosure is shown.

[0027] Figure 7 A block diagram illustrating a device for predicting total power generation across the entire grid, as shown in an exemplary embodiment of this disclosure;

[0028] Figure 8 A block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure. Detailed Implementation

[0029] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0030] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following examples do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0031] It should be noted that the phrase "at least one of several items" in this disclosure refers to three parallel cases: "any one of the several items", "a combination of any number of the several items", and "all of the several items". For example, "including at least one of A and B" includes the following three parallel cases: (1) including A; (2) including B; (3) including A and B. As another example, "performing at least one of step one and step two" indicates the following three parallel cases: (1) performing step one; (2) performing step two; (3) performing both step one and step two.

[0032] Currently, there are relatively few technologies for predicting the power output of renewable energy across the entire power grid. Most power grids use a single-field modeling method, which involves summing the predicted power of each field in the province to obtain the total predicted power. Therefore, the single-field modeling method relies on the prediction results of each field, requiring basic information and observation data of all renewable energy fields, as well as sufficient computing resources to perform predictions for all single fields. Furthermore, summing the predicted power of each field will cause the prediction errors of each field to accumulate, and the lack of exploration of regional characteristics makes it difficult to improve the accuracy of the overall power prediction.

[0033] To address the aforementioned issues, this disclosure provides a method, apparatus, electronic device, storage medium, and computer program product for predicting the total power generation of the entire power grid. Based on partially known electric field data and the actual total power output of the entire grid, it constructs intermediate targets to mine missing information about undetermined electric fields, thereby predicting the power output of new energy sources across the entire grid. Furthermore, this disclosure can simultaneously mine the correlation between electric fields by modeling single electric fields, thereby reducing the error accumulation caused by the summation method of single-field predictions and improving the accuracy of predicting the power generation of new energy sources across the entire grid.

[0034] First, the technical terms defined in the exemplary embodiments of this disclosure are introduced.

[0035] Total renewable energy output: refers to the total power generated by all renewable energy power generation facilities within the power grid.

[0036] The LGB algorithm, or LightGBM (Light Gradient Boosting Machine), is an efficient gradient boosting framework that uses a tree-based learning algorithm.

[0037] DBSCAN clustering algorithm: It is a density-based clustering algorithm that divides high-density regions into clusters and separates these clusters by low-density regions, thereby classifying the data.

[0038] Below, we will refer to Figures 1 to 8 This disclosure specifically describes the methods, apparatus, electronic devices, storage media, and computer program products for predicting the total power generation of the entire network.

[0039] Figure 1 A flowchart illustrating a method for predicting total power generation of the entire grid in an exemplary embodiment of this disclosure is shown.

[0040] Reference Figure 1 In step 101, the total power generation of the entire network and the power generation of multiple known electric fields are obtained.

[0041] An electric field can be considered a known electric field if one or more of the following information is available: basic electric field information (installed capacity, latitude and longitude, grid connection time, etc.), actual wind speed, actual power generation, and historical predicted power generation.

[0042] Since there are many power plants involved in the new energy network, the observation data of the known power plants are subject to power outages and maintenance, and there are also situations where power plants are being expanded or new power plants are being connected to the grid. Therefore, the acquired data can be preprocessed.

[0043] The preprocessing of the acquired data can be performed in the following ways:

[0044] According to an exemplary embodiment of this disclosure, power generation data of each of a plurality of known electric fields are acquired; the power generation data of each known electric field is subjected to quality inspection to identify abnormal electric fields with anomalous power generation data; the power generation data of the abnormal electric fields is filled with data; and the power generation data of the plurality of known electric fields is acquired based on the data-filled power generation data of each known electric field. This allows for preprocessing of the acquired data, improving prediction efficiency.

[0045] Specifically, as an example, quality checks can identify anomalous data such as jumps, straight lines, and consecutive missing data. Understandably, other anomalous data can also be identified.

[0046] Afterwards, data can be filled in for electric fields with continuously missing data, installed capacity changes, and newly added electric fields. As an example, the main method is to fill in data using adjacent electric fields.

[0047] The adjacent electric fields can be filled in the following ways:

[0048] According to an exemplary embodiment of this disclosure, the power generation data of each known electric field is normalized based on the installed capacity of each field; a third correlation coefficient is calculated between the normalized power generation data of each known electric field; from among the known electric fields, a reference electric field with the largest third correlation coefficient with the abnormal electric field is determined; the power generation data of the reference electric field after normalization is inversely normalized based on the installed capacity of the abnormal electric field to obtain filled data; and the power generation data of the abnormal electric field is filled based on the filled data. This allows for the processing of abnormal data using adjacent electric field filling, improving prediction efficiency.

[0049] Specifically, as an example, we can first normalize the actual power generation (power generation data) of all known electric fields by dividing the actual power generation by the installed capacity.

[0050] Next, a third correlation coefficient can be calculated between the normalized actual power generation of each known electric field. It is understood that the term "third" here is used to distinguish similar objects, not necessarily to describe a specific order or sequence. This third correlation coefficient can be obtained by calculating the Pearson correlation coefficient. Of course, other feasible correlation indicators can also be used to calculate it.

[0051] Subsequently, for a known electric field containing anomalous data to be filled, the actual power generation of the known electric field with the largest third correlation coefficient can be selected to obtain the filling data of the known electric field containing anomalous data.

[0052] Finally, the data used for filling, namely the actual power generation of the known electric field with the largest third correlation coefficient with the known electric field to be filled, can be multiplied by the installed capacity of the known electric field to be filled for inverse normalization to generate a set of filling data. This set of filling data is then used as the historical actual power generation of the known electric field to be filled.

[0053] Figure 2 This illustration shows a flowchart of the preprocessing of the acquired actual power generation in an exemplary embodiment of the present disclosure.

[0054] Reference Figure 2 According to an exemplary embodiment of this disclosure, after obtaining the total power of the entire network (i.e., the actual power generation of all known electric fields in the entire network), data quality control can be performed. For example, data anomaly detection control can be performed to identify abnormal data, installed capacity change data conversion can be performed to adjust the corresponding actual power generation according to the capacity before and after the installed capacity change, and continuous missing stations can be removed to remove known stations with continuously missing data.

[0055] Next, data filling can be performed on known electric fields with data anomalies. For example, first, the actual power generation of all known electric fields can be normalized. Then, the most relevant power station can be searched, that is, the known electric field with the highest correlation to the known electric field to be filled can be searched. Finally, power data restoration and filling can be performed, that is, the actual power generation of the most relevant known electric field can be restored and then the data filling can be performed on the known electric field to be filled. After that, the whole network modeling and optimization scheme can be executed.

[0056] The specific optimization plan for the entire network modeling can be as follows:

[0057] Return to reference Figure 1 In step 102, the first correlation coefficient between the total power generation of the entire network and the sum of the power generation of multiple known electric fields is calculated.

[0058] Specifically, as an example only, the first correlation coefficient can be calculated in the same way as the third correlation coefficient, i.e., it can be obtained by calculating the Pearson correlation coefficient. Of course, other feasible correlation indicators can also be used to calculate it.

[0059] Understandably, the terms "first" and "third" here are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0060] Based on relevant data from the entire network, the modeling of the power output of new energy sources across the entire network based on a finite electric field can be divided into two types: a modeling method based on the difference and a modeling method based on the ratio, based on the first correlation coefficient.

[0061] Figure 3This illustration shows a flowchart of an exemplary embodiment of the present disclosure, in which a difference-based modeling method and a ratio-based modeling method are distinguished based on a first correlation coefficient.

[0062] Reference Figure 3 Let the total power generation of the entire network be P. Assuming that the basic information and observation data (i.e., power generation, actual power generation) of n electric fields are known, the total actual power generation of the n (known) electric fields can be calculated as P. A =P1+…+P n Furthermore, other unknown electric fields can be considered as a whole, i.e., P n+1 =PP A Next, we can analyze the total power P and P0. A The correlation r is used to calculate the first correlation coefficient. When the correlation r is high, i.e., r>=θ (θ is the first preset value), it indicates that the actual power generation distribution of the known electric field is similar to the actual power generation distribution of the entire grid, and the ratio-based modeling method (ratio modeling method, ratio method) can be used. When the correlation is low, i.e., r<θ, it indicates that the actual power generation distribution of the known electric field cannot represent the entire electric field (i.e., the actual power generation distribution of the entire grid), and the difference-based modeling method (difference modeling method, interpolation method) can be used.

[0063] Next, field filtering can be performed on the known electric field.

[0064] According to exemplary embodiments of this disclosure, at least one known electric field among a plurality of known electric fields can be determined based on the accuracy of historical power predictions of a plurality of known electric fields. Accuracy filtering can improve prediction efficiency.

[0065] As an example only, we can consider the actual power P1, P2, ..., P of n known electric fields. n and historical predicted power Accuracy statistics were performed, and a set A′ of stations with an accuracy greater than 80% was selected, which means that there is at least one known electric field.

[0066] In addition to accuracy, station filtering and screening can also be performed using indicators such as relative error, absolute error, and root mean square error.

[0067] Next, we will introduce the specific process of the ratio-based modeling method.

[0068] The modeling objective of the ratio method can be constructed as follows:

[0069] Specifically, as an example only, the ratio method can calculate the total actual power P of a known electric field (i.e., the known electric field in the field set A′) with an accuracy greater than 80%. A′ =P1+…+P n′and total predicted power Define the ratio between the total actual power of the entire network and the total actual power of the known electric field in A′, i.e. The ratio method can calculate the total predicted power of the entire network using the prediction ratio R.

[0070] Return to reference Figure 1 In step 103, in response to the first correlation coefficient being greater than or equal to the first preset value, the ratio of the predicted value of the total power generation of the entire network to the sum of the predicted values ​​of the power generation of at least one known electric field is calculated, and the predicted value of the total power generation of the entire network is obtained based on the product of the sum of the predicted values ​​of the power generation of at least one known electric field and the ratio.

[0071] Specifically, as an example only, the sum of the predicted values ​​of the total power generation of the entire network p^ and the predicted values ​​of the power generation of the known electric fields in A′ can be used. The ratio r, and the power generated by the known electric field in A′ To make a prediction, that is, to sum the predicted values ​​of the generated power from the known electric field in A′. The predicted total power generation of the entire network is calculated by multiplying the power generation by the ratio r mentioned above.

[0072] Since the ratios differ for different weather types and at different times, meteorological and temporal characteristics based primarily on weather types can be constructed. Based on these weather type and temporal characteristics, the total power p of the known electric field in A′ can then be established. A′ The model between the ratio R and the model.

[0073] More specifically, in step 103, the temporal and meteorological characteristics of the region where at least one known electric field is located can be obtained; a first model is established between the temporal and meteorological characteristics of the region where at least one known electric field is located, the sum of the power generation of at least one known electric field, and the ratio of the total power generation of the entire grid to the sum of the power generation of at least one known electric field; based on the first model, the ratio of the predicted value of the total power generation of the entire grid to the sum of the predicted values ​​of the power generation of at least one known electric field is obtained. By constructing a model based on the temporal and meteorological characteristics, the total power p of the known electric fields in A′ can be obtained. A′ The correspondence between the ratio R and the ratio.

[0074] Specifically, temporal characteristics can include, but are not limited to, different time periods of the day, such as different hours; meteorological characteristics can include, but are not limited to, different weather types. It is understood that temporal characteristics can also be time features other than different hours, and meteorological characteristics can also be meteorological features other than weather types.

[0075] Taking time characteristics as an example, Figure 4This illustrates an exemplary embodiment of the present disclosure, showing p in a certain province during different time periods. A′ A comparison chart between the R scatter plot and the R scatter plot, see reference. Figure 4 (a), (b), (c), and (d) represent the p values ​​for 0-6, 6-12, 12-18, and 18-24 hours of the day, respectively. A′ The scatter distribution between p and R shows that p A′ The distribution of R scatter points varies significantly across different time periods.

[0076] The total predicted power of the power station can be based on weather and / or time characteristics T, as well as the known electric field in A′. The IGBT algorithm is used to build a model to the modeling target R, i.e., the first model, in order to predict the corresponding ratio r.

[0077] Figure 5 This diagram illustrates a flowchart of a method for predicting total power generation across the entire power grid based on the ratio method, as shown in an exemplary embodiment of this disclosure.

[0078] Reference Figure 5 According to an exemplary embodiment of this disclosure, let the total power (generation) of the entire network be P. Assuming that the basic information and observation data (i.e., power generation, actual power generation) of n electric fields (known electric fields) are known, a set A′ of electric fields with a historical accuracy greater than 80% can be selected, and a modeling target can be constructed. Next, the time features T1…T can be constructed. 24 Then, classify the weather type: cloudy, sunny, rainy, strong wind, light wind. Afterwards, it can be based on T1…T 24 and p A′ The IGBT algorithm is used to build a model, predict the ratio R(R), and calculate the total predicted power of the entire network. That is, the predicted value of the total power generation of the entire network.

[0079] Next, we will introduce the specific process of the difference-based modeling method.

[0080] According to an exemplary embodiment of this disclosure, all electric fields in the entire network, except for at least one known electric field among a plurality of known electric fields, are considered as undetermined electric fields. Using the total power generation of the undetermined electric fields as the modeling target can improve prediction efficiency.

[0081] Specifically, as an example only, the modeling objective of the interpolation method can be constructed in the following way:

[0082] It can calculate the total actual power P of all known electric fields (i.e., the known electric fields in the field set A′) with an accuracy greater than 80%. A′ =P1+…+P n′ And it can represent the total actual power PP of all electric fields other than A′, i.e., the electric field to be determined. A′Viewed as a whole, that is, P n+1 =PP A′ The modeling target is the total power generation of the unknown electric field.

[0083] Return to reference Figure 1 In step 104, in response to the first correlation coefficient being less than the first preset value, the predicted value of the power generation of the undetermined electric field is predicted, and based on the sum of the predicted value of the power generation of the undetermined electric field and the predicted value of the power generation of at least one known electric field, the predicted value of the total power generation of the entire network is obtained.

[0084] Specifically, as an example only, we can consider the power generation of a given electric field. Make predictions and forecast the power generation of the known electric field in A′. This can be obtained by summing the predicted power output from the known electric field in A′. And the predicted power generation of the undetermined electric field. Calculate the predicted value of the total power generation of the entire network.

[0085] Next, we will introduce how to construct power generation data for predicting an unknown electric field by utilizing power characteristics. The model.

[0086] According to exemplary embodiments of this disclosure, the power generation of a to-be-determined electric field can be obtained based on the difference between the total power generation of the entire power grid and the sum of the power generation of at least one known electric field; a second correlation coefficient is calculated between the sum of the power generation of at least one known electric field and the power generation of the to-be-determined electric field; several known electric fields among the at least one known electric field whose second correlation coefficient is greater than a second preset value are identified; meteorological characteristics of the area where the to-be-determined electric field is located are obtained; a second model is established between the sum of the power generation of the several known electric fields, the meteorological characteristics of the area where the to-be-determined electric field is located, and the power generation of the to-be-determined electric field; and a predicted value of the power generation of the to-be-determined electric field is obtained based on the second model. Meteorological characteristics can be constructed through power correlation analysis to efficiently predict the power generation of the to-be-determined electric field.

[0087] Specifically, as an example only, we can compare the actual generated power in A′ with the generated power P of the undetermined electric field. n+1 Normalize the output and calculate the actual power generation of each station in A′ and P. n+1 The correlation index between them, namely the second correlation coefficient, can be calculated in the same way as the first correlation coefficient, i.e., by calculating the Pearson correlation coefficient. Of course, other feasible correlation index calculation methods can also be used.

[0088] Understandably, the terms "second" and "third" here are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0089] Next, based on the second correlation coefficient, we can screen from A′ for correlations with P. n+1 The actual power generation P of several known electric fields with high correlation (i.e., the second correlation coefficient is greater than the second preset value) α1 ,...P αi Furthermore, it is possible to obtain the meteorological characteristics β of the area where the unknown electric field is located. i ,...β k For example, this includes, but is not limited to, average wind speed (wind farm) or average irradiance (photovoltaic power plant); next, the IGBT algorithm can be used to establish the data from P... α1 ,...P αi ,β i ,...β k To P n+1 The model (i.e., the sum of the power generation of several known electric fields and the meteorological characteristics of the area where the unknown electric field is located, and a second model relating it to the power generation of the unknown electric field) is used to obtain the total predicted power of the unknown electric field. Finally, the predicted value of the total power generation of the entire network.

[0090] According to exemplary embodiments of this disclosure, the undetermined electric field may include multiple undetermined electric fields. Therefore, in the process of acquiring the meteorological characteristics of the region where the undetermined electric field is located, first meteorological characteristics of the region where the multiple undetermined electric fields are located can be acquired; based on the first meteorological characteristics of the region where the multiple undetermined electric fields are located, the multiple undetermined electric fields are divided into multiple predetermined regions; second meteorological characteristics of the multiple predetermined regions are acquired as the meteorological characteristics of the region where the undetermined electric field is located. Regions can be divided and regional characteristics constructed through correlation analysis of power and meteorology.

[0091] Specifically, as an example only, the predicted weather for electric fields other than known electric fields, i.e., fields of unknown electric field, can be divided into regions. The division method can be as follows:

[0092] First, calculate the daily maximum wind speed, daily minimum wind speed, and daily wind speed standard deviation for each undetermined electric field as the first meteorological characteristic of each undetermined electric field (if the undetermined electric field is a photovoltaic electric field, then statistically analyze the relevant indicators of irradiance).

[0093] Next, the DBSCAN clustering algorithm can be used to divide the multiple undetermined electric fields into multiple regions (predetermined regions) based on the first meteorological characteristics of the areas where the undetermined electric fields are located. The second meteorological characteristic of each region, namely the daily average wind speed (or average irradiance for photovoltaic applications), is then obtained as the meteorological characteristic β of the region where the undetermined electric fields are located. i ,...β k .

[0094] According to exemplary embodiments of this disclosure, both the first meteorological feature and the second meteorological feature are either wind speed features or both are irradiance features. Wind speed features are used for wind farms, and irradiance features are used for photovoltaic power plants.

[0095] Figure 6 This diagram illustrates a flowchart of a method for predicting total power generation across the entire power grid based on the difference method in an exemplary embodiment of this disclosure.

[0096] Reference Figure 6 According to an exemplary embodiment of this disclosure, let the total power (generation) of the entire network be P. Assuming that the basic information and observation data (i.e., power generation, actual power generation) of n electric fields (known electric fields) are known, a set A′ of electric fields with a historical accuracy greater than 80% can be selected, and a modeling target P can be constructed. n+1 =PP A′ Next, feature construction can be performed to obtain the power representation in the known electric field A′ (the power in A′ selected from P). n+1 (Actual power generation with high correlation) P α1 ,...P αi And the meteorological representative (meteorological characteristics of the area where the undetermined power station is located) β of the selected area outside of the entire network A′. i ,...β k Subsequently, based on power and meteorological data, an IGBT algorithm can be used to build a model to predict the power of the missing electric field set. That is, the total predicted power of the undetermined electric field Finally, the total predicted power of the entire network can be calculated. That is, the predicted value of the total power generation of the entire network.

[0097] It is understood that the LGB algorithm mentioned in this disclosure can be used with other machine learning and deep learning algorithms.

[0098] According to exemplary embodiments of this disclosure, a method for modeling the power output of renewable energy across the entire network based on finite electric fields is proposed. Specifically, two methods for modeling the power output of renewable energy across the entire network by constructing intermediate modeling targets are proposed: a modeling method based on differences and a modeling method based on ratios. This method can fully utilize the sum of the total power of known electric fields, or it can obtain the sum of the total power of known electric fields using modeling methods. It is applicable to the prediction of electricity and power output in both network-wide and cluster scenarios. Furthermore, while ensuring full utilization of information related to individual electric fields, this disclosure can reduce the error accumulation caused by the single-field prediction and summation method by transforming the modeling target and mining local electric field information and inter-field correlations based on single-field data, thereby improving the accuracy of the network-wide model.

[0099] Figure 7 A block diagram of a device for predicting total power generation of the entire grid is shown in an exemplary embodiment of this disclosure.

[0100] Reference Figure 7 The exemplary embodiments of this disclosure also provide a prediction device 700 for the total power generation of the entire network, which may include, but is not limited to, a data acquisition unit 701, a coefficient calculation unit 702, and a power prediction unit 703.

[0101] The data acquisition unit 701 can acquire the total power generation of the entire network and the power generation of multiple known electric fields. Among them, all electric fields in the entire network except for at least one known electric field are undetermined electric fields.

[0102] The coefficient calculation unit 702 can calculate the first correlation coefficient between the total power generation of the entire network and the sum of the power generation of multiple known electric fields.

[0103] The power prediction unit 703 can, in response to a first correlation coefficient being greater than or equal to a first preset value, predict the ratio of the predicted value of the total power generation of the entire network to the sum of the predicted values ​​of the power generation of at least one known electric field, and obtain the predicted value of the total power generation of the entire network based on the product of the sum of the predicted values ​​of the power generation of at least one known electric field and the ratio; in response to a first correlation coefficient being less than the first preset value, predict the predicted value of the power generation of the undetermined electric field, and obtain the predicted value of the total power generation of the entire network based on the sum of the predicted values ​​of the power generation of the undetermined electric field and the sum of the predicted values ​​of the power generation of at least one known electric field.

[0104] It is understood that the specific implementation process of the above-described exemplary embodiment of the total power generation prediction device 700 is largely the same as that of the above-described exemplary embodiment of the total power generation prediction method, and will not be described in detail here. The total power generation prediction device 700 can be configured as software, hardware, firmware, or any combination thereof to perform specific functions. For example, these devices can correspond to dedicated integrated circuits, pure software code, or modules combining software and hardware. Furthermore, one or more functions implemented by these devices can also be uniformly executed by components in physical entity devices (e.g., processors, clients, or servers).

[0105] Figure 8 A block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure.

[0106] Reference Figure 8 The electronic device 800 includes at least one memory 801 and at least one processor 802. The at least one memory 801 stores a set of computer-executable instructions. When the set of computer-executable instructions is executed by the at least one processor 802, a method for predicting the total power generation of the entire network according to an exemplary embodiment of the present disclosure is performed.

[0107] As an example, electronic device 800 may be a PC, tablet, personal digital assistant, smartphone, or other device capable of executing the aforementioned set of instructions. Here, electronic device 800 is not necessarily a single electronic device, but may be any collection of devices or circuits capable of executing the aforementioned instructions (or instruction sets) individually or in combination. Electronic device 800 may also be part of an integrated control system or system manager, or may be configured to interconnect with a portable electronic device locally or remotely (e.g., via wireless transmission) through an interface.

[0108] In electronic device 800, processor 802 may include a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller, or a microprocessor. By way of example and not limitation, processor may also include analog processors, digital processors, microprocessors, multi-core processors, processor arrays, network processors, etc.

[0109] The processor 802 can execute instructions or code stored in the memory 801, which can also store data. Instructions and data can also be sent and received over a network via a network interface device, which can employ any known transmission protocol.

[0110] The memory 801 may be integrated with the processor 802, for example, by placing RAM or flash memory within an integrated circuit microprocessor. Alternatively, the memory 801 may include a separate device, such as an external disk drive, a storage array, or other storage device that can be used by any database system. The memory 801 and the processor 802 may be operatively coupled, or may communicate with each other, for example, via I / O ports, network connections, etc., enabling the processor 802 to read files stored in the memory.

[0111] In addition, the electronic device 800 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, mouse, touch input device, etc.). All components of the electronic device 800 can be interconnected via a bus and / or network.

[0112] According to exemplary embodiments of the present disclosure, a computer-readable storage medium storing instructions may also be provided, wherein when the instructions are executed by at least one computing device, they cause at least one computing device to perform the above-described method for predicting the total power generation of the entire network.

[0113] Examples of computer-readable storage media herein include: read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disc storage, hard disk drive (HDD), solid-state drive (SSD), card storage (such as multimedia cards, secure digital (SD) cards, or ultra-fast digital (XD) cards), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state drive, and any other device configured to store a computer program and any associated data, data files, and data structures in a non-transitory manner and to provide the computer program and any associated data, data files, and data structures to a processor or computer so that the processor or computer can execute the computer program. The computer program in the aforementioned computer-readable storage medium can run in an environment deployed in computer devices such as clients, hosts, agent devices, servers, etc. Furthermore, in one example, the computer program and any associated data, data files, and data structures are distributed across a networked computer system, such that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed manner through one or more processors or computers. It should be noted that the instructions can also be used to perform additional steps in addition to those described above, or to perform more specific processing while performing the above steps. The details of these additional steps and further processing have already been mentioned in the description of the relevant methods, so they will not be repeated here to avoid repetition.

[0114] Another embodiment of this disclosure relates to a system comprising at least one computing device and at least one storage device for storing instructions, wherein the instructions, when executed by at least one computing device, cause at least one computing device to perform the above-described method for predicting the total power generation of the entire network.

[0115] It should be noted that the system according to the exemplary embodiments of this disclosure may rely entirely on the operation of computer programs or instructions to achieve the corresponding functions. That is, each unit corresponds to each step in the functional architecture of the computer program, so that the entire system is called through a special software package (e.g., a lib library) to achieve the corresponding functions.

[0116] On the other hand, when the aforementioned system is implemented as software, firmware, middleware, or microcode, the program code or code segment used to perform the corresponding operation can be stored in a computer-readable medium such as a storage medium, so that at least one processor or at least one computing device can perform the corresponding operation by reading and running the corresponding program code or code segment.

[0117] According to exemplary embodiments of this disclosure, the storage device may be integrated with the computing device, for example, by arranging RAM or flash memory within an integrated circuit microprocessor. Alternatively, the storage device may include a separate device, such as an external disk drive, a storage array, or other storage device usable by any database system. The storage device and the computing device may be operatively coupled, or may communicate with each other, for example, via I / O ports, network connections, etc., enabling the computing device to read instructions stored in the storage device.

[0118] Another embodiment of this disclosure relates to a computer program product, including a computer program / instructions that, when executed by a processor, implement the method for predicting the total power generation of the entire network as described in any of the above embodiments.

[0119] According to the prediction method, device, electronic equipment, storage medium and computer program product of the total power generation of the whole network provided in this disclosure, it is possible to predict the power output of the whole network's new energy by constructing intermediate targets and mining the missing undetermined electric field information, based on some known electric field data and the actual total power output of the whole network. Furthermore, this disclosure can mine the correlation between electric fields while modeling single electric fields, thereby reducing the error accumulation caused by the summation method of single electric field prediction and improving the accuracy of the prediction of the power generation of the whole network's new energy.

[0120] The foregoing has described various exemplary embodiments of this disclosure. It should be understood that the foregoing description is exemplary only and not exhaustive, and this disclosure is not limited to the disclosed exemplary embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. A method for predicting the total power generation of the entire power grid, characterized in that, The method for predicting the total power generation of the entire network includes: Obtain the total power generation of the entire network and the power generation of multiple known electric fields, wherein all electric fields in the entire network other than at least one of the multiple known electric fields are undetermined electric fields; Calculate the first correlation coefficient between the total power generation of the entire network and the sum of the power generation of the multiple known electric fields; In response to the first correlation coefficient being greater than or equal to a first preset value, the predicted value of the total power generation of the entire network is calculated as the ratio of the predicted value of the total power generation of the entire network to the sum of the predicted values ​​of the power generation of the at least one known electric field. Based on the product of the sum of the predicted values ​​of the power generation of the at least one known electric field and the ratio, the predicted value of the total power generation of the entire network is obtained. In response to the first correlation coefficient being less than the first preset value, a predicted value of the power generation of the undetermined electric field is predicted, and based on the sum of the predicted value of the power generation of the undetermined electric field and the predicted value of the power generation of at least one known electric field, a predicted value of the total power generation of the entire network is obtained.

2. The method for predicting the total power generation of the entire network as described in claim 1, characterized in that, The ratio of the predicted value of the total power generation of the entire network to the sum of the predicted values ​​of the power generation of the at least one known electric field includes: Obtain the temporal and meteorological characteristics of the area where the at least one known electric field is located; A first model is established between the temporal and meteorological characteristics of the region where the at least one known electric field is located, the total power generation of the at least one known electric field, and the ratio of the total power generation of the entire grid to the total power generation of the at least one known electric field; The ratio of the predicted total power generation of the entire network to the sum of the predicted power generation of the at least one known electric field, obtained based on the first model.

3. The method for predicting the total power generation of the entire network as described in claim 1, characterized in that, The predicted value of the power generation of the undetermined electric field includes: The power generation of the undetermined electric field is obtained based on the difference between the total power generation of the entire network and the sum of the power generation of the at least one known electric field. Calculate the second correlation coefficient between the sum of the power generated by the at least one known electric field and the power generated by the undetermined electric field; Determine a plurality of known electric fields in which the second correlation coefficient is greater than a second preset value among the at least one known electric fields; Obtain the meteorological characteristics of the area where the unknown electric field is located; A second model is established between the sum of the power generation of the known electric fields and the meteorological characteristics of the region where the unknown electric field is located, and the power generation of the unknown electric field. The predicted power generation value of the undetermined electric field is obtained based on the second model.

4. The method for predicting the total power generation of the entire network as described in claim 1, characterized in that, Obtain the power generation of multiple known electric fields, including: Obtain the power generation data of each of the multiple known electric fields; The power generation data of each known electric field are subjected to quality inspection to determine whether there are abnormal electric fields in the power generation data. Data filling is performed on the power generation data of the abnormal electric field; Based on the power generation data of each known electric field after data filling, the power generation of the multiple known electric fields is obtained.

5. The method for predicting the total power generation of the entire network as described in claim 4, characterized in that, Data filling is performed on the power generation data of the abnormal electric field, including: The power generation data of each known electric field are normalized based on the installed capacity of each field. Calculate the third correlation coefficient among the power generation data after normalization of each known electric field; From the known electric fields of the plurality of known electric fields, determine the reference electric field with the largest third correlation coefficient with the anomalous electric field; Based on the installed capacity of the abnormal electric field, the normalized power generation data of the reference electric field is denormalized to obtain the filled data; The power generation data of the abnormal electric field is populated based on the filling data.

6. The method for predicting the total power generation of the entire network as described in claim 1, characterized in that, The method further includes: Based on the accuracy of historical power predictions of the plurality of known electric fields, at least one known electric field among the plurality of known electric fields is determined.

7. The method for predicting the total power generation of the entire network as described in claim 3, characterized in that, The undetermined electric field includes multiple undetermined electric fields. The step of obtaining the meteorological characteristics of the area where the unknown electric field is located includes: Obtain the first meteorological characteristics of the area where the multiple undetermined electric fields are located; Based on the first meteorological characteristics of the areas where the multiple undetermined electric fields are located, the multiple undetermined electric fields are divided into multiple predetermined areas; The second meteorological characteristics of the multiple predetermined areas are obtained as the meteorological characteristics of the area where the undetermined electric field is located.

8. The method for predicting the total power generation of the entire network as described in claim 7, characterized in that, Both the first meteorological feature and the second meteorological feature are either wind speed features or both are radiation features.

9. A device for predicting the total power generation of the entire power grid, characterized in that, The device for predicting the total power generation of the entire network includes: The data acquisition unit is configured to acquire the total power generation of the entire network and the power generation of multiple known electric fields, wherein all electric fields in the entire network except for at least one of the known electric fields are undetermined electric fields; The coefficient calculation unit is configured to: calculate the first correlation coefficient between the total power generation of the entire network and the sum of the power generation of the multiple known electric fields; The power prediction unit is configured as follows: In response to the first correlation coefficient being greater than or equal to a first preset value, the predicted value of the total power generation of the entire network is calculated as the ratio of the predicted value of the total power generation of the entire network to the sum of the predicted values ​​of the power generation of the at least one known electric field. Based on the product of the sum of the predicted values ​​of the power generation of the at least one known electric field and the ratio, the predicted value of the total power generation of the entire network is obtained. In response to the first correlation coefficient being less than the first preset value, a predicted value of the power generation of the undetermined electric field is predicted, and based on the sum of the predicted value of the power generation of the undetermined electric field and the predicted value of the power generation of at least one known electric field, a predicted value of the total power generation of the entire network is obtained.

10. An electronic device, characterized in that, include: At least one processor; At least one memory that stores computer-executable instructions. Wherein, when the computer-executable instructions are executed by the at least one processor, the at least one processor causes the at least one processor to execute the method for predicting the total power generation of the entire network as described in any one of claims 1-8.

11. A computer-readable storage medium for storing instructions, characterized in that, When the instruction is executed by at least one processor, it causes the at least one processor to perform the method for predicting the total power generation of the entire network as described in any one of claims 1-8.

12. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the method for predicting the total power generation of the entire network as described in any one of claims 1-8.

Images