CN118153737A - Method and device for predicting line loss of power distribution network - Google Patents

Abstract

The invention provides a method and a device for predicting line loss of a power distribution network, and belongs to the technical field of power engineering. The method comprises the following steps: acquiring a characteristic data set of a power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the period to be predicted; projecting the characteristic data set of the first time period under a first coordinate system to obtain a first test set, and projecting the characteristic data set of the second time period under a second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period; and inputting the weather difference values and the holiday difference values of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into a trained line loss prediction model, and performing line loss prediction on the power distribution network. The invention can improve the accuracy of line loss prediction.

Description

Method and device for predicting line loss of power distribution network

Technical Field

The invention relates to the technical field of power engineering, in particular to a method and a device for predicting line loss of a power distribution network.

Background

In power system development, line loss is a comprehensive index reflecting the level of power grid equipment, production operations, and management of a power supply enterprise. In the operation and maintenance process of the power station, accurate prediction of line loss is particularly important for the operation of the power station, not only affects the accuracy of electricity declaration of an industrial enterprise park, but also can provide necessary data support for reducing the line loss in the future.

At present, a mode of predicting the line loss of the power distribution network is generally adopted, wherein the line loss is predicted according to the historical total output power and the total power load, or the historical power attribute information is subjected to clustering division processing according to the preset radius, and the line loss is predicted through a clustering center of each cluster.

However, due to the large amount of power attribute information, how to improve the accuracy of the power distribution network line loss prediction is a technical problem that needs to be solved at present.

Disclosure of Invention

The embodiment of the invention provides a method and a device for predicting line loss of a power distribution network, which are used for solving the problem that the existing line loss prediction method is inaccurate.

In a first aspect, an embodiment of the present invention provides a method for predicting a line loss of a power distribution network, including:

acquiring a characteristic data set of a power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the period to be predicted, wherein the first period is a historical period adjacent to the period to be predicted for a preset interval, and the second period is a period in the same year as the previous year of the period to be predicted;

Projecting the characteristic data set of the first time period under a first coordinate system to obtain a first test set, and projecting the characteristic data set of the second time period under a second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period;

And inputting the weather difference values and the holiday difference values of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into a trained line loss prediction model, and performing line loss prediction on the power distribution network.

In one possible implementation, the feature data set of the first period includes n features;

projecting the feature data set of the first period of time into a first coordinate system to obtain a first test set, comprising:

performing standardization processing on a matrix formed by the characteristic data sets of the first time period, and determining a first divergence matrix of the characteristic data sets of the first time period after the standardization processing;

calculating the eigenvalue of the first divergence matrix and the eigenvector corresponding to each eigenvalue;

Based on the magnitude of the eigenvalues of the first divergence matrix, sorting the eigenvectors corresponding to all the eigenvalues from large to small, and forming a first coordinate system from the first n eigenvectors;

and performing dimension reduction processing on the feature data set of the first period after the normalization processing based on the first coordinate system to obtain a first test set.

In one possible implementation, the feature data set of the second period includes n features;

projecting the feature data set of the second period of time into a second coordinate system to obtain a second test set, comprising:

Performing standardization processing on a matrix formed by the characteristic data sets of the second time period, and determining a second divergence matrix of the characteristic data sets of the first time period after the standardization processing;

Calculating the eigenvalue of the second divergence matrix and the eigenvector corresponding to each eigenvalue;

based on the magnitude of the eigenvalues of the second divergence matrix, sorting the eigenvectors corresponding to all the eigenvalues from large to small, and forming a second coordinate system from the first n eigenvectors;

and performing dimension reduction processing on the feature data set of the second period after the normalization processing based on the second coordinate system to obtain a second test set.

In one possible implementation, the line loss prediction model comprises a prediction module and a correction module, wherein the correction module comprises a weather correction module and a holiday correction module;

The weather correction module is used for determining a first correction factor based on weather features of a period to be predicted, weather features of a second period and weather features of a first period;

The holiday correction module is used for determining a second correction factor based on holiday data of a period to be predicted and holiday data of a first period;

the first correction factor and the second correction factor are used for correcting the prediction result of the prediction module so as to obtain the final prediction result of the line loss prediction model.

In one possible implementation, the line loss prediction model is trained based on a historical data set of the power distribution network to be predicted for at least 2 years of history, the historical data set including actual line loss of the transformer area, meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity, and cable cross section.

In one possible implementation, the training process of the line loss prediction model is as follows:

Determining a first input data set and a second input data set of the training sample, wherein the first input data set is obtained based on the fact that a first historical data set is projected under a projection coordinate system of the first historical data set, and the projection coordinate system of the first historical data set is a preset number of feature vectors of a divergence matrix of the first historical data set; the second input data set is obtained based on the projection of the second historical data set under a projection coordinate system of the second historical data set, wherein the projection coordinate system of the second historical data set is a preset number of eigenvectors of a divergence matrix of the second historical data set; the first historical data set is a historical data set in a period adjacent to the prediction period by a preset interval, and the second historical data set is a historical data set of a period contemporaneous with the previous year of the prediction period;

determining a weather data difference set and a holiday difference set corresponding to the first input data set and the second input data set;

training the neural network model based on the first input data set, the second input data set, the meteorological data difference set and the holiday difference set to obtain a line loss prediction model.

In one possible implementation, the first input data set is based on normalizing the first historical data set, and forming the normalized data into a first training data set; calculating a divergence matrix of the first training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the first training data set, sorting the eigenvectors corresponding to all eigenvalues according to the order from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the first historical data set; performing dimension reduction processing on the first training data set based on a projection coordinate system of the first historical data set to obtain a first input data set;

The second input data set is based on normalization processing of the second historical data set, and the normalized data form a second training data set; calculating a divergence matrix of the second training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the second training data set, sorting the eigenvectors corresponding to all eigenvalues according to the order from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the second historical data set; and performing dimension reduction processing on the second training data set based on the projection coordinate system of the second historical data set to obtain a second input data set.

In one possible implementation manner, after obtaining the feature data sets of the power distribution network to be predicted in the first period and the second period, the method further includes:

And carrying out missing value and/or abnormal value processing on the characteristic data sets of the first time period and the second time period.

In one possible implementation, the feature data set includes meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity, and cable cross-section.

In a second aspect, an embodiment of the present invention provides a device for predicting a line loss of a power distribution network, including:

The acquisition module is used for acquiring characteristic data sets of the power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the period to be predicted, wherein the first period is a historical period adjacent to the period to be predicted for a preset interval, and the second period is a period contemporaneous with the previous year of the period to be predicted;

The projection module is used for projecting the characteristic data set of the first period to the first coordinate system to obtain a first test set, and projecting the characteristic data set of the second period to the second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period;

the prediction module is used for inputting the weather difference and the holiday difference of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into the trained line loss prediction model, and predicting the line loss of the power distribution network.

The embodiment of the invention provides a method and a device for predicting line loss of a power distribution network, wherein firstly, a characteristic data set of the power distribution network to be predicted in a first time period and a second time period, weather data and holiday data of the power distribution network to be predicted in the first time period are acquired, then, the characteristic data set of the first time period is projected to a first coordinate system to obtain a first test set, and the characteristic data set of the second time period is projected to a second coordinate system to obtain a second test set. And finally, inputting the weather difference values and the holiday difference values of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into a trained line loss prediction model, and carrying out line loss prediction on the power distribution network. The influence of weather and holidays on line loss can be fully considered by carrying out line loss prediction by adopting a trained line loss prediction model through the historical data adjacent to the period to be predicted, the historical data in the same period as the previous year of the period to be predicted and the weather difference value and holiday difference value of the period to be predicted and the second period and the first period respectively. In addition, the line loss of the period to be predicted is predicted by adopting the historical data adjacent to the period to be predicted and the historical data in the same period as the previous year of the period to be predicted, so that the line loss can be predicted more accurately.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments or the description of the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flowchart of an implementation of a method for predicting line loss of a power distribution network according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a power distribution network line loss prediction device according to an embodiment of the present invention.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the following description will be made by way of specific embodiments with reference to the accompanying drawings.

Fig. 1 is a flowchart of an implementation of a method for predicting line loss of a power distribution network according to an embodiment of the present invention, which is described in detail below:

Step S110, acquiring characteristic data sets of the power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the power distribution network to be predicted in the period.

The first time period is a historical time period adjacent to the time period to be predicted by a preset interval, and the second time period is a time period which is in the same year as the previous year of the time period to be predicted. The preset interval may be 6 months, 3 months, 1 month, or the same as the interval of the period to be predicted, which is not limited herein, and is less than 12 months. And the last period of the first period is a period immediately adjacent to the period at which the period to be predicted starts.

Line loss is a power loss phenomenon which cannot be avoided by a power distribution network, and is line loss, in particular to energy loss caused when electric energy flows through a power transmission line.

For example, the period to be predicted is 2024 years 4 months to 6 months, then the first period may be 2024 years 1 month to 3 months, and the second period may be 2023 years 4 months to 6 months.

The meteorological data can comprise temperature, humidity and wind force, and the meteorological data of the period to be predicted can be obtained through a meteorological prediction platform. Holidays include weekends and national legal holidays, and holiday data may be determined by a holiday calendar. Holiday and meteorological data can severely impact the load and power of a distribution network.

In some embodiments, to be able to accurately predict line losses of the power distribution network, the feature data set may include meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radii, distribution transformer capacity, and cable cross-sections. Each feature in the feature dataset has a certain correlation with line loss, and therefore line loss can be predicted from these features.

The meteorological data, the average running years of equipment, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section data of the first period and the second period are real data of historical actual measurement, and can be checked through historical records. The frequency of recording may be average data of each day or average data of multiple days, and may be set according to a specific application scenario, which is not limited herein.

In this embodiment, the meteorological data may include temperature, humidity, and wind.

In some embodiments, after the feature data sets of the first period and the second period and the weather data and holiday data of the period to be predicted are obtained, processing of these data, such as processing of missing values or processing of abnormal values, is also required to improve the accuracy of the original input data.

Step S120, projecting the feature data set of the first period under the first coordinate system to obtain a first test set, and projecting the feature data set of the second period under the second coordinate system to obtain a second test set.

The first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period. The second coordinate system is a preset number of eigenvectors of the divergence matrix formed by the characteristic data sets of the second period.

In some embodiments, since the feature data set includes not only a plurality of features, but also each feature includes data of a plurality of time periods, the feature data set is a high-dimensional data set, and if the high-dimensional data set is directly used, the analysis amount of the data is increased, and the prediction difficulty is increased. And the data relevance between each feature data in the feature data sets cannot be obtained, so that the accuracy of prediction is improved. In order to further mine the association of the feature data in the feature data set, extract accurate power fingerprint data for prediction, the feature data set needs to be processed, and the specific processing procedure is as follows:

For example, the feature data set of the first period includes 7 features, and the feature data set of the second period includes 7 features. These 7 features are respectively meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section. Each feature includes data for m periods.

Since the types of data in the feature data set are different, it is necessary to perform normalization processing on the feature data set first. And carrying out standardization processing on a matrix formed by the characteristic data sets of the first time period, and determining a first divergence matrix of the characteristic data sets of the first time period after the standardization processing.

For example, the feature data set m1= { x1, x2, x3, x4, x5, x6, x7} of the first period after the normalization process is performed, and covariance between feature data sets of each dimension is calculated to form a covariance matrix Cov (x 1, x2, x3, x4, x5, x6, x 7). Cov is a covariance calculation function representing the degree of correlation of two feature data.

Then, a first divergence matrix Q1, q1= (M-1) ×cov (x 1, x2, x3, x4, x5, x6, x 7) of the feature data set M1 of the first period after the normalization processing is determined, and M is data containing M periods for each feature.

Next, the eigenvalues of the first divergence matrix Q1 and the eigenvectors corresponding to each eigenvalue are calculated. And sorting the feature vectors corresponding to all the feature values according to the size from large to small based on the size of the feature values of the first divergence matrix Q1. The first 7 feature vectors are combined into a first coordinate system R1. Or the number of the feature values to be reserved can be determined through the accumulated contribution rate of the feature values, and the feature vector corresponding to the feature values is correspondingly selected according to the reserved feature values.

Finally, the feature data set M1 of the first period after the normalization processing is subjected to dimension reduction processing based on the first coordinate system R1, so as to obtain a first test set S1, s1=r1×m1.

Similarly, the feature data set m2= { y1, y2, y3, y4, y5, y6, y7} of the second period after the normalization processing calculates covariance between feature data sets of each dimension, and constitutes covariance matrix Cov (y 1, y2, y3, y4, y5, y6, y 7). Cov is a covariance calculation function representing the degree of correlation of two feature data.

Then, a second divergence matrix Q2, q2= (M-1) ×cov (y 1, y2, y3, y4, y5, y6, y 7) of the feature data set M2 of the second period after the normalization processing is determined, and M is data containing M periods for each feature.

Next, the eigenvalues of the second divergence matrix Q2 and the eigenvectors corresponding to each eigenvalue are calculated. And sorting the eigenvectors corresponding to all the eigenvalues from large to small based on the magnitude of the eigenvalues of the second divergence matrix Q2. The first 7 feature vectors are combined into a second coordinate system R2. Or the number of the feature values to be reserved can be determined through the accumulated contribution rate of the feature values, and the feature vector corresponding to the feature values is correspondingly selected according to the reserved feature values.

Finally, the feature data set M2 of the second period after the normalization processing is subjected to dimension reduction processing based on the second coordinate system R2, so as to obtain a second test set S2, s2=r2×m2.

And step 130, inputting the weather difference and the holiday difference of the first test set, the second test set and the period to be predicted and the second period and the first period respectively into a trained line loss prediction model, and predicting the line loss of the power distribution network.

In some embodiments, the line loss prediction model includes a prediction module and a correction module, the correction module including a weather correction module and a holiday correction module.

Since holiday and meteorological data can seriously affect the load and power of a power distribution network, so that the line loss can be greatly affected, the influence of holiday and meteorological data on line loss prediction needs to be considered.

In this embodiment, the weather modification module is configured to determine the first modification factor based on the weather characteristics of the period to be predicted, the weather characteristics of the second period, and the weather characteristics of the first period.

Specifically, a first difference between the first time period and the time period to be predicted is determined based on the weather features of the first time period and the weather features of the time period to be predicted. And determining a second difference value between the second period and the period to be predicted based on the weather features of the second period and the weather features of the period to be predicted. A first correction factor is then determined based on the first difference and the second difference. For example, different weights may be respectively assigned to the first difference and the second difference, and the respective weights of the first difference and the second difference may be obtained based on training of the training sample.

The meteorological features here include temperature, humidity and wind power, so that the first difference is a fitted curve of the temperature difference, the humidity difference and the wind power difference, and the second difference is also a fitted curve of the temperature difference, the humidity difference and the wind power difference.

In this embodiment, the holiday correction module is configured to determine the second correction factor based on holiday data of the period to be predicted, holiday data of the first period, and holiday data of the second period.

Specifically, a third difference value between the first period and the period to be predicted is determined based on the holiday data of the first period and the holiday data of the period to be predicted. And determining a fourth difference value between the second period and the period to be predicted based on the holiday data of the second period and the holiday data of the period to be predicted. A second correction factor is then determined based on the third difference and the fourth difference. For example, different weights may be respectively assigned to the third difference and the fourth difference, and the respective weights of the third difference and the fourth difference may be obtained based on training of the training sample.

In some embodiments, the line loss prediction model is trained based on a historical data set of the power distribution network to be predicted for at least 2 years of history. The historical data set includes actual line loss of the transformer area, meteorological data, equipment average operating life, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section.

For example, the actual line loss, meteorological data, the average running life of equipment, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section of a transformer area of the power distribution network to be predicted in the year 2020 from 1 month to 2022 are obtained.

A training set is constructed, the training set comprising training samples and test samples. The training set comprises a first historical data set, a second historical data set, a meteorological data difference set and a holiday difference set, wherein the meteorological data difference set comprises a first meteorological data difference set between a period to be predicted and the first historical data set, and a second meteorological data difference set between a period to be predicted and the second historical data set. Meteorological data and holidays for the period to be predicted are also known. The holiday difference comprises a first holiday difference set in the period to be predicted and the first historical data set, and a second holiday difference set in the period to be predicted and the second historical data set.

The time period corresponding to the first historical data set is a time period adjacent to the time period to be predicted, and the time period corresponding to the second historical data set is a time period of the same year as the previous year of the time period to be predicted.

For ease of understanding, the method of constructing one of the samples in the training set is described, and since the training set needs to include a plurality of samples, a detailed description thereof is omitted.

For example, the line loss to be predicted for the period of 2022, 11 months, and the first historical data are meteorological data for 2022, 10 months of equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section. The second historical data is meteorological data of month 11 of 2021, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable section. Then, based on the first history data, the second history data and the period to be predicted, a weather data difference value and a holiday difference value are determined. Note that, the weather data differences here include 2 weather data differences, including a difference of weather data of 2022 at 11 and weather data of 2022 at 10, and a difference of weather data of 2022 at 11 and weather data of 2021 at 11. Holiday differences also include 2 differences, a holiday difference of 2022 for month 11 meteorological data and 2022 for month 10, and a holiday difference of 2022 for month 11 meteorological data and 2021 for month 11.

Based on the same construction method, a plurality of groups of historical data are constructed to form a training sample, so that the training accuracy of the constructed line loss prediction model can be improved.

After the first historical data set and the second historical data set are obtained, normalization processing is needed to be carried out on the first historical data set and the second historical data set respectively, and the first training data set and the second training data set are obtained.

Then, a first training divergence matrix corresponding to the first training data set and a second training divergence matrix corresponding to the second training data set are respectively calculated. The calculating method of the divergence matrix is the same as that of the step S120, and will not be described here.

Then, the eigenvalues of the first training divergence matrix and the second training divergence matrix and the eigenvectors corresponding to the eigenvalues are respectively obtained. And sorting the eigenvalues of the first training divergence matrix and the eigenvalues of the second training divergence matrix from large to small respectively. Based on the ordering of the eigenvalues of the first training divergence matrix, the first 7 eigenvectors are selected to form a projection coordinate system of the first historical data set. And selecting the first 7 eigenvectors to form a projection coordinate system of the second historical data set based on the ordering of the eigenvalues of the second training divergence matrix.

And finally, performing dimension reduction processing on the first training data set based on the projection coordinate system of the first historical data set to obtain a first input data set, and performing dimension reduction processing on the second training data set based on the projection coordinate system of the second historical data set to obtain a second input data set.

After the first input data set, the second input data set, the weather data difference set and the holiday difference set are obtained, the data can be used for training the constructed neural network model.

The weather data difference sets comprise a first weather data difference set and a second weather data difference set, and the holiday difference sets comprise a first holiday difference set and a second holiday difference set.

The neural network model may include an input layer, a multi-layered LSTM unit, and an output layer, wherein the multi-layered LSTM unit includes a prediction module and a correction module, and the correction module includes a weather correction module and a holiday correction module.

Therefore, when the line loss is predicted, not only is the damage of the power data considered, but also the influence of weather, holidays and equipment aging on the line loss is fully considered, so that the line loss of the power distribution network can be predicted more accurately.

According to the prediction method provided by the embodiment of the invention, firstly, the characteristic data sets of the power distribution network to be predicted in the first time period and the second time period respectively, and the meteorological data and holiday data of the power distribution network to be predicted in the time period are obtained, then, the characteristic data sets of the first time period are projected to the first coordinate system to obtain the first test set, and the characteristic data sets of the second time period are projected to the second coordinate system to obtain the second test set. And finally, inputting the weather difference values and the holiday difference values of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into a trained line loss prediction model, and carrying out line loss prediction on the power distribution network. The influence of weather and holidays on line loss can be fully considered by carrying out line loss prediction by adopting a trained line loss prediction model through the historical data adjacent to the period to be predicted, the historical data in the same period as the previous year of the period to be predicted and the weather difference value and holiday difference value of the period to be predicted and the second period and the first period respectively. In addition, the line loss of the period to be predicted is predicted by adopting the historical data adjacent to the period to be predicted and the historical data in the same period as the previous year of the period to be predicted, so that the line loss can be predicted more accurately.

It should be understood that the sequence number of each step in the foregoing embodiment does not mean that the execution sequence of each process should be determined by the function and the internal logic, and should not limit the implementation process of the embodiment of the present invention.

The following are device embodiments of the invention, for details not described in detail therein, reference may be made to the corresponding method embodiments described above.

Fig. 2 is a schematic structural diagram of a power distribution network line loss prediction device according to an embodiment of the present invention, and for convenience of explanation, only the portions relevant to the embodiment of the present invention are shown, which are described in detail below:

As shown in fig. 2, the predicting device 200 for a power distribution network line loss includes:

The obtaining module 210 is configured to obtain a feature data set of the power distribution network to be predicted in a first period and a second period, and weather data and holiday data of the period to be predicted, where the first period is a historical period adjacent to the period to be predicted by a preset interval, and the second period is a period that is the same year as a previous year of the period to be predicted;

The projection module 220 is configured to project the feature data set of the first period into a first coordinate system to obtain a first test set, and project the feature data set of the second period into a second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period;

The prediction module 230 is configured to input the weather difference and the holiday difference between the first test set, the second test set, and the period to be predicted and the second period and the first period respectively to a trained line loss prediction model, and predict the line loss of the power distribution network.

In one possible implementation, the feature data set of the first period includes n features;

The projection module 220 is configured to perform normalization processing on a matrix formed by the feature data sets of the first period, and determine a first divergence matrix of the feature data sets of the first period after the normalization processing;

calculating the eigenvalue of the first divergence matrix and the eigenvector corresponding to each eigenvalue;

Based on the magnitude of the eigenvalues of the first divergence matrix, sorting the eigenvectors corresponding to all the eigenvalues from large to small, and forming a first coordinate system from the first n eigenvectors;

and performing dimension reduction processing on the feature data set of the first period after the normalization processing based on the first coordinate system to obtain a first test set.

In one possible implementation, the feature data set of the second period includes n features;

The projection module 220 is configured to perform normalization processing on a matrix formed by the feature data sets of the second period, and determine a second divergence matrix of the feature data sets of the first period after the normalization processing;

Calculating the eigenvalue of the second divergence matrix and the eigenvector corresponding to each eigenvalue;

based on the magnitude of the eigenvalues of the second divergence matrix, sorting the eigenvectors corresponding to all the eigenvalues from large to small, and forming a second coordinate system from the first n eigenvectors;

and performing dimension reduction processing on the feature data set of the second period after the normalization processing based on the second coordinate system to obtain a second test set.

In one possible implementation, the line loss prediction model includes a prediction module and a correction module, and the correction module includes a weather correction module and a holiday correction module;

The weather correction module is used for determining a first correction factor based on weather features of a period to be predicted, weather features of a second period and weather features of a first period;

The holiday correction module is used for determining a second correction factor based on holiday data of a period to be predicted and holiday data of a first period;

the first correction factor and the second correction factor are used for correcting the prediction result of the prediction module so as to obtain the final prediction result of the line loss prediction model.

In one possible implementation, the line loss prediction model is trained based on a historical data set of the power distribution network to be predicted for at least 2 years of history, the historical data set including actual line loss of the transformer area, meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity, and cable cross section.

In one possible implementation, the training process of the line loss prediction model is as follows:

Determining a first input data set and a second input data set of the training sample, wherein the first input data set is obtained based on the fact that a first historical data set is projected under a projection coordinate system of the first historical data set, and the projection coordinate system of the first historical data set is a preset number of feature vectors of a divergence matrix of the first historical data set; the second input data set is obtained based on the projection of the second historical data set under a projection coordinate system of the second historical data set, wherein the projection coordinate system of the second historical data set is a preset number of eigenvectors of a divergence matrix of the second historical data set; the first historical data set is a historical data set in a period adjacent to the prediction period by a preset interval, and the second historical data set is a historical data set of a period contemporaneous with the previous year of the prediction period;

determining a weather data difference set and a holiday difference set corresponding to the first input data set and the second input data set;

training the neural network model based on the first input data set, the second input data set, the meteorological data difference set and the holiday difference set to obtain a line loss prediction model.

In one possible implementation, the first input data set is based on normalizing the first historical data set, and forming the normalized data into a first training data set; calculating a divergence matrix of the first training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the first training data set, sorting the eigenvectors corresponding to all eigenvalues according to the order from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the first historical data set; performing dimension reduction processing on the first training data set based on a projection coordinate system of the first historical data set to obtain a first input data set;

The second input data set is based on normalization processing of the second historical data set, and the normalized data form a second training data set; calculating a divergence matrix of the second training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the second training data set, sorting the eigenvectors corresponding to all eigenvalues according to the order from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the second historical data set; and performing dimension reduction processing on the second training data set based on the projection coordinate system of the second historical data set to obtain a second input data set.

In a possible implementation manner, the acquiring module is configured to perform missing value and/or outlier processing on the feature data sets of the first period and the second period.

In one possible implementation, the feature data set includes meteorological data, equipment average operating years, holiday data, three-phase load data, power supply radius, distribution transformer capacity, and cable cross-section.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the templates, elements, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

The modules/units, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the flow of the method of the foregoing embodiment, or may be implemented by instructing related hardware by a computer program, where the computer program may be stored in a computer readable storage medium, and the computer program may implement the steps of the method embodiment of predicting line loss of each power distribution network when executed by a processor. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier wave signal, a telecommunication signal, a software distribution medium, and so forth.

The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.

Claims (10)

1. The method for predicting the line loss of the power distribution network is characterized by comprising the following steps of:

Acquiring characteristic data sets of a power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the period to be predicted, wherein the first period is a historical period adjacent to the period to be predicted for a preset interval, and the second period is a period contemporaneous with the previous year of the period to be predicted;

Projecting the characteristic data set of the first time period under a first coordinate system to obtain a first test set, and projecting the characteristic data set of the second time period under a second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period;

And inputting the weather difference values and the holiday difference values of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into a trained line loss prediction model, and performing line loss prediction on the power distribution network.

2. The method for predicting line loss of a power distribution network of claim 1, wherein the feature dataset of the first period of time includes n features;

The projecting the characteristic data set of the first period into a first coordinate system to obtain a first test set includes:

performing standardization processing on a matrix formed by the characteristic data sets of the first time period, and determining a first divergence matrix of the characteristic data sets of the first time period after the standardization processing;

calculating the eigenvalue of the first divergence matrix and the eigenvector corresponding to each eigenvalue;

based on the magnitude of the eigenvalues of the first divergence matrix, sorting all eigenvectors corresponding to each eigenvalue according to the sequence from large to small, and forming the first coordinate system by the first n eigenvectors;

And performing dimension reduction processing on the feature data set of the first period after the normalization processing based on the first coordinate system to obtain the first test set.

3. The method for predicting line losses of a power distribution network according to claim 1 or 2, wherein the feature data set of the second period of time includes n features;

The projecting the feature data set of the second period of time into a second coordinate system to obtain a second test set includes:

Performing standardization processing on a matrix formed by the characteristic data sets of the second time period, and determining a second divergence matrix of the characteristic data sets of the first time period after the standardization processing;

calculating the eigenvalue of the second divergence matrix and the eigenvector corresponding to each eigenvalue;

based on the magnitude of the eigenvalues of the second divergence matrix, sorting all eigenvectors corresponding to each eigenvalue according to the sequence from large to small, and forming the first n eigenvectors into the second coordinate system;

and performing dimension reduction processing on the feature data set of the second period after the standardization processing based on the second coordinate system to obtain the second test set.

4. The method for predicting line loss of a power distribution network according to claim 1, wherein the line loss prediction model comprises a prediction module and a correction module, and the correction module comprises a weather correction module and a holiday correction module;

the weather correction module is used for determining a first correction factor based on the weather characteristics of the period to be predicted, the weather characteristics of the second period and the weather characteristics of the first period;

the holiday correction module is used for determining a second correction factor based on the holiday data of the period to be predicted, the holiday data of the first period and the holiday data of the second period;

the first correction factor and the second correction factor are used for correcting the prediction result of the prediction module so as to obtain the final prediction result of the line loss prediction model.

5. The method for predicting line losses of a power distribution network according to claim 4, wherein the line loss prediction model is trained based on a historical dataset of at least 2 years of history of the power distribution network to be predicted, the historical dataset including actual line losses of a region, meteorological data, equipment average operating life, holiday data, three-phase load data, power supply radius, distribution transformer capacity and cable cross-section.

6. The method for predicting line loss of power distribution network according to claim 5, wherein the training process of the line loss prediction model is as follows:

Determining a first input data set and a second input data set of a training sample, wherein the first input data set is obtained based on the projection of a first historical data set under a projection coordinate system of the first historical data set, and the projection coordinate system of the first historical data set is a preset number of feature vectors of a divergence matrix of the first historical data set; the second input data set is obtained based on the fact that a second historical data set is projected under a projection coordinate system of the second historical data set, and the projection coordinate system of the second historical data set is a preset number of eigenvectors of a divergence matrix of the second historical data set; the first historical data set is a historical data set in a period adjacent to the prediction period by a preset interval, and the second historical data set is a historical data set of a period contemporaneous with the prediction period;

Determining a weather data difference set and a holiday difference set corresponding to the first input data set and the second input data set;

And training a neural network model based on the first input data set, the second input data set, the meteorological data difference set and the holiday difference set to obtain the line loss prediction model.

7. The method for predicting line loss of a power distribution network according to claim 6, wherein the first input data set is based on normalizing the first historical data set, and forming the normalized data into a first training data set; calculating a divergence matrix of the first training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the first training data set, sorting eigenvectors corresponding to all eigenvalues according to the sequence from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the first historical data set; performing dimension reduction processing on the first training data set based on a projection coordinate system of the first historical data set to obtain the first input data set;

The second input data set is based on normalization processing of the second historical data set, and the normalized data form a second training data set; calculating a divergence matrix of the second training data set, and solving a characteristic value and a characteristic vector of the divergence matrix; based on the magnitude of the eigenvalues of the divergence matrix of the second training data set, sorting eigenvectors corresponding to all eigenvalues according to the sequence from large to small, and forming the first 7 eigenvectors into a projection coordinate system of the second historical data set; and performing dimension reduction processing on the second training data set based on a projection coordinate system of the second historical data set to obtain the second input data set.

8. The method for predicting line loss of a power distribution network according to claim 1, wherein the obtaining the feature data sets of the power distribution network to be predicted after the first period and the second period respectively further comprises:

and carrying out missing value and/or abnormal value processing on the characteristic data sets of the first time period and the second time period.

9. The method of claim 1, wherein the characteristic data set includes meteorological data, equipment average operating life, holiday data, three-phase load data, power supply radius, distribution transformer capacity, and cable cross section.

10. The utility model provides a prediction device of distribution network line loss which characterized in that includes:

The power distribution network prediction system comprises an acquisition module, a prediction module and a prediction module, wherein the acquisition module is used for acquiring a characteristic data set of a power distribution network to be predicted in a first period and a second period respectively, and meteorological data and holiday data of the period to be predicted, wherein the first period is a historical period adjacent to the period to be predicted for a preset interval, and the second period is a period in the same year as the previous year of the period to be predicted;

The projection module is used for projecting the characteristic data set of the first time period under a first coordinate system to obtain a first test set, and projecting the characteristic data set of the second time period under a second coordinate system to obtain a second test set; the first coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the first time period, and the second coordinate system is a preset number of eigenvectors of a divergence matrix formed by the characteristic data sets of the second time period;

the prediction module is used for inputting the weather difference and the holiday difference of the first test set, the second test set and the period to be predicted, the second period and the first period respectively into the trained line loss prediction model, and predicting the line loss of the power distribution network.