Classified statistical accounting method for regional electricity sales

Abstract

The invention relates to a classified statistical accounting method for regional electricity sales, and the method comprises the steps: obtaining the real-time data of the regional electricity sales, carrying out the processing of the real-time data of the regional electricity sales, and determining the feature data of the regional electricity sales; performing category division on the regional electricity sale quantity characteristic data, and determining regional electricity sale quantity characteristic data of sub-industries; and performing statistical accounting on the regional power sale quantity characteristic data of the sub-industries, and determining regional power sale quantity accounting analysis results of the sub-industries. The method solves the problems that the power supply reliability of a power grid system is low and the power supply quality of a power grid is reduced due to the fact that the regional power sale quantity cannot be well classified, counted and calculated in the prior art. According to the method, the regional electricity sale quantity can be well classified, counted and accounting, the regional electricity sale quantity can be visually displayed in real time, the regional electricity sale quantity can be intelligently controlled in time, the power supply reliability of a power grid system is improved, and the power supply quality of a power grid can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of regional electricity sales, and in particular to a classification statistical accounting method for regional electricity sales. Background Art

[0002] Electricity sales refers to the electricity sold by power companies to users (including wholesalers) and the electricity used by the company's non-power production, capital construction, overhaul and non-production departments (such as canteens and dormitories). With the continuous development of the power market, the classification and statistical accounting of regional electricity sales has become an important research direction. Through the classification and statistical accounting of regional electricity sales, we can better understand the electricity demand and supply situation in various regions, and provide scientific data support for policy making.

[0003] The Chinese patent with publication number CN105096159A discloses a method and device for predicting regional electricity sales, wherein the method includes: determining the time domain and frequency domain characteristics of the historical electricity sales curve of each area to be predicted, and clustering each area to be predicted accordingly, within the same cluster group, the historical electricity sales curve of each area to be predicted has the same characteristics in the time domain and frequency domain, and the characteristics are used as the characteristic labels of the cluster group, and finally, for each cluster group, a preset prediction algorithm corresponding to the characteristic label of the cluster group is selected to predict the electricity sales of each area to be predicted in the cluster group; by clustering different areas to be predicted, the areas with the same historical electricity sales curve characteristics are uniformly predicted using the same matching prediction algorithm, thereby improving the accuracy of regional electricity sales prediction; however, the patent has the following defects:

[0004] The existing system cannot properly classify and calculate regional electricity sales, and cannot intuitively display regional electricity sales in real time, resulting in the inability to timely and intelligently control regional electricity sales, which makes the power supply reliability of the power grid system low and reduces the power supply quality of the power grid. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a classified statistical accounting method for regional electricity sales, which can better classify and statistically calculate regional electricity sales, can intuitively display regional electricity sales in real time, and can timely perform intelligent management and control of regional electricity sales, thereby increasing the power supply reliability of the power grid system and improving the power supply quality of the power grid.

[0006] The present invention solves the technical problem by adopting the following technical solutions:

[0007] The classification statistical accounting method of regional electricity sales includes the following steps:

[0008] S1: Acquire the real-time data of regional electricity sales, process the real-time data of regional electricity sales, and determine the characteristic data of regional electricity sales;

[0009] S2: According to the industry classification standards, the regional electricity sales characteristic data is divided into different electricity consumption categories to determine the regional electricity sales characteristic data by industry;

[0010] S3: Statistically calculate the characteristic data of regional electricity sales by industry, determine the calculation and analysis results of regional electricity sales by industry, and display them intuitively in real time in a visual form.

[0011] Moreover, the specific implementation method for obtaining the real-time data of regional electricity sales in step 1 is: real-time monitoring of the electricity consumption time in the region to determine the regional electricity consumption time data; real-time monitoring of the electricity consumption category in the region to determine the regional electricity consumption category data; real-time monitoring of the electricity consumption season in the region to determine the regional electricity consumption season data; real-time monitoring of the electricity consumption quantity in the region to determine the regional electricity consumption quantity data; based on the regional electricity consumption time data, regional electricity consumption category data, regional electricity consumption season data and regional electricity consumption quantity data, determine the real-time data of regional electricity sales.

[0012] Moreover, the specific implementation method of processing the regional power sales real-time data in step 1 to determine the regional power sales characteristic data is: obtaining the regional power sales real-time data; retrieving the regional power sales real-time data based on the sequential retrieval method; performing consistency check on the regional power sales real-time data; checking whether the regional power sales real-time data meets the requirements according to the reasonable value range and mutual relationship of each variable in the regional power sales real-time data; removing inconsistent data that exceeds the normal range, is logically unreasonable or contradictory in the regional power sales real-time data; processing invalid values ​​and missing values ​​in the regional power sales real-time data; removing invalid data and missing data that are of no value to the regional power sales classification statistics and accounting in the regional power sales real-time data; determining the regional power sales real-time data that is valuable to the regional power sales classification statistics and accounting; sorting the regional power sales real-time data that is valuable to the regional power sales classification statistics and accounting based on the internal sorting method; determining the regional power sales real-time data with an arrangement order; performing feature extraction on the regional power sales real-time data with an arrangement order; and determining the regional power sales characteristic data.

[0013] Moreover, the specific implementation method of step 2 is: obtaining regional electricity sales characteristic data; and dividing the regional electricity sales characteristic data into different electricity consumption categories according to the industry classification standard and the electricity consumption characteristics.

[0014] Moreover, the specific implementation method of dividing the regional electricity sales characteristic data into different electricity consumption categories is: indexing the regional electricity sales characteristic data one by one; comparing and analyzing the regional electricity sales characteristic data indexed one by one with the industry classification standards; determining the electricity consumption category of the regional electricity sales characteristic data indexed one by one; determining the regional electricity sales characteristic data of different electricity consumption categories; uniformly storing the regional electricity sales characteristic data with the same electricity consumption category; and determining the regional electricity sales characteristic data by industry.

[0015] Moreover, the specific implementation method of step 3 is: according to the statistical accounting requirements of regional electricity sales, a large amount of historical regional electricity sales data that can reflect the characteristics of regional electricity sales are collected; the collected historical regional electricity sales data are divided to determine the training set and the test set; according to the statistical accounting requirements of regional electricity sales, a model architecture suitable for statistical accounting of regional electricity sales is selected; based on the training set, the selected model architecture suitable for statistical accounting of regional electricity sales is trained; a regional electricity sales statistical accounting model is determined; the regional electricity sales statistical accounting model is tested and optimized; and the optimal regional electricity sales statistical accounting model is determined.

[0016] Moreover, the selected model architecture suitable for statistical accounting of regional electricity sales, and the specific implementation method of adjusting the corresponding proportion of the number of samples in the training set according to the complexity of the model architecture before model architecture training is: selecting a model architecture suitable for statistical accounting of regional electricity sales; extracting the number of input units and the number of output units of the fully connected layer contained in the selected model architecture; obtaining the complexity evaluation parameter corresponding to the model architecture according to the number of input units and the number of output units of the fully connected layer contained in the model architecture:

[0017] ;

[0018] in, represents the complexity evaluation parameter; m represents the total number of fully connected layers; C ri represents the number of input units of the i-th fully connected layer; C si Represents the number of output units of the i-th fully connected layer; Represents the number of weight parameters of the i-th fully connected layer; Represents the number of weight parameters of the i-1th fully connected layer; and Represents the maximum and minimum values ​​of the number of weight parameters corresponding to n fully connected layers; Represents the average number of weight parameters corresponding to m fully connected layers;

[0019] The complexity evaluation parameter is used to adjust the corresponding ratio of the number of samples in the training set.

[0020] Moreover, the specific implementation method of adjusting the corresponding proportion of the sample quantity of the training set by using the complexity evaluation parameter is: extracting the historical data of the regional power sales volume to divide and determine the initial proportion of the training set corresponding to the training set and the test set; extracting the data amount corresponding to the invalid data in the sample data in the training set; obtaining the data quality evaluation coefficient corresponding to the training set according to the data amount corresponding to the invalid data in the sample data in the training set:

[0021] ;

[0022] Where f represents the data quality evaluation coefficient; n represents the number of data types contained in the invalid data; C i Indicates the amount of invalid data corresponding to the i-th data type; C zi represents the total amount of historical data corresponding to the i-th data type; w rpi represents the average value of the data weight values ​​corresponding to all the data contained in the invalid data in the i-th data type; w zi Represents the median value of the data weight values ​​corresponding to all the data contained in the i-th data type;

[0023] Extracting the complexity evaluation parameter;

[0024] Using the complexity evaluation parameter and the data quality evaluation coefficient, the initial proportion corresponding to the training set is adjusted to obtain the proportion corresponding to the adjusted number of samples in the training set;

[0025] Among them, the corresponding ratio of the adjusted sample quantity of the training set is:

[0026] ;

[0027] Among them, B t Indicates the proportion of the adjusted sample size in the training set; B 0 Indicates the initial proportion corresponding to the training set; represents the complexity evaluation parameter; f represents the data quality evaluation coefficient; k represents the preset compensation coefficient.

[0028] Moreover, the specific implementation method for testing and optimizing the regional electricity sales statistical accounting model is: obtaining the regional electricity sales statistical accounting model; performing performance testing on the regional electricity sales statistical accounting model based on a test set; determining the performance test results based on the regional electricity sales statistical accounting model; conducting in-depth mining and analysis on the performance test results based on the regional electricity sales statistical accounting model; determining an optimization adjustment plan based on the regional electricity sales statistical accounting model; optimizing and adjusting the regional electricity sales statistical accounting model based on the optimization adjustment plan; and determining the optimal regional electricity sales statistical accounting model.

[0029] The advantages and positive effects of the present invention are:

[0030] The present invention obtains the real-time data of regional electricity sales, processes the real-time data of regional electricity sales, determines the characteristic data of regional electricity sales, divides the characteristic data of regional electricity sales into different electricity consumption categories according to the industry classification standard, determines the characteristic data of regional electricity sales by industry, statistically calculates the characteristic data of regional electricity sales by industry, determines the calculation and analysis results of regional electricity sales by industry, intuitively displays them in real time in a visualized form, and intelligently controls the regional electricity sales based on the regional electricity sales control plan, can better classify and statistically calculate the regional electricity sales, can intuitively display the regional electricity sales in real time, and can timely intelligently control the regional electricity sales, thereby increasing the power supply reliability of the power grid system and improving the power supply quality of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described in detail below with reference to the accompanying drawings.

[0033] In order to solve the existing problem that the regional electricity sales cannot be well classified and statistically calculated, the regional electricity sales cannot be intuitively displayed in real time, resulting in the inability to timely intelligently control the regional electricity sales, making the power supply reliability of the power grid system low and reducing the power supply quality of the power grid, such as Figure 1 As shown, this embodiment provides the following technical solutions:

[0034] The classification statistical accounting method of regional electricity sales includes the following steps:

[0035] S1: Acquire the real-time data of regional electricity sales, process the real-time data of regional electricity sales, and determine the characteristic data of regional electricity sales;

[0036] In this embodiment, as a preferred technical solution of the present invention, real-time data of regional electricity sales is obtained by performing the following operations:

[0037] Monitor the electricity usage time in the area in real time and determine the regional electricity usage time data;

[0038] Conduct real-time monitoring of the electricity consumption categories in the region and determine the regional electricity consumption category data;

[0039] Conduct real-time monitoring of the electricity consumption season in the region and determine the regional electricity consumption season data;

[0040] Monitor the amount of electricity used in the area in real time and determine the regional electricity consumption data;

[0041] Among them, the real-time data of regional electricity sales volume is determined based on the regional electricity consumption time data, regional electricity consumption category data, regional electricity consumption season data and regional electricity consumption quantity data.

[0042] It should be noted that by obtaining real-time data on regional electricity sales, it is convenient to grasp the real-time situation of regional electricity sales, and facilitate the subsequent classification, statistics and accounting of regional electricity sales.

[0043] In this embodiment, as a preferred technical solution of the present invention, the real-time data of regional power sales is processed, and the following operations are performed:

[0044] Obtain real-time data on regional electricity sales;

[0045] Based on the sequential retrieval method, the real-time data of regional electricity sales is retrieved;

[0046] Conduct consistency checks on real-time data on regional electricity sales;

[0047] Check whether the real-time data of regional electricity sales meets the requirements according to the reasonable value range and mutual relationship of each variable in the real-time data of regional electricity sales;

[0048] Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of regional electricity sales;

[0049] Process invalid and missing values ​​in the real-time data of regional electricity sales;

[0050] Remove invalid data and missing data that are of no value to the classification and statistical calculation of regional electricity sales in the real-time data of regional electricity sales;

[0051] Determine the real-time data of regional electricity sales which is valuable for regional electricity sales classification statistics and accounting;

[0052] It should be noted that by searching the real-time data on regional electricity sales, inconsistent data, invalid data and missing data can be removed, and the real-time data on regional electricity sales that is valuable for the classification and statistical accounting of regional electricity sales can be determined, which can improve the processing speed and efficiency of subsequent real-time data on regional electricity sales.

[0053] Based on the internal sorting method, the regional electricity sales classification statistics and accounting valuable regional electricity sales real-time data are sorted;

[0054] Determine the real-time data of regional electricity sales with ranking order;

[0055] It should be noted that by sorting the valuable real-time data of regional electricity sales volume classified and statistically calculated, the real-time data of regional electricity sales volume with an arrangement order can be determined, which is convenient for quickly finding the rules of the real-time data of regional electricity sales volume.

[0056] Extract features from real-time data of regional electricity sales with ranking order;

[0057] Determine the regional electricity sales characteristic data.

[0058] It should be noted that by extracting features from the real-time regional electricity sales data with an arrangement order, the regional electricity sales characteristic data can be determined, which facilitates the subsequent classification of the regional electricity sales characteristic data into different electricity consumption categories.

[0059] S2: According to the industry classification standards, the regional electricity sales characteristic data is divided into different electricity consumption categories to determine the regional electricity sales characteristic data by industry;

[0060] In this embodiment, as a preferred technical solution of the present invention, the regional electricity sales characteristic data is divided into different electricity consumption categories, and the following operations are performed:

[0061] Obtain regional electricity sales characteristic data;

[0062] According to the industry classification standards, the regional electricity sales characteristic data is divided into different electricity consumption categories according to the electricity consumption characteristics;

[0063] Among them, different electricity consumption categories include residential electricity consumption, commercial electricity consumption, and industrial electricity consumption;

[0064] Among them, the regional electricity sales characteristic data is divided into different electricity consumption categories, including:

[0065] Index the regional electricity sales characteristic data one by one;

[0066] Compare and analyze the regional electricity sales characteristic data indexed one by one with the industry classification standards;

[0067] Determine the electricity consumption category of the regional electricity sales characteristic data indexed one by one;

[0068] Determine the regional electricity sales characteristic data for different electricity consumption categories;

[0069] The characteristic data of regional electricity sales with the same electricity consumption category are stored uniformly;

[0070] Determine the regional electricity sales characteristic data for each industry.

[0071] It should be noted that according to the industry classification standards, the regional electricity sales characteristic data can be divided into different electricity consumption categories according to the electricity consumption characteristics, so that the regional electricity sales characteristic data of each industry can be determined, which is convenient for the subsequent statistical calculation of the regional electricity sales characteristic data of each industry.

[0072] S3: Statistically calculate the characteristic data of regional electricity sales by industry, determine the calculation and analysis results of regional electricity sales by industry, and display them intuitively in real time in a visual form.

[0073] In this embodiment, as a preferred technical solution of the present invention, the calculation and analysis results of regional electricity sales by industry are determined, and the following operations are performed:

[0074] Obtain regional electricity sales characteristic data by industry;

[0075] Conduct statistical accounting of regional electricity sales characteristic data by industry;

[0076] Determine the regional electricity sales accounting data by industry to reflect the electricity consumption behavior of regional electricity sales by industry;

[0077] Conduct in-depth mining and relevant analysis of regional electricity sales accounting data by industry;

[0078] Mining and analyzing the potential electricity consumption patterns and trends of regional electricity sales accounting data by industry;

[0079] Determine the calculation and analysis results of regional electricity sales by industry;

[0080] The calculation and analysis results of regional electricity sales by industry are organized into reports or charts, which are displayed intuitively in real time in a visual form;

[0081] Formulate regional electricity sales control plans based on the regional electricity sales accounting and analysis results by industry;

[0082] Intelligently control regional electricity sales based on the regional electricity sales control solution.

[0083] It should be noted that by statistically calculating the characteristic data of regional electricity sales by industry, determining the calculation and analysis results of regional electricity sales by industry, and intuitively displaying them in real time in a visual form, and intelligently controlling the regional electricity sales based on the regional electricity sales control plan, the regional electricity sales can be better classified and statistically calculated, the regional electricity sales can be intuitively displayed in real time, and the regional electricity sales can be intelligently controlled in a timely manner, which increases the power supply reliability of the power grid system and improves the power supply quality of the power grid.

[0084] In this embodiment, as a preferred technical solution of the present invention, statistical calculation is performed on the regional power sales characteristic data of different industries, and the following operations are performed:

[0085] According to the statistical accounting requirements of regional electricity sales, a large amount of historical data on regional electricity sales that can reflect the characteristics of regional electricity sales are collected;

[0086] Divide the collected historical data of regional electricity sales into training sets and test sets;

[0087] According to the statistical accounting requirements of regional electricity sales, select a model architecture suitable for statistical accounting of regional electricity sales;

[0088] Based on the training set, the selected model architecture suitable for regional electricity sales statistics and accounting is trained;

[0089] Determine the statistical accounting model for regional electricity sales;

[0090] Test and optimize the statistical accounting model for regional electricity sales;

[0091] Determine the optimal statistical accounting model for regional electricity sales.

[0092] Specifically, a model architecture suitable for statistical accounting of regional electricity sales is selected, and the corresponding proportion of the number of samples in the training set is adjusted according to the complexity of the model architecture before training the model architecture, including:

[0093] The selected model architecture is suitable for statistical accounting of regional electricity sales;

[0094] Extracting the number of input units and the number of output units of the fully connected layer included in the selected model architecture;

[0095] The complexity evaluation parameter corresponding to the model architecture is obtained according to the number of input units and the number of output units of the fully connected layer included in the model architecture; wherein the complexity evaluation parameter is obtained by the following formula:

[0096] ;

[0097] in, represents the complexity evaluation parameter; m represents the total number of fully connected layers; C ri represents the number of input units of the i-th fully connected layer; C si Represents the number of output units of the i-th fully connected layer; Represents the number of weight parameters of the i-th fully connected layer; Represents the number of weight parameters of the i-1th fully connected layer; and Represents the maximum and minimum values ​​of the number of weight parameters corresponding to n fully connected layers; Represents the average number of weight parameters corresponding to m fully connected layers;

[0098] The complexity evaluation parameter is used to adjust the corresponding ratio of the number of samples in the training set.

[0099] The technical effects of the above technical solution are as follows: First, by selecting a model architecture suitable for statistical accounting of regional electricity sales, it is ensured that the model can accurately reflect the characteristics and laws of regional electricity sales, and improve the accuracy and reliability of statistical accounting. This targeted selection avoids the inadaptability that may be caused by using a general model, thereby improving the effect of the model in practical applications.

[0100] By extracting the number of input units and output units of the fully connected layer in the model architecture, and calculating the complexity evaluation parameters based on this information , providing a scientific basis for adjusting the number of samples before model training. This complexity evaluation not only takes into account the number and size of the fully connected layers, but also comprehensively evaluates the complexity of the model through the changes in the number of weight parameters (including maximum, minimum and average values), making the adjustment more precise and reasonable.

[0101] Using complexity evaluation parameters to adjust the proportion of the number of samples in the training set can effectively avoid overfitting or underfitting problems caused by the mismatch between the number of samples and the complexity of the model. When the model complexity is high, appropriately increasing the number of samples can improve the generalization ability of the model; conversely, when the model complexity is low, reducing the number of samples can speed up training and reduce computing costs. This dynamic adjustment strategy makes model training more efficient and stable.

[0102] Through the implementation of the above technical solutions, it is possible to ensure that the training set is scientifically and reasonably adjusted before model training, thereby avoiding various problems caused by inappropriate sample numbers during the training process. This not only improves the efficiency and stability of model training, but also makes the trained model more in line with actual needs, and improves the accuracy and reliability of regional electricity sales statistics.

[0103] Enhance the applicability and scalability of the model: This technical solution is not only applicable to the currently selected model architecture and regional electricity sales statistical accounting tasks, but can also be adjusted and optimized according to actual needs. For example, when it is necessary to handle more complex statistical accounting tasks or adopt a new model architecture, it is only necessary to recalculate the complexity evaluation parameters and adjust the corresponding proportion of the sample number, thereby enhancing the applicability and scalability of the model.

[0104] Specifically, the complexity evaluation parameter is used to adjust the corresponding ratio of the number of samples in the training set, including:

[0105] Extracting the historical data of power sales in the region, dividing the data and determining the initial ratios of the training set and the test set corresponding to the training set;

[0106] Extracting the amount of data corresponding to invalid data in the sample data in the training set;

[0107] The data quality evaluation coefficient corresponding to the training set is obtained according to the data amount corresponding to the invalid data in the sample data in the training set, wherein the data quality evaluation coefficient is obtained by the following formula:

[0108] ;

[0109] Where f represents the data quality evaluation coefficient; n represents the number of data types contained in the invalid data; C i Indicates the amount of invalid data corresponding to the i-th data type; C zi represents the total amount of historical data corresponding to the i-th data type; w rpi represents the average value of the data weight values ​​corresponding to all the data contained in the invalid data in the i-th data type; w zi Represents the median value of the data weight values ​​corresponding to all the data contained in the i-th data type;

[0110] Extracting the complexity evaluation parameter;

[0111] Using the complexity evaluation parameter and the data quality evaluation coefficient, the initial proportion corresponding to the training set is adjusted to obtain the proportion corresponding to the adjusted number of samples in the training set;

[0112] The corresponding ratio of the adjusted sample quantity of the training set is obtained by the following formula:

[0113] ;

[0114] Among them, B t Indicates the proportion of the adjusted sample size in the training set; B 0 Indicates the initial proportion corresponding to the training set; represents a complexity evaluation parameter; f represents a data quality evaluation coefficient; k represents a preset compensation coefficient, and the value range of the compensation coefficient is 0.03-0.12.

[0115] The technical effect of the above technical solution is: by introducing the data quality evaluation coefficient (f), the solution can carefully evaluate the impact of invalid data in the training set on the overall data quality. This helps to identify which data types or data subsets have more invalid data, so as to deal with these problems more targetedly. Improving data quality is an important guarantee for the effect of model training, because high-quality data can reduce noise and bias in the model training process.

[0116] The scheme not only considers the complexity of the model architecture (through the complexity evaluation parameter The data quality of the training set is also combined (measured by the data quality evaluation coefficient f). This dual consideration makes the adjustment of the proportion of the number of training set samples more comprehensive and scientific. Models with high complexity require more data to avoid overfitting, while low data quality may require the removal of some invalid data or the increase of the proportion of valid data to ensure the effectiveness of model training.

[0117] Through the preset compensation coefficient k, the scheme provides a flexible adjustment space for adjusting the proportion of the number of training set samples. The value range of the compensation coefficient (0.03-0.12) allows fine-tuning according to actual conditions to adapt to scenarios of different complexity and data quality. This flexibility helps to obtain the optimal proportion of the number of training set samples, thereby improving the efficiency and effectiveness of model training. By comprehensively considering the model complexity, data quality and the proportion of the number of samples, the scheme can ensure that the model is neither overfitting nor underfitting during the training process, thereby improving the stability and accuracy of the model. Stable models can generalize better to new data, while accurate models can provide more reliable prediction results.

[0118] In the case of low data quality or low model complexity, computing resources and time can be saved by reducing the number of samples in the training set. On the contrary, in the case of high data quality or high model complexity, appropriately increasing the number of samples in the training set can improve the training effect of the model. This strategy of optimizing resource utilization helps to achieve better economic and social benefits in practical applications. Since the scheme takes into account the dual factors of data quality and model complexity, and adjusts the corresponding proportion of the number of training set samples accordingly, the trained model has stronger adaptability and robustness. This model can better adapt to changes in different scenarios and needs, thereby improving its practicality and reliability in practical applications.

[0119] In this embodiment, as a preferred technical solution of the present invention, the regional electricity sales statistical accounting model is tested and optimized, and the following operations are performed:

[0120] Obtain the statistical accounting model for regional electricity sales;

[0121] Perform performance test on regional electricity sales statistics and accounting model based on the test set;

[0122] Determine the performance test results based on the regional electricity sales statistical accounting model;

[0123] Conduct in-depth analysis on the performance test results based on the regional electricity sales statistical accounting model;

[0124] Determine the optimization adjustment plan based on the regional electricity sales statistical accounting model;

[0125] Optimize and adjust the regional electricity sales statistics and accounting model based on the optimization adjustment plan;

[0126] Determine the optimal statistical accounting model for regional electricity sales.

[0127] In this embodiment, as a preferred technical solution of the present invention, statistical calculation is performed on the regional power sales characteristic data of different industries, and the following operations are further performed:

[0128] Obtain the optimal statistical accounting model for regional electricity sales;

[0129] Obtain regional electricity sales characteristic data by industry;

[0130] Based on the optimal regional electricity sales statistical accounting model, statistical accounting of regional electricity sales characteristic data by industry is performed;

[0131] Determine the regional electricity sales accounting data for each industry.

[0132] Therefore, by acquiring the real-time data of regional electricity sales, processing the real-time data of regional electricity sales, determining the characteristic data of regional electricity sales, dividing the characteristic data of regional electricity sales into different electricity consumption categories according to the industry classification standards, determining the characteristic data of regional electricity sales by industry, and performing statistical calculations on the characteristic data of regional electricity sales by industry, determining the calculation and analysis results of regional electricity sales by industry, and intuitively displaying them in real time in a visual form, and intelligently controlling the regional electricity sales based on the regional electricity sales control plan, the regional electricity sales can be better classified and statistically calculated, the regional electricity sales can be intuitively displayed in real time, and the regional electricity sales can be intelligently controlled in a timely manner, thereby increasing the power supply reliability of the power grid system and improving the power supply quality of the power grid.

[0133] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation manner. Any other implementation manners derived by those skilled in the art based on the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A classification statistical accounting method for regional electricity sales, characterized by: The following steps are involved: S1: Acquire the real-time data of regional electricity sales, process the real-time data of regional electricity sales, and determine the characteristic data of regional electricity sales; S2: According to the industry classification standards, the regional electricity sales characteristic data is divided into different electricity consumption categories to determine the regional electricity sales characteristic data by industry; S3: Statistically calculate the characteristic data of regional electricity sales by industry, determine the calculation and analysis results of regional electricity sales by industry, and display them intuitively in real time in a visual form.

2. The classification statistical accounting method of regional electricity sales according to claim 1 is characterized by: The specific implementation method of obtaining the real-time data of regional electricity sales in step 1 is: real-time monitoring of the electricity consumption time in the region to determine the regional electricity consumption time data; real-time monitoring of the electricity consumption category in the region to determine the regional electricity consumption category data; real-time monitoring of the electricity consumption season in the region to determine the regional electricity consumption season data; Monitor the amount of electricity used in the area in real time and determine the regional electricity consumption data; Based on the regional electricity consumption time data, regional electricity consumption category data, regional electricity consumption season data and regional electricity consumption quantity data, the real-time data of regional electricity sales volume is determined.

3. The classification statistical accounting method of regional electricity sales according to claim 2 is characterized by: The specific implementation method of processing the regional power sales real-time data in step 1 to determine the regional power sales characteristic data is: obtaining the regional power sales real-time data; searching the regional power sales real-time data based on the sequential retrieval method; performing consistency check on the regional power sales real-time data; checking whether the regional power sales real-time data meets the requirements according to the reasonable value range and mutual relationship of each variable in the regional power sales real-time data; Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of regional electricity sales; process invalid values ​​and missing values ​​in the real-time data of regional electricity sales; Remove invalid data and missing data that are of no value to the regional electricity sales classification statistics and accounting in the regional electricity sales real-time data; determine the regional electricity sales real-time data that are valuable to the regional electricity sales classification statistics and accounting; sort the regional electricity sales real-time data that are valuable to the regional electricity sales classification statistics and accounting based on the internal sorting method; Determine the real-time data of regional electricity sales with ranking order; Feature extraction is performed on the real-time data of regional electricity sales volume with an arrangement order; and characteristic data of regional electricity sales volume is determined.

4. The classification statistical accounting method of regional electricity sales according to claim 1 is characterized by: The specific implementation method of step 2 is: obtaining regional electricity sales characteristic data; and dividing the regional electricity sales characteristic data into different electricity consumption categories according to the industry classification standards and electricity consumption characteristics.

5. The classification statistical accounting method of regional electricity sales according to claim 4 is characterized by: The specific implementation method of dividing the regional electricity sales characteristic data into different electricity consumption categories is: indexing the regional electricity sales characteristic data one by one; comparing and analyzing the indexed regional electricity sales characteristic data one by one with the industry classification standard; determining the electricity consumption category of the indexed regional electricity sales characteristic data one by one; determining the regional electricity sales characteristic data of different electricity consumption categories; uniformly storing the regional electricity sales characteristic data with the same electricity consumption category; and determining the regional electricity sales characteristic data by industry.

6. The classification statistical accounting method of regional electricity sales according to claim 1 is characterized by: The specific implementation method of step 3 is: according to the statistical accounting requirements of regional electricity sales, a large amount of regional electricity sales history data that can reflect the characteristics of regional electricity sales are collected; the collected regional electricity sales history data are divided to determine the training set and the test set; According to the statistical accounting requirements of regional electricity sales, select a model architecture suitable for statistical accounting of regional electricity sales; Based on the training set, the selected model architecture suitable for regional electricity sales statistics and accounting is trained; the regional electricity sales statistics and accounting model is determined; the regional electricity sales statistics and accounting model is tested and optimized; and the optimal regional electricity sales statistics and accounting model is determined.

7. The classification statistical accounting method of regional electricity sales according to claim 6 is characterized by: The selected model architecture suitable for statistical accounting of regional electricity sales, and the specific implementation method of adjusting the corresponding proportion of the number of samples in the training set according to the complexity of the model architecture before model architecture training are: selecting a model architecture suitable for statistical accounting of regional electricity sales; extracting the number of input units and the number of output units of the fully connected layer contained in the selected model architecture; obtaining the complexity evaluation parameter corresponding to the model architecture according to the number of input units and the number of output units of the fully connected layer contained in the model architecture: ； in, represents the complexity evaluation parameter; m represents the total number of fully connected layers; C ri represents the number of input units of the i-th fully connected layer; C si Represents the number of output units of the i-th fully connected layer; Represents the number of weight parameters of the i-th fully connected layer; Represents the number of weight parameters of the i-1th fully connected layer; and Represents the maximum and minimum values ​​of the number of weight parameters corresponding to n fully connected layers; Represents the average number of weight parameters corresponding to m fully connected layers; The complexity evaluation parameter is used to adjust the corresponding ratio of the number of samples in the training set.

8. The classification statistical accounting method for regional electricity sales according to claim 7 is characterized by: The specific implementation method of adjusting the proportion of the sample quantity of the training set by using the complexity evaluation parameter is as follows: extracting the historical data of the regional power sales volume to divide and determine the initial proportion of the training set corresponding to the training set and the test set; extracting the amount of data corresponding to the invalid data in the sample data in the training set; and obtaining the data quality evaluation coefficient corresponding to the training set according to the amount of data corresponding to the invalid data in the sample data in the training set: ； Where f represents the data quality evaluation coefficient; n represents the number of data types contained in the invalid data; C i Indicates the amount of invalid data corresponding to the i-th data type; C zi represents the total amount of historical data corresponding to the i-th data type; w rpi represents the average value of the data weight values ​​corresponding to all the data contained in the invalid data in the i-th data type; w zi Represents the median value of the data weight values ​​corresponding to all the data contained in the i-th data type; Extracting the complexity evaluation parameter; Using the complexity evaluation parameter and the data quality evaluation coefficient, the initial proportion corresponding to the training set is adjusted to obtain the proportion corresponding to the adjusted number of samples in the training set; Among them, the corresponding ratio of the adjusted sample quantity of the training set is: ； Among them, B t represents the proportion of the adjusted sample size in the training set; B0 represents the initial proportion of the training set; represents the complexity evaluation parameter; f represents the data quality evaluation coefficient; k represents the preset compensation coefficient.

9. The classification statistical accounting method of regional electricity sales according to claim 6 is characterized by: The specific implementation method of testing and optimizing the regional electricity sales statistical accounting model is: obtaining the regional electricity sales statistical accounting model; performing performance testing on the regional electricity sales statistical accounting model based on the test set; Determine the performance test results based on the regional electricity sales statistical accounting model; Conduct in-depth analysis on the performance test results based on the regional electricity sales statistical accounting model; Determine the optimization adjustment plan based on the regional electricity sales statistical accounting model; Optimize and adjust the regional electricity sales statistics and accounting model based on the optimization adjustment plan; Determine the optimal statistical accounting model for regional electricity sales.

Images