Electricity consumption behavior analysis method considering customer electricity consumption and time-space separation of payment

Abstract

The invention relates to an electricity consumption behavior analysis method considering customer electricity consumption and time-space separation of payment, and the method comprises the specific steps: respectively collecting electricity consumption data and payment data of electricity consumption customers from each power system according to time periods and geographic positions, and the data comprise a time sequence and a spatial position; carrying out data preprocessing on the collected power consumption data and payment data, wherein the data preprocessing comprises data cleaning, data conversion and data reconstruction; integrating the preprocessed power consumption data and payment data with other data influencing the power consumption behavior to form a comprehensive data set; identifying a main component in the comprehensive data set by using an empirical orthogonal function, namely, a main power utilization mode; the method comprises the following steps: firstly, decomposing a comprehensive data set by using a spatial-temporal characteristic matrix, extracting characteristic values of power consumption mode data in the comprehensive data set, and sorting the characteristic values of the power consumption mode data from large to small;

Description

Technical Field

[0001] The present invention relates to the technical field of power marketing, and in particular to an electricity consumption behavior analysis method that takes into account the time and space separation of customers' electricity consumption and payment. Background Art

[0002] In the field of power marketing, analyzing the electricity consumption behavior of electricity users is an important means to optimize the allocation of power resources and improve the efficiency of power grid operations.

[0003] A substation is a basic unit in the power system, usually containing a certain number of electricity users. Through cluster analysis, customers with similar electricity consumption characteristics can be divided into different groups. However, traditional clustering methods are often limited to the analysis of a single time point or a single spatial dimension, and it is difficult to fully reveal the periodicity and spatial distribution patterns of electricity consumption behaviors of different types of customers.

[0004] As a large electricity consumer, the electricity consumption behavior of industrial parks has an important impact on the load management and power resource allocation of the power grid. Traditional forecasting methods, such as time series analysis or machine learning models, may be able to predict the electricity consumption of industrial parks to a certain extent. However, these methods often ignore the spatial distribution characteristics of electricity consumption behavior, as well as the mutual influence and synergy between different industrial parks.

[0005] The periodicity of electricity consumption behavior is an important aspect of power system analysis, which involves changes in electricity consumption patterns over days, weeks, months, seasons, and even years. For example, there may be significant differences in electricity consumption between weekdays and weekends, and seasonal changes may also affect electricity consumption. Traditional analysis methods may have difficulty accurately describing and predicting such periodic changes, especially when considering the interaction of multiple periodic factors.

[0006] The spatial distribution pattern of electricity consumption behavior is of great significance for the spatial planning of power grids and the spatial allocation of power resources. Different geographical locations, climate conditions, economic activities, and population density may affect the spatial distribution of electricity consumption behavior. Traditional analysis methods may not fully consider these spatial factors, resulting in insufficient understanding of the spatial distribution pattern of electricity consumption behavior.

[0007] The prior art CN118396326A discloses a system and method for predicting and dispatching residential electricity consumption based on big data. The system constructs an electricity consumption prediction neural network model and electricity consumption threshold, divides the electricity price of residential areas into grades according to time periods, optimizes the power supply mode of residential areas, and establishes a renewable backup power supply. At the same time, a variety of simulation scenarios of emergencies and power failures are preset, and emergency mechanism plans are established in advance for emergencies and power failures. When the neural network model predicts that the electricity consumption of a residential area exceeds the threshold, the power supply mode is dispatched according to the predicted time point to balance the supply and demand relationship. The residential area adopts a time-of-use electricity price policy during the off-peak period to guide residents to reduce electricity consumption. In this way, the residents' power resources are reasonably allocated and the supply and demand relationship of the power grid is optimized.

[0008] In the existing technology, the electricity price in a specific area is only analyzed by time period, without fully considering the temporal and spatial characteristics of users' electricity consumption behavior for a comprehensive analysis. At the same time, the cyclical changes in electricity consumption behavior are not considered, and only the changes in fixed time periods every day are used as a reference. At the same time, the possible impact of other external factors on users' electricity consumption behavior is not considered, and the results are not accurate enough. The spatial distribution of users' electricity consumption behavior is not combined with geographic information, and the massive electricity consumption data and geographic information data are not fully utilized. Summary of the invention

[0009] In order to solve the above problems existing in the prior art, the present invention proposes an electricity consumption behavior analysis method which takes into account the time and space separation of customers' electricity consumption and payment.

[0010] The technical solution of the present invention is as follows:

[0011] On the one hand, the present invention provides a method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment, and the specific steps include:

[0012] Collect electricity consumption data and payment data of electricity users from each power system according to time period and geographical location, including time series and spatial location;

[0013] Preprocess the collected electricity consumption data and payment data, including data cleaning, data conversion and data reconstruction; then integrate the preprocessed electricity consumption data and payment data with other data that affect electricity consumption behavior to form a comprehensive data set;

[0014] The principal components, i.e., the main electricity consumption patterns, in the comprehensive data set are identified using the empirical orthogonal function. First, the comprehensive data set is decomposed using the spatiotemporal feature matrix, and the eigenvalues ​​of the electricity consumption pattern data in the comprehensive data set are extracted. The eigenvalues ​​of the electricity consumption pattern data are sorted in descending order. At the same time, the electricity consumption patterns ranked by the eigenvalues ​​are selected in turn according to the preset number of principal components to calculate the variance contribution rate. Several electricity consumption pattern data whose variance contribution rates meet the preset accuracy range are selected for reconstruction to obtain the spatiotemporal distribution of each principal component.

[0015] Analyze the reconstructed power consumption mode data, calculate the contribution of each power consumption mode data to the total variance, and determine the impact of each power consumption mode on the user's power consumption behavior;

[0016] According to the impact of each power consumption mode on user behavior, the spatial location information and geographic information of each power consumption mode data are used to draw maps, and the time series information is used to identify the periodic change law of the main components. The allocation of power resources is adjusted according to the map distribution and periodic change law of each main component.

[0017] As a preferred implementation manner, the power systems that collect power consumption data and payment data of power users from each power system according to time periods and geographical locations include smart meters, electric energy systems, marketing management systems, power consumption collection systems, and geographic information systems; the power consumption data and payment data include power consumption, power consumption time, power consumption address, payment time, payment method, and payment location.

[0018] As a preferred implementation, the data cleaning, data conversion and data reconstruction steps are specifically as follows:

[0019] Data cleaning: Identify and handle missing values, outliers, and erroneous data;

[0020] Data conversion: convert non-numeric data into numeric data;

[0021] Data reconstruction includes time series construction and spatial data reconstruction, specifically:

[0022] Time series construction: by presetting the sampling period, the target time series data within each time period is collected;

[0023] Spatial data reconstruction: Reconstruct the data into gridded spatial data according to the geographic location information of each data, divide the spatial units according to the preset conditions, form a power consumption micro-unit grid, and number each micro-unit grid in the order of rows first and columns, and establish a two-way mapping table; after converting the electricity consumption address of each user into longitude and latitude spatial coordinates through the place name address coding service, fill in the corresponding user micro-unit grid, and at the same time fill in the corresponding user's remaining electricity consumption data and payment data.

[0024] As a preferred implementation, the step of first decomposing the comprehensive data set using the spatiotemporal feature matrix is ​​specifically as follows:

[0025] The comprehensive data set includes electricity consumption data at m time points and n spatial locations, and the matrix formed is shown in the following formula:

[0026]

[0027] Preset matrix X with spatiotemporal characteristics ij , spatial function v ik and time function y kj (k=1,2,…,m), the functional relationship between the three is as follows:

[0028]

[0029] The function expression is converted into a matrix expression as follows:

[0030]

[0031] Next, for the matrix x ij Perform SVD decomposition, the specific expression is as follows:

[0032]

[0033] Where E is a diagonal matrix composed of singular values, V L 、V R are the left and right singular vector matrices respectively.

[0034] As a preferred implementation, the power consumption modes ranked by eigenvalues ​​are selected in sequence according to the preset number of principal components to calculate the variance contribution rate in the variance contribution rate step, specifically:

[0035]

[0036] In the formula, R k Indicates the proportion of the variance of the kth principal component to the variance of all components; σ k Represents the variance of the kth principal component, and the variance is represented by the kth element on the diagonal of the diagonal matrix E; σ i represents the variance of the i-th principal component, is the sum of the cumulative variances of each component.

[0037] As a preferred implementation, the step of selecting a plurality of power consumption mode data whose variance contribution rate meets a preset accuracy range for reconstruction is as follows:

[0038] PC i =V iX a×b

[0039] Where X is the original power consumption characteristic collection data; V i Refers to the eigenvector corresponding to the i-th eigenroot; PC i is the i-th spatial eigenvector or spatial mode, which reflects the spatial distribution characteristics of the factor field to a certain extent; a is the number of electricity grids, and b is the number of days.

[0040] As a preferred implementation, the method of sequentially using the spatial location information and geographic information of each power consumption mode data to draw a map is specifically as follows:

[0041] Combine spatial patterns with geographic information, use GIS tools to mark areas with higher electricity consumption or areas with specific electricity consumption patterns on the map; use heat maps to display electricity consumption or electricity intensity in different geographic locations through color depth.

[0042] As a preferred implementation, the method of using time series information to identify the periodic variation law of the principal component is specifically to use Fourier transform or other time series analysis methods based on the time series information to identify the periodic features in the time series information.

[0043] On the other hand, the present invention proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements an electricity usage behavior analysis method that takes into account the time and space separation of customers' electricity usage and payment as described in any embodiment of the present invention.

[0044] On the other hand, the present invention proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an electricity usage behavior analysis method that takes into account the time and space separation of customer electricity usage and payment as described in any embodiment of the present invention.

[0045] The present invention has the following beneficial effects:

[0046] 1. The present invention uses the EOF algorithm to quickly identify electricity consumption patterns, and the visualization tool helps decision makers quickly understand complex electricity consumption data, providing instant decision support for power grid management and power marketing. The spatial distribution of electricity consumption can be used as an indicator of economic development level, helping governments and enterprises understand the intensity and development trends of economic activities in different regions.

[0047] 2. The present invention uses the spatiotemporal characteristics analyzed by the EOF algorithm to understand the electricity consumption characteristics of different regions. Power companies can implement demand-side management strategies, such as peak and valley electricity prices and load control, to balance the grid load and improve grid stability.

[0048] 3. The present invention formulates a more reasonable power grid maintenance and repair plan through the temporal and spatial characteristics of power consumption patterns. By analyzing power consumption patterns, it helps power companies to evaluate and disperse operational risks, such as the impact that natural disasters, equipment failures, etc. may have on the power grid, and formulate risk management strategies.

[0049] 4. The present invention provides personalized power services according to the user's power consumption pattern, guides the user to participate in the demand response plan by analyzing the user's power consumption behavior, and realizes dynamic management of the power grid load. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of the steps of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] It should be understood that the step numbers used in this document are only for convenience of description and are not intended to limit the order in which the steps are executed.

[0053] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0054] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0055] The term "and / or" means and includes any and all possible combinations of one or more of the associated listed items.

[0056] Embodiment 1:

[0057] See also Figure 1 , a method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment, characterized in that the specific steps include:

[0058] Collect electricity consumption data and payment data of electricity users from each power system according to time period and geographical location, including time series and spatial location;

[0059] In this embodiment, electricity consumption data and payment data of electricity users in different time periods (such as day, week, month, season, year) and different geographical locations (such as city, region, street) are collected from business systems such as smart electricity meters, electric energy systems, marketing management systems, electricity consumption collection systems, and geographic information systems (GIS), including electricity consumption, electricity consumption time, electricity consumption address, payment time, payment method, payment location, etc.

[0060] Preprocess the collected electricity consumption data and payment data, including data cleaning, data conversion and data reconstruction; then integrate the preprocessed electricity consumption data and payment data with other data that affect electricity consumption behavior to form a comprehensive data set;

[0061] In this embodiment, data from different sources are integrated, such as combining electricity consumption data with meteorological data, demographic data, etc. It is ensured that the timestamps and spatial locations of different data sources are accurately aligned to facilitate subsequent analysis.

[0062] Verify data consistency and ensure that data between different data sources are logically consistent. Check to ensure that the dataset covers the time and space range required for analysis.

[0063] The principal components, i.e., the main electricity consumption patterns, in the comprehensive data set are identified using the empirical orthogonal function. First, the comprehensive data set is decomposed using the spatiotemporal feature matrix, and the eigenvalues ​​of the electricity consumption pattern data in the comprehensive data set are extracted. The eigenvalues ​​of the electricity consumption pattern data are sorted in descending order. At the same time, the electricity consumption patterns ranked by the eigenvalues ​​are selected in turn according to the preset number of principal components to calculate the variance contribution rate. Several electricity consumption pattern data whose variance contribution rates meet the preset accuracy range are selected for reconstruction to obtain the spatiotemporal distribution of each principal component.

[0064] In this embodiment, the empirical orthogonal function (EOF) algorithm is a statistical method for dimensionality reduction and feature extraction, which is used to identify the main patterns in multivariate data sets. It decomposes the complex spatiotemporal changes in the original data set into several orthogonal principal components (PCs), each of which corresponds to a specific spatiotemporal pattern. In meteorology and oceanography, EOF is used to identify spatial patterns and their changes over time. In the power marketing business, EOF can reveal the spatiotemporal characteristics of electricity consumption data.

[0065] EOF finds a set of orthogonal bases through the eigendecomposition of the covariance matrix, so that the original data set can be represented as a linear combination of this set of bases. Through eigendecomposition, we can find a set of eigenvectors (EOFs) that are not only orthogonal but also sorted by the size of their variance contribution.

[0066] Analyze the reconstructed power consumption mode data, calculate the contribution of each power consumption mode data to the total variance, and determine the impact of each power consumption mode on the user's power consumption behavior;

[0067] In this embodiment, EOF decomposition is used to identify the dominant spatiotemporal patterns in the electricity consumption data, and each extracted principal component is interpreted to determine the physical meaning it represents, such as seasonal changes, differences between weekdays and weekends, regional electricity consumption characteristics, etc. The contribution of each principal component to the total variance is calculated to evaluate its importance in the overall spatiotemporal changes.

[0068] According to the impact of each power consumption mode on user behavior, the spatial location information and geographic information of each power consumption mode data are used to draw maps, and the time series information is used to identify the periodic change law of the main components. The allocation of power resources is adjusted according to the map distribution and periodic change law of each main component.

[0069] In this embodiment, the relationship between the principal components and electricity consumption behavior is studied, such as the change in electricity consumption during peak and valley electricity price periods. The correlation between the principal components and external factors (such as weather conditions and economic activities) is analyzed to understand the impact of external factors on electricity consumption patterns. The specific analysis steps include:

[0070] (1) Result visualization

[0071] It can display complex spatiotemporal data in a graphical way, which is easy to understand. It can provide visual decision support tools for power grid operation and power marketing. It can also serve as an effective means to communicate analysis results with non-professionals.

[0072] (2) Drawing of spatiotemporal feature maps

[0073] Draw the spatiotemporal characteristics of electricity consumption data to intuitively display the distribution and intensity of different modes. Use a line graph to show the trend of electricity consumption patterns over time, a bar graph to compare electricity consumption in different time periods or regions, and a scatter plot to show the relationship between two principal components to analyze the correlation between different electricity consumption patterns.

[0074] (3) Geographic Information Mapping

[0075] Use GIS tools to mark areas with high electricity consumption or specific electricity consumption patterns on the map. Use heat maps and other forms to show the electricity consumption or intensity of electricity consumption in different geographical locations through color depth.

[0076] (4) Time series analysis

[0077] Decompose time series data to reveal trend, seasonality, and random components.

[0078] (5) Interactive tool development

[0079] Develop interactive tools and design simple and intuitive user interfaces, allowing users to adjust analysis parameters and view results according to different needs. Implement dynamic interactive functions such as rolling curtain comparison analysis, map drawing, 3D flight animation, etc.

[0080] As a preferred implementation manner of this embodiment, the power system that collects electricity consumption data and payment data of electricity users from each power system according to time periods and geographical locations includes smart meters, electric energy systems, marketing management systems, electricity consumption collection systems, and geographic information systems; the electricity consumption data and payment data include electricity consumption, electricity consumption time, electricity consumption address, payment time, payment method, and payment location.

[0081] As a preferred implementation of this embodiment, the data cleaning, data conversion and data reconstruction steps are specifically as follows:

[0082] Data cleaning: Identify and handle missing values, outliers, and erroneous data;

[0083] Data conversion: convert non-numeric data into numeric data;

[0084] Data reconstruction includes time series construction and spatial data reconstruction, specifically:

[0085] Time series construction: by presetting the sampling period, the target time series data within each time period is collected;

[0086] Spatial data reconstruction: Reconstruct the data into gridded spatial data according to the geographic location information of each data, divide the spatial units according to the preset conditions, form a power consumption micro-unit grid, and number each micro-unit grid in the order of rows first and columns, and establish a two-way mapping table; after converting the electricity consumption address of each user into longitude and latitude spatial coordinates through the place name address coding service, fill in the corresponding user micro-unit grid, and at the same time fill in the corresponding user's remaining electricity consumption data and payment data.

[0087] In this embodiment, each preprocessing step is specifically as follows:

[0088] Data cleaning: Identify and handle missing values, outliers, or erroneous data to ensure data accuracy and reliability. Use statistical methods (such as mean, median, mode) to fill or delete missing values, depending on the distribution of missing values ​​and the characteristics of the data set. Use standard deviation to identify outliers and handle outliers through strategies such as interpolation, deletion, or replacement. Check the data to ensure that there are no duplicate records in the data set to avoid bias in the analysis results.

[0089] Data conversion: Convert non-numeric data (such as time and geographic location) into numerical data to meet the analysis requirements of the EOF algorithm.

[0090] Convert dates to consecutive days or convert times to hours. Convert addresses and geographic location information to latitude and longitude coordinates, or use regional codes in geographic information systems (GIS). Convert electricity consumption data to a unified unit, such as kilowatt-hours (kWh), and convert payment data to a unified unit, such as yuan, and perform logarithmic transformation to reduce data skewness.

[0091] Data reconstruction:

[0092] Time series construction: This invention focuses on exploring the spatiotemporal characteristics of electricity consumption and payment behavior over multiple days. In order to reduce the impact of peak and valley electricity consumption in the morning and evening and reduce the frequency of electricity calculation, 24 hours is set as the minimum electricity consumption time unit. The daily electricity consumption of each electricity user is formed according to the natural day, and the payment amount and date time series data are issued after the electricity bill is issued.

[0093] Spatial data reconstruction: Reconstruct the data into gridded spatial data based on geographic location information for easy spatial analysis. Divide the spatial units according to the voltage level and the radius of the power consumption area to form a "micro-unit grid" for electricity consumption, and number each "micro-unit grid" in the order of first row and then column to establish a two-way mapping table. Convert the electricity consumption address into longitude and latitude spatial coordinates through the place name address coding service, calculate which "micro-unit grid" the address is located in, and add the electricity consumption, electricity consumption date, payment amount, and payment date of this electricity consumption address to the electricity consumption and payment amount of the "micro-unit grid" by date, and count them.

[0094] As a preferred implementation of this embodiment, the step of first decomposing the comprehensive data set using the spatiotemporal feature matrix is ​​specifically as follows:

[0095] The comprehensive data set includes electricity consumption data at m time points and n spatial locations, and the matrix formed is shown in the following formula:

[0096]

[0097] Preset matrix X with spatiotemporal characteristics ij , spatial function v ik and time function y kj (k=1,2,…,m), the functional relationship between the three is as follows:

[0098]

[0099] The function expression is converted into a matrix expression as follows:

[0100]

[0101] Next, for the matrix x ij Perform SVD decomposition, the specific expression is as follows:

[0102]

[0103] Where E is a diagonal matrix composed of singular values, V L 、V R are the left and right singular vector matrices respectively.

[0104] In this embodiment, E is a diagonal matrix composed of singular values, which can be expressed as follows:

[0105] E=diag(σ1,σ2,…,σ r ),r≤{m,n}

[0106] And the singular values ​​are arranged from large to small:

[0107] σ1≥σ2≥…≥σ r

[0108] V L 、V R They are the left and right singular vector matrices, with dimensions of m×r and n×r respectively.

[0109] As a preferred implementation of this embodiment, the variance contribution rate calculation in the variance contribution rate step is performed by selecting the power consumption modes ranked by eigenvalues ​​in sequence according to the preset number of principal components, specifically:

[0110]

[0111] In the formula, R k Indicates the proportion of the variance of the kth principal component to the variance of all components; σ k Represents the variance of the kth principal component, and the variance is represented by the kth element on the diagonal of the diagonal matrix E; σ i represents the variance of the i-th principal component, is the sum of the cumulative variances of each component.

[0112] As a preferred implementation of this embodiment, the step of selecting a plurality of power consumption mode data whose variance contribution rate meets a preset accuracy range for reconstruction is as follows:

[0113] PC i =V i X a×b

[0114] Where X is the original power consumption characteristic collection data; V i Refers to the eigenvector corresponding to the i-th eigenroot; PC i is the i-th spatial eigenvector or spatial mode, which reflects the spatial distribution characteristics of the factor field to a certain extent; a is the number of electricity grids, and b is the number of days.

[0115] As a preferred implementation of this embodiment, the method of sequentially using the spatial location information and geographic information of each power consumption mode data to draw a map is specifically as follows:

[0116] Combine spatial patterns with geographic information, use GIS tools to mark areas with higher electricity consumption or areas with specific electricity consumption patterns on the map; use heat maps to display electricity consumption or electricity intensity in different geographic locations through color depth.

[0117] As a preferred implementation of this embodiment, the method of using time series information to identify the periodic variation law of the main component is specifically to use Fourier transform or other time series analysis methods based on the time series information to identify the periodic characteristics in the time series information.

[0118] Embodiment 2:

[0119] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for analyzing electricity usage behavior that takes into account the separation of customer electricity usage and payment time and space as described in any embodiment of the present invention.

[0120] Embodiment three:

[0121] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements a method for analyzing electricity usage behavior that takes into account the time and space separation of customer electricity usage and payment as described in any embodiment of the present invention.

[0122] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment, characterized in that: The specific steps include: Collect electricity consumption data and payment data of electricity users from each power system according to time period and geographical location, including time series and spatial location; Preprocess the collected electricity consumption data and payment data, including data cleaning, data conversion and data reconstruction; then integrate the preprocessed electricity consumption data and payment data with other data that affect electricity consumption behavior to form a comprehensive data set; The principal components, i.e., the main electricity consumption patterns, in the comprehensive data set are identified using the empirical orthogonal function. First, the comprehensive data set is decomposed using the spatiotemporal feature matrix, and the eigenvalues ​​of the electricity consumption pattern data in the comprehensive data set are extracted. The eigenvalues ​​of the electricity consumption pattern data are sorted in descending order. At the same time, the electricity consumption patterns ranked by the eigenvalues ​​are selected in turn according to the preset number of principal components to calculate the variance contribution rate. Several electricity consumption pattern data whose variance contribution rates meet the preset accuracy range are selected for reconstruction to obtain the spatiotemporal distribution of each principal component. Analyze the reconstructed power consumption mode data, calculate the contribution of each power consumption mode data to the total variance, and determine the impact of each power consumption mode on the user's power consumption behavior; According to the impact of each power consumption mode on user behavior, the spatial location information and geographic information of each power consumption mode data are used to draw maps, and the time series information is used to identify the periodic change law of the main components. The allocation of power resources is adjusted according to the map distribution and periodic change law of each main component.

2. According to claim 1, a method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment, characterized in that: The power systems that collect electricity consumption data and payment data of electricity users from each power system according to time periods and geographical locations include smart meters, electric energy systems, marketing management systems, electricity consumption collection systems, and geographic information systems; the electricity consumption data and payment data include electricity consumption, electricity consumption time, electricity consumption address, payment time, payment method, and payment location.

3. The method for analyzing electricity consumption behavior according to claim 1, which takes into account the time and space separation of customers' electricity consumption and payment, is characterized in that: The data cleaning, data conversion and data reconstruction steps are specifically as follows: Data cleaning: Identify and handle missing values, outliers, and erroneous data; Data conversion: convert non-numeric data into numeric data; Data reconstruction includes time series construction and spatial data reconstruction, specifically: Time series construction: by presetting the sampling period, the target time series data within each time period is collected; Spatial data reconstruction: Reconstruct the data into gridded spatial data according to the geographic location information of each data, divide the spatial units according to the preset conditions, form a power consumption micro-unit grid, and number each micro-unit grid in the order of rows first and columns, and establish a two-way mapping table; after converting the electricity consumption address of each user into longitude and latitude spatial coordinates through the place name address coding service, fill in the corresponding user micro-unit grid, and at the same time fill in the corresponding user's remaining electricity consumption data and payment data.

4. The method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment according to claim 1, characterized in that: The step of first decomposing the comprehensive data set using the spatiotemporal feature matrix is ​​specifically as follows: The comprehensive data set includes electricity consumption data at m time points and n spatial locations, and the matrix X formed is shown as follows: Preset matrix X with spatiotemporal characteristics ij , spatial function v ik and time function y kj (k=1,2,…,m), the functional relationship between the three is as follows: The function expression is converted into a matrix expression as follows: Next, for the matrix x ij Perform SVD decomposition, the specific expression is as follows: Where E is a diagonal matrix composed of singular values, V L 、V R They are the left and right singular vector matrices respectively. E is usually arranged from large to small.

5. The method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment according to claim 1, characterized in that: The calculation of the variance contribution rate in the variance contribution rate step is performed by selecting the power consumption modes ranked by the eigenvalues ​​in sequence according to the preset number of principal components, specifically: In the formula, R k Indicates the proportion of the variance of the kth principal component to the variance of all components; σ k Represents the variance of the kth principal component, and the variance is represented by the kth element on the diagonal of the diagonal matrix E; σ i represents the variance of the i-th principal component, is the sum of the cumulative variances of each component.

6. The method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment according to claim 1, characterized in that: The step of selecting a plurality of power consumption mode data whose variance contribution rate meets the preset accuracy range for reconstruction is as follows: PC i =V i X a×b In the formula, X is the original electricity consumption characteristic collection data; V i Refers to the eigenvector corresponding to the i-th eigenroot; PC i is the i-th spatial eigenvector or spatial mode, which reflects the spatial distribution characteristics of the factor field to a certain extent; a is the number of electricity grids, and b is the number of days.

7. The method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment according to claim 1, characterized in that: The method for sequentially using the spatial location information and geographic information of each power consumption mode data to draw a map is specifically as follows: Combine spatial patterns with geographic information, use GIS tools to mark areas with higher electricity consumption or areas with specific electricity consumption patterns on the map; use heat maps to display electricity consumption or electricity intensity in different geographic locations through color depth.

8. The method for analyzing electricity consumption behavior taking into account the time and space separation of customer electricity consumption and payment according to claim 1, characterized in that: The method of using time series information to identify the periodic variation law of the principal component is specifically to use Fourier transform or other time series analysis methods based on the time series information to identify the periodic features in the time series information.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the electricity usage behavior analysis method according to any one of claims 1 to 8 that takes into account the separation of customer electricity usage and payment time and space.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements a method for analyzing electricity usage behavior that takes into account the separation of customer electricity usage and payment time and space as described in any one of claims 1 to 8.

Images