Electric power data analysis method and device

Abstract

The invention provides a power data analysis method and device. The method comprises the following steps: acquiring historical power consumption data and related external variable data of a target area; establishing a multiple regression relation model between power consumption and external variables according to the historical power consumption data and related external variable data; defining a probability distribution function for each external variable; performing Monte Carlo simulation based on the multiple regression relation model, randomly extracting a value from probability distribution of each external variable for each iteration, substituting the extracted external variable value into the multiple regression relation model, and calculating a corresponding power consumption prediction value; accumulating and storing the power consumption predicted value obtained by each iteration to form a simulation result database; from the simulation result database, key external variable combinations and contexts that result in significant changes in power consumption of the target region are identified. According to the invention, the accuracy of power consumption prediction can be improved, and key external variable combinations and situations can be identified.

Description

Technical Field

[0001] The present invention relates to the technical field of power data analysis, and in particular to a power data analysis method and a device thereof. Background Art

[0002] With the development of social economy and population growth, the demand for electricity continues to rise, and the stable operation of the power system and the rational allocation of power resources are becoming more and more important. In order to improve the operating efficiency of the power system and the management level of power resources, power data analysis technology has emerged. Traditional power data analysis methods mainly rely on historical data for simple trend prediction. This method often ignores the impact of external environmental factors on power consumption, resulting in inaccurate prediction results. In addition, existing methods rarely consider the interaction between multiple external variables, making it difficult to identify the key factors affecting power consumption. Summary of the invention

[0003] The purpose of the present invention is to propose a power data analysis method and device thereof to improve the accuracy of power consumption prediction and identify key external variable combinations and scenarios that affect power consumption.

[0004] To achieve the above object, the present invention provides a power data analysis method, comprising:

[0005] Obtain historical electricity consumption data and relevant external variable data in the target area;

[0006] Establishing a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data;

[0007] Define probability distribution functions for each external variable;

[0008] Performing Monte Carlo simulation based on the multivariate regression relationship model, for each iteration, randomly extracting a value from the probability distribution of each external variable, substituting the extracted external variable value into the multivariate regression relationship model, and calculating the corresponding power consumption prediction value;

[0009] The power consumption prediction values ​​obtained in each iteration are accumulated and stored to form a simulation result database;

[0010] Based on the simulation results database, key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area are identified.

[0011] Preferably, the external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

[0012] Preferably, the establishing of a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data includes:

[0013] Normalizing the historical power consumption data and related external variable data;

[0014] Using multiple linear regression analysis, the normalized electricity consumption was used as the dependent variable and the external variables as the independent variables to establish a multiple regression relationship model.

[0015] Preferably, defining a probability distribution function for each external variable comprises:

[0016] The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

[0017] Preferably, based on the simulation result database, identifying key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area includes:

[0018] Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

[0019] As the same inventive concept, the present invention also provides a power data analysis device, comprising:

[0020] A data collection module is used to obtain historical electricity consumption data and related external variable data of the target area;

[0021] A model building module, used to build a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data;

[0022] Distribution definition module, used to define the probability distribution function for each external variable;

[0023] A prediction module, configured to perform a Monte Carlo simulation based on the multivariate regression relationship model, randomly extract a value from the probability distribution of each external variable for each iteration, substitute the extracted external variable value into the multivariate regression relationship model, and calculate a corresponding power consumption prediction value;

[0024] A database module is used to accumulate and store the power consumption prediction values ​​obtained in each iteration to form a simulation result database;

[0025] The result output module is used to identify key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the simulation result database.

[0026] Preferably, the external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

[0027] Preferably, the model building module is used to:

[0028] Normalizing the historical power consumption data and related external variable data;

[0029] Using multiple linear regression analysis, the normalized electricity consumption was used as the dependent variable and the external variables as the independent variables to establish a multiple regression relationship model.

[0030] Preferably, the distribution definition module is used to:

[0031] The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

[0032] Preferably, the result output module is used to:

[0033] Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

[0034] The present invention has the following beneficial effects:

[0035] Traditional power consumption prediction methods often fail to fully consider the impact of external environmental factors on power consumption, resulting in large errors in the prediction results. The present invention can more accurately predict power consumption by establishing a multivariate regression relationship model between power consumption and external variables and introducing Monte Carlo simulation.

[0036] In a complex power system, multiple external variables may affect power consumption at the same time, and the interaction between different variables may lead to significant changes in power consumption. By simulating scenarios with different external variable combinations, the present invention can effectively identify key external variable combinations that lead to significant changes in power consumption, and provide decision support for the optimal dispatch of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0038] Figure 1 The figure is a flow chart of a power data analysis method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The detailed description of the accompanying drawings is intended as an illustration of the current embodiments of the present invention and is not intended to represent the only form in which the present invention can be implemented. It should be understood that the same or equivalent functions can be accomplished by different embodiments that are intended to be included in the spirit and scope of the present invention.

[0040] See also Figure 1 The present invention provides a method for analyzing power data, comprising the following steps:

[0041] Step S10, obtaining historical power consumption data and related external variable data of the target area.

[0042] Specifically, the target area may be a city, and the relevant external variable data may be data that may be related to the power consumption of the target area, such as weather conditions, population mobility, etc.

[0043] Step S20: establishing a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data.

[0044] Specifically, the multivariate regression relationship model can be expressed as:

[0045] text{power consumption}=beta_0+beta_1*text{variable 1}+beta_2*text{variable 2}+...+beta_n*text{variable n}+ε;

[0046] Among them, beta_0 is the intercept term, beta_1, beta_2, ..., beta_n are the regression coefficients of each external variable, and ε is the error term.

[0047] Step S30, defining a probability distribution function for each external variable.

[0048] Step S40, performing Monte Carlo simulation based on the multiple regression relationship model, for each iteration, randomly extracting a value from the probability distribution of each external variable, substituting the extracted external variable value into the multiple regression relationship model, and calculating the corresponding power consumption prediction value.

[0049] Specifically, Monte Carlo simulation is a numerical calculation method based on probability and statistical theory, which simulates the probability distribution of one or more random variables through random sampling to estimate possible results. In power data analysis, Monte Carlo simulation can be used to predict power consumption under the influence of different external variables.

[0050] In each iteration of the Monte Carlo simulation, we need to generate a random value for each external variable, and these random values ​​are drawn according to the previously defined probability distribution function (step S30). For example, if an external variable follows a normal distribution, we will randomly draw a value from the normal distribution.

[0051] The extracted external variable values ​​are substituted into the multivariate regression relationship model. This means that we use these randomly generated values ​​as inputs to the model to calculate the corresponding power consumption prediction values.

[0052] Using the regression model, we can get a predicted value for electricity consumption. This process is repeated in each iteration, each time using different randomly drawn values ​​of the external variables.

[0053] The specific steps are as follows:

[0054] Start iteration and set the number of iterations (e.g., 1000 or more, depending on the required accuracy and computing resources);

[0055] In each iteration, for each external variable, the following operations are performed:

[0056] According to the probability distribution function of the external variable, a value is randomly selected;

[0057] Substitute all extracted external variable values ​​into the multiple regression model together;

[0058] Calculate the model output, i.e., the predicted value of electricity consumption;

[0059] Record the predicted value of this iteration;

[0060] Repeat the above process until all iterations are completed;

[0061] Eventually, through a large number of iterations, we can get a distribution of predicted values, which helps us understand the possible range and probability distribution of power consumption under different combinations of external variables.

[0062] Step S50, accumulating and storing the power consumption prediction values ​​obtained in each iteration to form a simulation result database.

[0063] Specifically, in each iteration of the Monte Carlo simulation, a predicted value of power consumption is generated; these predicted values ​​are the possible results of random variables under different scenarios; in order to be able to analyze these results, they need to be stored; the simulation result database should have a clear structure to facilitate the storage and retrieval of data; the database contains the following fields: number of iterations, external variable values ​​for each iteration, corresponding power consumption prediction values, etc.; after each iteration is completed, the number of iterations, the extracted external variable values, and the calculated power consumption prediction values ​​are stored as a record in the database; this process is repeated until all iterations are completed; the database can be a relational database (such as MySQL, PostgreSQL) or a non-relational database (such as MongoDB, Cassandra), and the specific choice depends on the amount of data, query requirements, and system environment.

[0064] The specific steps are as follows:

[0065] Create an empty database or data table to store simulation results.

[0066] After each Monte Carlo simulation iteration, the following information is inserted into the database as a record: an iteration identifier (e.g., iteration number or unique identifier), the extracted value of each external variable in this iteration, and the power consumption prediction value calculated in this iteration;

[0067] Make sure the database can handle the large number of inserts and perform performance optimizations where necessary to avoid bottlenecks in data storage.

[0068] After all iterations are completed, the database will contain a complete set of simulation results, which can be used for further analysis.

[0069] In this way, the simulation results database becomes a valuable data resource that not only contains a large amount of forecast data, but also retains the contextual information of each forecast (i.e. the specific values ​​of external variables). This enables analysts to conduct in-depth data mining, such as identifying potential patterns in electricity consumption, assessing risks, and predicting extreme situations.

[0070] Step S60, identifying key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the simulation result database.

[0071] Specifically, step S60 is the analysis phase in the power data analysis method, which aims to extract valuable information from the simulation result database to identify which external variable combinations and scenarios will lead to significant changes in power consumption in the target area. The identified key external variable combinations and scenarios are explained to understand how they affect power consumption. These findings can be combined with actual conditions, such as seasonal changes, market trends, policy impacts, etc., to provide deeper insights. The analysis results are organized into reports, and key findings are presented through visualization tools such as charts, heat maps, and scatter plots to facilitate understanding by decision makers and stakeholders.

[0072] Through step S60, power companies and policymakers can better understand the impact of external factors on power consumption, and thus take corresponding measures to optimize the allocation of power resources, improve the efficiency and stability of the power grid, and cope with possible peaks or valleys in power demand.

[0073] Furthermore, the external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

[0074] Specifically, in the power data analysis method, considering external variables is crucial to improving prediction accuracy. The following is an explanation of how the mentioned external variables are further refined and applied in step S60.

[0075] Temperature: Analyze the impact of temperature changes on electricity consumption, especially under extreme temperature conditions (such as high or low temperatures), and identify the nonlinear relationship between temperature and electricity consumption, such as the use of air conditioning and heating equipment.

[0076] Humidity: Investigate the potential impact of humidity on electricity consumption, especially the impact on air conditioning energy consumption in high humidity environments, and analyze the interaction between humidity and other variables, such as the increase in electricity consumption under high temperature and high humidity conditions.

[0077] Precipitation: Study the short-term and long-term effects of precipitation on electricity consumption, such as the cooling effect after rainfall that may reduce air conditioning use, and analyze the relationship between precipitation and hydropower generation.

[0078] Population density: Explore the direct impact of population density on electricity consumption. Generally, densely populated areas have higher electricity demand. Analyze the long-term impact of population density changes on electricity consumption.

[0079] Urbanization rate: examines the impact of urbanization on electricity consumption patterns. Increased urbanization rates are usually accompanied by an increase in electricity demand. Analyzes the relationship between urbanization rates and power infrastructure investment and energy efficiency improvements.

[0080] GDP: Study the correlation between GDP growth and electricity consumption. Usually, GDP growth will lead to an increase in electricity demand. Analyze the relationship between the contribution of different industries to GDP and electricity consumption.

[0081] Industrial output: Assess the impact of industrial output on electricity consumption, especially the electricity demand of heavy industry and high energy consumption industries, and analyze the short-term and long-term impact of changes in industrial output on electricity consumption.

[0082] In step S60, the analysis of these external variables may include the following aspects:

[0083] Variable Importance Ranking: Determine which external variables have the most significant impact on electricity consumption.

[0084] Scenario modeling: Construct different scenarios of external variable combinations and simulate the power consumption under these scenarios.

[0085] Critical scenario identification: By comparing the power consumption forecasts under different scenarios, the critical scenarios that cause significant changes are identified.

[0086] Risk analysis: Assessing the risk of power supply under a specific combination of external variables, such as the impact of extreme weather events or economic fluctuations on the power system.

[0087] Through these analyses, we can better understand how external variables affect electricity consumption individually or collectively, and provide a scientific basis for the planning, operation and risk management of power systems.

[0088] Furthermore, the step S20 includes:

[0089] Step S201, normalizing the historical power consumption data and related external variable data;

[0090] Specifically, normalization is an important step in data processing, which aims to scale data to a small specific range, usually [0, 1] or [-1, 1], for easy comparison and calculation.

[0091] The following are the specific steps of normalization:

[0092] Calculate maximum and minimum values: For each external variable and power consumption data, calculate the maximum and minimum values ​​in its historical data.

[0093] Apply normalization formula: Use the following formula to normalize the data:

[0094] For 0-1 normalization: text{normalized value}={text{original value}-text{minimum value}} / {text{maximum value}-text{minimum value}};

[0095] For -1 to 1 normalization: text{normalized value}=2*{text{original value}-text{minimum value}} / {text{maximum value}-text{minimum value}}-1;

[0096] Ensure that all data are normalized identically to maintain consistency in multiple regression analyses.

[0097] Step S202, using multiple linear regression analysis, taking the normalized power consumption as the dependent variable and the external variables as the independent variables, to establish a multiple regression relationship model.

[0098] Specifically, multiple linear regression is a statistical technique used to model the linear relationship between a dependent variable (in this case, electricity consumption) and multiple independent variables (external variables).

[0099] Here are the steps to build the model:

[0100] Select Model Type: Decide to use a multiple linear regression model to describe the relationship between electricity consumption and external variables.

[0101] Define the dependent and independent variables: take the normalized electricity consumption data as the dependent variable (Y), and the normalized external variable data (temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value) as the independent variables (X1, X2,..., Xn).

[0102] Fit the model: Use the method of least squares or other optimization algorithms to estimate regression coefficients that represent the relative impact of each external variable on electricity consumption.

[0103] Verify the model: Use cross-validation, residual analysis and other methods to test the accuracy, stability and reliability of the model.

[0104] Model Output: The resulting multiple regression model usually has the following form:

[0105] Y=beta_0+beta_1X_1+beta_2X_2+...+beta_nX_n+ε;

[0106] Where Y is the predicted electricity consumption, X1, X2, ..., Xn are external variables, β0 is the intercept, β1, β2, ..., βn are regression coefficients, and ε is the error term.

[0107] Through these steps, a multivariate regression model for predicting electricity consumption can be established, which considers the combined effects of multiple external variables.

[0108] Furthermore, the step S30 includes:

[0109] The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

[0110] Specifically, descriptive statistical analysis is performed on each external variable, including calculation of statistical quantities such as mean, median, standard deviation, minimum, maximum, skewness and kurtosis, and distribution characteristics of the data are observed by drawing histograms, density curves, box plots, etc., such as whether it is symmetrical, whether there are outliers, and the shape of the distribution. Based on descriptive statistics and distribution characteristics, the most suitable distribution type for each external variable is determined. Common distribution types include normal distribution, lognormal distribution, uniform distribution, exponential distribution, Weibull distribution, etc. If the data conforms to a specific distribution, the parameters of the distribution need to be estimated, such as the mean and standard deviation of the normal distribution.

[0111] According to the determined distribution type and parameters, a probability distribution function of each external variable is defined, and the probability distribution function can be a probability density function (PDF) or a cumulative distribution function (CDF).

[0112] Probability density function (PDF): describes the probability density of a random variable around a specific value.

[0113] Cumulative Distribution Function (CDF): describes the probability that a random variable is less than or equal to a certain value.

[0114] The following is an example of defining a probability distribution function:

[0115] For normal distribution: If the data is normally distributed, its probability density function is:

[0116]

[0117] Here, μ is the mean and σ is the standard deviation.

[0118] For other distributions: Depending on the type of distribution identified, use the corresponding formula to define the PDF or CDF.

[0119] Following these steps, a probability distribution function can be defined for each external variable, which will be used in the Monte Carlo simulation to randomly generate values ​​for the external variables, thereby simulating different scenarios. This helps to more accurately predict how electricity consumption will change under different external conditions.

[0120] Furthermore, based on the simulation result database, key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area are identified, including:

[0121] Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

[0122] Specifically, in step S60, identifying key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area is an analytical process that involves in-depth mining and comparison of the simulation result database. The following is a detailed explanation of this step:

[0123] Comparing power consumption forecasts corresponding to different external variable combinations: Extracting power consumption forecasts under different external variable combinations from the simulation result database, which are obtained through Monte Carlo simulation in step S40. Comparing the forecast results of different external variable combinations, this can be achieved by calculating various statistical indicators (such as mean, median, standard deviation, etc.), or by visualizing them through visualization tools (such as scatter plots, heat maps, etc.).

[0124] Identify key external variable combinations and scenarios: Perform significance tests, such as t-test or ANOVA, on the power consumption forecast values ​​for each external variable combination to determine which combinations have a significant impact on power consumption. Analyze the differences in power consumption forecast values ​​between different combinations to identify those combinations that lead to significant increases or decreases in power consumption.

[0125] Scenario identification: Combinations of variables with significant impacts are defined as “key scenarios”, which may include specific weather conditions (such as high temperature and humidity), economic activities (such as a significant increase in industrial output), or other external events (such as a temporary increase in population density caused by a large-scale event).

[0126] Here is an example:

[0127] Step 1: Filter out the high and low value intervals of the power consumption forecast value from the database;

[0128] Step 2: For each high and low value interval, check the corresponding external variable combination;

[0129] Step 3: Analyze the common characteristics of these combinations and identify which changes in external variables are associated with significant changes in electricity consumption;

[0130] Step 4: Determine the scenarios corresponding to the combinations of these key external variables, for example, “During the summer heat wave, when the temperature exceeds 30°C and the humidity exceeds 70%, electricity consumption increases significantly”;

[0131] Step 5: Verify the universality and reproducibility of these key scenarios to ensure that they are not just isolated anomalies.

[0132] Through this process, combinations and scenarios of external variables that have a significant impact on electricity consumption can be identified, thus providing a scientific basis for the planning, dispatching and risk management of the power system.

[0133] Another embodiment of the present invention further provides a power data analysis device, comprising:

[0134] A data collection module is used to obtain historical electricity consumption data and related external variable data of the target area;

[0135] A model building module, used to build a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data;

[0136] Distribution definition module, used to define the probability distribution function for each external variable;

[0137] A prediction module, configured to perform a Monte Carlo simulation based on the multivariate regression relationship model, randomly extract a value from the probability distribution of each external variable for each iteration, substitute the extracted external variable value into the multivariate regression relationship model, and calculate a corresponding power consumption prediction value;

[0138] A database module is used to accumulate and store the power consumption prediction values ​​obtained in each iteration to form a simulation result database;

[0139] The result output module is used to identify key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the simulation result database.

[0140] Furthermore, the external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

[0141] Furthermore, the model building module is used to:

[0142] Normalizing the historical power consumption data and related external variable data;

[0143] Using multiple linear regression analysis, the normalized electricity consumption was used as the dependent variable and the external variables as the independent variables to establish a multiple regression relationship model.

[0144] Furthermore, the distribution definition module is used to:

[0145] The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

[0146] Furthermore, the result output module is used to:

[0147] Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

[0148] The device of this embodiment corresponds to the method of the above embodiment. Therefore, the contents not described in detail in the device of this embodiment can be obtained by referring to the contents of the method of the above embodiment, and will not be repeated here.

[0149] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for analyzing power data, characterized in that: include: Obtain historical electricity consumption data and relevant external variable data in the target area; Establishing a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data; Define probability distribution functions for each external variable; Performing Monte Carlo simulation based on the multivariate regression relationship model, for each iteration, randomly extracting a value from the probability distribution of each external variable, substituting the extracted external variable value into the multivariate regression relationship model, and calculating the corresponding power consumption prediction value; The power consumption prediction values ​​obtained in each iteration are accumulated and stored to form a simulation result database; Based on the simulation results database, key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area are identified.

2. The method according to claim 1, characterized in that The external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

3. The method according to claim 2, characterized in that The establishing of a multivariate regression relationship model between power consumption and external variables according to the historical power consumption data and relevant external variable data includes: Normalizing the historical power consumption data and related external variable data; Using multiple linear regression analysis, the normalized electricity consumption was used as the dependent variable and the external variables as the independent variables to establish a multiple regression relationship model.

4. The method according to claim 3, characterized in that The method of defining a probability distribution function for each external variable includes: The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

5. The method according to claim 4, characterized in that The method of identifying key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the simulation result database includes: Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

6. A power data analysis device, characterized in that: include: A data collection module is used to obtain historical electricity consumption data and related external variable data of the target area; A model building module, used to build a multivariate regression relationship model between power consumption and external variables based on the historical power consumption data and relevant external variable data; Distribution definition module, used to define the probability distribution function for each external variable; A prediction module, configured to perform a Monte Carlo simulation based on the multivariate regression relationship model, randomly extract a value from the probability distribution of each external variable for each iteration, substitute the extracted external variable value into the multivariate regression relationship model, and calculate a corresponding power consumption prediction value; A database module is used to accumulate and store the power consumption prediction values ​​obtained in each iteration to form a simulation result database; The result output module is used to identify key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the simulation result database.

7. The device according to claim 6, characterized in that The external variables include temperature, humidity, precipitation, population density, urbanization rate, GDP and industrial output value of the target area.

8. The device according to claim 7, characterized in that The model building module is used to: Normalizing the historical power consumption data and related external variable data; Using multiple linear regression analysis, the normalized electricity consumption was used as the dependent variable and the external variables as the independent variables to establish a multiple regression relationship model.

9. The device according to claim 8, characterized in that The distribution definition module is used to: The distribution law of the external variable is obtained by statistically analyzing the historical data of the external variable, and a probability distribution function suitable for describing the distribution of the external variable data is defined according to the distribution law.

10. The device according to claim 9, characterized in that The result output module is used to: Compare the electricity consumption forecasts corresponding to different combinations of external variables, and determine the key external variable combinations and scenarios that lead to significant changes in electricity consumption in the target area based on the comparison results.

Images