A User Profile-Based Smart Electricity Carbon Delivery Method

Abstract

This invention discloses a user-profile-based intelligent push method for electricity carbon emissions, belonging to the field of intelligent push for electricity carbon emissions. The method includes the following steps: S1, collecting and classifying users' annual historical load data and time-of-use electricity price datasets; S2, extracting typical daily load curves and typical time-of-use electricity price curves for each season; S3, calculating the daily electricity price demand elasticity for weekdays and weekends; S4, constructing user profiles; S5, classifying users' electricity price demand elasticity; S6, pushing differentiated electricity purchase strategies and carbon emission reduction regulatory strategies to price-sensitive and non-price-sensitive users respectively. This invention employs the above-mentioned user-profile-based intelligent push method for electricity carbon emissions, accurately characterizing user profiles based on the user's electricity price demand elasticity coefficient. This allows the platform to customize differentiated strategies for different types of users, improving the operational capabilities of the regulatory platform and demonstrating significant application value.

Description

Technical Field

[0001] This invention relates to the field of electric carbon collaborative intelligent push, and in particular to an electric carbon collaborative intelligent push method based on user profiles. Background Technology

[0002] In the field of carbon emissions reduction, research on intelligent push technology targeting user electricity consumption behavior often focuses on minimizing user costs, neglecting the user's load characteristics. Considering user characteristics, if a user is price-sensitive, time-of-use pricing can effectively reduce customer electricity costs and alleviate grid load dispatching pressure by guiding users to change their electricity purchasing behavior at different times, thus contributing to the achievement of carbon emission reduction targets. However, if a user is price-insensitive, the effect of time-of-use pricing is limited.

[0003] Furthermore, current research on user profiling only focuses on seasonal analysis of user electricity consumption behavior, without delving into the distinction between weekdays and weekends. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a user-profile-based intelligent push method for electricity carbon pricing. This method accurately profiles users based on their electricity price demand elasticity coefficient, enabling the platform to customize differentiated strategies for different user types, thereby enhancing the operational capabilities of the regulatory platform and demonstrating significant application value.

[0005] To achieve the above objectives, this invention provides a user profile-based intelligent push method for electricity carbon-coordinated delivery, comprising the following steps:

[0006] S1. Collect and categorize users' annual historical load data and time-of-use electricity price datasets according to the season, and perform data preprocessing;

[0007] S2. Classify the daily load curves and time-of-use electricity price curves of the same season, and classify the typical load curves and time-of-use electricity price curves of weekdays and weekends. Then, use the particle swarm optimization algorithm to extract the typical daily load curves and typical time-of-use electricity price curves of the season.

[0008] S3. Under the same season, based on the typical daily load curve and time-of-use electricity price curve, calculate the daily electricity price demand elasticity on weekdays and weekends.

[0009] S4. Use the GMM algorithm to perform cluster analysis on the elasticity of electricity price demand, construct user profiles, and form a user tag library;

[0010] S5. Classification of users’ electricity price demand elasticity: Based on the constructed user profile, users are divided into price-sensitive users and price-insensitive users;

[0011] S6. Push differentiated electricity purchase strategies and carbon emission reduction regulatory strategies to price-sensitive users and insensitive users respectively.

[0012] Preferably, the data preprocessing in step S1 includes missing value processing and malformed value processing;

[0013] Missing value handling: Missing values ​​are supplemented using the average daily load or electricity price data for the corresponding season, i.e., mean interpolation, with the following formula:

[0014]

[0015] In the formula, This represents the average daily load or electricity price data for the corresponding season; n represents the number of days in the corresponding season; x i This refers to the load or electricity price data for day i.

[0016] The steps for handling abnormal values ​​are as follows: Using the Laida criterion, first calculate the average value of the load data. X i Let V be the load data for day i, and the standard deviation V of the load data. i =X i -X(i = 1, 2, 3, ... n);

[0017] Then, the standard deviation δ is calculated using Bessel's formula, and the measured value x is set. b Standard deviation V b (1≤b≤n);

[0018] When: |V b |=|x b When -X|>3δ, determine x b If the value is an outlier, filter it out and replace it with X;

[0019] After outliers, X and V are recalculated. i Then, a new round of screening is carried out to check for other abnormal data in the remaining data until all data meets the requirements.

[0020] Preferably, the extraction steps for the typical weekday load curve and the typical weekend load curve in step S2 are as follows:

[0021] Step 1: Assume that a user has m+1 workdays or weekends in the j-th season. Then, find the set of workday or weekend loads Y for that user in this season. j ∈(m+1)×24:

[0022]

[0023] In the formula, Let t be the load value of the user on the nth day and tth hour during the jth season's weekday or weekend, where n = 0, 1, 2, ..., m, and t = 0, 1, 2, ..., 23;

[0024] Step 2: Define the typical load X of the user on the j-th seasonal workday or weekend. j for:

[0025] X j = [x0, x1, x2, ... x t , ...x 23 In equation (3), x t This represents the typical load value for users in the t-th hour during a weekday or weekend in the j-th season.

[0026] Step 3: Use the shortest distance between all coordinates of the curves as the objective function.

[0027]

[0028] In the formula, S j Y is the objective function value; j [n] represents the user's load value on day n of the j-th season; X j [t] represents the load value for hour t in the j-th season;

[0029] Among them, X j Note that the load constraints must be met to ensure that the load remains within the specified range, i.e.:

[0030] min{(Y j [n][t]}≤x j [t]≤max{(Y j [n][t]} (5)

[0031] In the formula, min{(Y j [n][t]} represents the minimum value of the load interval; max{(Y j [n][t]} represents the maximum value of the load interval; and t = 0, 1, 2, ..., 23; n = 0, 1, 2, ..., m;

[0032] Fourth step: Solve the objective function described in formula (4) using the particle swarm optimization algorithm, and obtain the optimal X through iterative solution. j As a typical load curve for the corresponding season;

[0033] The steps for extracting the seasonal typical weekday time-of-use electricity price curve and the seasonal typical weekend time-of-use electricity price curve are as follows:

[0034] Step 1: Assume that a user has m+1 working days or weekends in the j-th season. Then, find the set Z of the time-of-use electricity prices for the user during the working days or weekends in this season.j ∈(m+1)×24:

[0035]

[0036] In the formula, The electricity price for a user on the nth day of the tth hour during a weekday or weekend in the jth season;

[0037] Step 2: Let the typical time-of-use electricity price curve for the user on the j-th seasonal workday or weekend be...

[0038] A j = [a0, a1, a2, ... a t , ...a 23 (7)

[0039] In the formula: a t The electricity price for a user during the t-th hour on a weekday or weekend in the j-th season;

[0040] Step 3: Use the shortest distance between all coordinates of the curves as the objective function.

[0041]

[0042] In the formula: P j Z represents the objective function value; m represents the number of weekdays or weekend days in the j-th season; Z represents the objective function value. j [n] represents the electricity price for the user on day n in the j-th season; a j [t] represents the typical time-of-use electricity price for the user at time t in the j-th season;

[0043] Among them, a j The value at any given time cannot exceed the upper or lower limit of that time.

[0044] min{Z j [n][t]}≤a j [t]≤max{(Z j [n][t]} (9)

[0045] In the formula, min{Z j [n][t]} represents the lower bound at time t; and max{(Z j [n][t]} represents the upper limit value at time t;

[0046] Fourth step: Solve the objective function described in formula (8) using the particle swarm optimization algorithm, and obtain the optimal a through iterative solution. j This serves as a typical time-of-use electricity price curve for the corresponding season.

[0047] Preferably, in step S3, the typical load curve and time-of-use electricity price curve for the same season and corresponding date are combined to form a typical load-electricity price curve. Then, the daily electricity price demand elasticity of users is calculated using the electricity price elasticity formula, and an electricity price demand elasticity curve is formed.

[0048] Furthermore, the elasticity coefficient E of electricity load in the electricity price elasticity formula represents the percentage change in user electricity demand ΔL caused by a change in electricity price ΔP within a certain period, and its expression is as follows:

[0049]

[0050] In the formula, P0 and L0 represent typical electricity price and electricity demand, respectively.

[0051] Preferably, step S4 specifically includes the following steps:

[0052] S41. Determine the number of clusters k for the elasticity of electricity prices on weekdays and weekends under the same season, and use the sum of squared errors (SSE) as the evaluation index for determining the number of clusters:

[0053]

[0054] In the formula, C i For a specific cluster in the clustering; u i C i The mean of all data points in the cluster; x is a point in the cluster;

[0055] S42. Assuming the true number of clusters is k*, compare the true number of clusters k* and k, and determine the optimal number of clusters based on the comparison results and the changes in SSE.

[0056] S43. Use GMM to fit the dataset to a mixture of multiple Gaussian distributions to discover the underlying structure of the data:

[0057] S431. Assume that each class of data follows a normal distribution, and the probability model of the normal distribution is:

[0058]

[0059] In the formula, k is the number of clusters; α k As weight; The probability density function is a normal distribution; θ k y represents the probability density parameter; y represents the sample.

[0060] The probability density expression is:

[0061]

[0062] In the formula: σ k The standard deviation of sample y; μk Let y be the mean of the sample y;

[0063] Using the training data of the normal distribution probability model, estimate the probability density function of k classes of normal distributions and the weight α of each class. k ;

[0064] S432. Calculate the probability of each data point appearing in each of the k normal distributions, that is, substitute the data point into each of the k normal distributions to find the probability P(y) of belonging to each class. i ) k :

[0065]

[0066] In the formula: y i Let be a data point in the dataset; k represents the k-th normal distribution.

[0067] S433, By comparing probabilities P(y) i ) k The value of is used to assign the sample to the cluster with the highest probability value.

[0068] Preferably, in step S5, a user electricity consumption behavior profile is generated based on the clustering results of the user's price demand elasticity, and the user is divided into price-sensitive users and price-insensitive users.

[0069] Preferably, in step S5, the user's price demand elasticity curve is updated periodically based on the user's real-time feedback on time-of-use electricity pricing, and the user type is iterated to improve the accuracy of push notifications.

[0070] The present invention has the following beneficial effects:

[0071] 1. By using the GMM algorithm to perform user clustering analysis and build user profiles, the demand characteristics of electricity users can be discovered, thereby customizing differentiated strategies, improving the operational capabilities of the regulatory platform, and having good application value.

[0072] 2. A cost-effective and sophisticated real-time monitoring solution helps users clearly identify their current emissions status, assess the economics of carbon reduction measures, and promote the research and application of low-carbon technologies.

[0073] 3. Help users accurately grasp their own electricity purchasing behavior characteristics, reduce electricity costs, and assist users in formulating more reasonable electricity purchasing and emission reduction measures.

[0074] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0075] Figure 1This is a flowchart of an intelligent push method for electricity carbon based on user profiles according to the present invention. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages disclosed in the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments in this application without creative effort are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout.

[0077] It should be noted that the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0078] Similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0079] like Figure 1 As shown, a user profile-based intelligent push method for electricity carbon-coordinated delivery includes the following steps:

[0080] S1. Collect and categorize users' annual historical load data and time-of-use electricity price datasets according to the season, and perform data preprocessing;

[0081] The data preprocessing described in step S1 includes handling missing values ​​and handling malformed values;

[0082] Missing value handling: Since the data has been categorized by season and fluctuations caused by seasonal factors have been eliminated, the electricity consumption data distribution for each season is similar. Therefore, missing values ​​are supplemented using the average daily load or electricity price data for the corresponding season, i.e., mean interpolation. The formula is as follows:

[0083]

[0084] In the formula, This represents the average daily load or electricity price data for the corresponding season; n represents the number of days in the corresponding season; x i This refers to the load or electricity price data for day i.

[0085] The steps for handling outlier values ​​are as follows: The dataset contains a large amount of user load data that conforms to a normal distribution. Therefore, the Raida criterion is adopted to first calculate the average value of the load data. X i Let V be the load data for day i, and the standard deviation V of the load data. i =X i -X(i = 1, 2, 3, ..., n);

[0086] Then, the standard deviation δ is calculated using Bessel's formula, and the measured value x is set. b Standard deviation V b (1≤b≤n);

[0087] When: |V b |=|x b When -X|>3δ, determine x b If the value is an outlier, filter it out and replace it with X;

[0088] After outliers, X and V are recalculated. i Then, a new round of screening is carried out to check for other abnormal data in the remaining data until all data meets the requirements.

[0089] S2. Classify the daily load curves and time-of-use electricity price curves of the same season, and classify the typical load curves and time-of-use electricity price curves of weekdays and weekends. Use the particle swarm optimization algorithm to extract the typical daily load curves and typical time-of-use electricity price curves of the season, ensuring the objectivity of the basic load extraction.

[0090] The steps for extracting the typical weekday load curve and the typical weekend load curve in step S2 are as follows:

[0091] Step 1: Assume that a user has m+1 workdays or weekends in the j-th season. Then, find the set of workday or weekend loads Y for that user in this season. j ∈(m+1)×24:

[0092]

[0093] In the formula, Let t be the load value of a user on the nth day and tth hour during a weekday or weekend in the jth season, where n = 0, 1, 2, ..., m, and t = 0, 1, 2, ..., 23;

[0094] Step 2: Define the typical load X of the user on the j-th seasonal workday or weekend. j for:

[0095] X j = [x0, x1, x2, ... x t , ...x 23In equation (3), x t This represents the typical load value for users in the t-th hour during a weekday or weekend in the j-th season.

[0096] Step 3: Use the shortest distance between all coordinates of the curves as the objective function.

[0097]

[0098] In the formula, S j Y is the objective function value; j [n] represents the user's load value on day n of the j-th season; X j [t] represents the load value for hour t in the j-th season;

[0099] Among them, X j Note that the load constraints must be met to ensure that the load remains within the specified range, i.e.:

[0100] min{(Y j [n][t]}≤x j [t]≤max{(Y j [n][t]} (5)

[0101] In the formula, min{(Y j [n][t]} represents the minimum value of the load interval; max{(Y j [n][t]} represents the maximum value of the load interval; and t = 0, 1, 2, ..., 23; n = 0, 1, 2, ..., m;

[0102] Fourth step: Solve the objective function described in formula (4) using the particle swarm optimization algorithm, and obtain the optimal X through iterative solution. j As a typical load curve for the corresponding season;

[0103] The steps for extracting the seasonal typical weekday time-of-use electricity price curve and the seasonal typical weekend time-of-use electricity price curve are as follows:

[0104] Step 1: Assume that a user has m+1 working days or weekends in the j-th season. Then, find the set Z of the time-of-use electricity prices for the user during the working days or weekends in this season. j ∈(m+1)×24:

[0105]

[0106] In the formula, The electricity price for a user on the nth day of the tth hour during a weekday or weekend in the jth season;

[0107] Step 2: Let the typical time-of-use electricity price curve for the user on the j-th seasonal workday or weekend be...

[0108] Aj = [a0, a1, a2, ... a t , ...a 23 (7)

[0109] In the formula: a t The electricity price for a user during the t-th hour on a weekday or weekend in the j-th season;

[0110] Step 3: Use the shortest distance between all coordinates of the curves as the objective function.

[0111]

[0112] In the formula: P j Z represents the objective function value; m represents the number of weekdays or weekend days in the j-th season; Z represents the objective function value. j [n] represents the electricity price for the user on day n in the j-th season; a j [t] represents the typical time-of-use electricity price for the user at time t in the j-th season;

[0113] Among them, a j The value at any given time cannot exceed the upper or lower limit of that time.

[0114] min{Z j [n][t]}≤a j [t]≤max{(Z j [n][t]} (9)

[0115] In the formula, min{Z j [n][t]} represents the lower bound at time t; and max{(Z j [n][t]} represents the upper limit value at time t;

[0116] Fourth step: Solve the objective function described in formula (8) using the particle swarm optimization algorithm, and obtain the optimal a through iterative solution. j This serves as a typical time-of-use electricity price curve for the corresponding season.

[0117] S3. Under the same season, based on the typical daily load curve and time-of-use electricity price curve, calculate the daily electricity price demand elasticity on weekdays and weekends.

[0118] In step S3, the typical load curve and time-of-use electricity price curve for the same season and corresponding date are combined to form a typical load-electricity price curve. Then, the daily electricity price demand elasticity of users is calculated using the electricity price elasticity formula, and an electricity price demand elasticity curve is formed.

[0119] Furthermore, the elasticity coefficient E of electricity load in the electricity price elasticity formula represents the percentage change in user electricity demand ΔL caused by the change in electricity price AP within a certain period, and its expression is as follows:

[0120]

[0121] In the formula, P0 and L0 represent typical electricity price and electricity demand, respectively.

[0122] S4. Use the GMM algorithm to perform cluster analysis on the elasticity of electricity price demand, construct user profiles, and form a user tag library;

[0123] Step S4 specifically includes the following steps:

[0124] S41. Determine the number of clusters k for the elasticity of electricity prices on weekdays and weekends under the same season, and use the sum of squared errors (SSE) as the evaluation index for determining the number of clusters:

[0125]

[0126] In the formula, C i For a specific cluster in the clustering; u i C i The mean of all data points in the cluster; x is a point in the cluster;

[0127] S42. Assume the actual number of clusters is k. * Compared to the actual number of clusters k * And k, based on the comparison results, determine the optimal number of clusters according to the changes in SSE;

[0128] S43. Use GMM to fit the dataset to a mixture of multiple Gaussian distributions to discover the underlying structure of the data:

[0129] S431. Assume that each class of data follows a normal distribution, and the probability model of the normal distribution is:

[0130]

[0131] In the formula, k is the number of clusters; α k As weight; The probability density function is a normal distribution; θ k y represents the probability density parameter; y represents the sample.

[0132] The probability density expression is:

[0133]

[0134] In the formula: σ k The standard deviation of sample y; μ k Let y be the mean of the sample y;

[0135] Using the training data of the normal distribution probability model, estimate the probability density function of k classes of normal distributions and the weight α of each class. k ;

[0136] S432. Calculate the probability of each data point appearing in each of the k normal distributions, that is, substitute the data point into each of the k normal distributions to find the probability P(y) of belonging to each class. i ) k :

[0137]

[0138] In the formula: y i Let be a data point in the dataset; k represents the k-th normal distribution.

[0139] S433, By comparing probabilities P(y) i ) k The value of is used to assign the sample to the cluster with the highest probability value.

[0140] S5. Classification of users’ electricity price demand elasticity: Based on the constructed user profile, users are divided into price-sensitive users and price-insensitive users;

[0141] In step S5, a user electricity consumption behavior profile is generated based on the clustering results of the user's price demand elasticity, and users are divided into price-sensitive users and price-insensitive users, which facilitates subsequent analysis of the user's inherent electricity consumption behavior attributes in this season.

[0142] In step S5, the user price demand elasticity curve is updated periodically based on the user's real-time feedback on time-of-use electricity pricing, and the user type is iterated to improve the accuracy of push notifications.

[0143] S6. Push differentiated electricity purchase strategies and carbon emission reduction regulatory strategies to price-sensitive users and insensitive users respectively.

[0144] For example, preferential electricity pricing policies can be offered to price-sensitive users, guiding them to reduce electricity consumption during peak hours and increase electricity purchases during off-peak hours to lower their electricity costs and simultaneously reduce carbon emissions during those periods. For price-insensitive users, time-of-use pricing has a relatively weaker regulatory effect. Instead, models that minimize the user's electricity purchase costs and carbon emission costs can be generated to guide them in adjusting their energy consumption structure, reducing the proportion of thermal power purchases and increasing the proportion of green electricity purchases, thus helping them reduce carbon emissions.

[0145] Therefore, the present invention adopts the above-mentioned user profile-based collaborative intelligent push method for electricity carbon, which accurately portrays user profiles based on the user's electricity price demand elasticity coefficient, so that the platform can customize differentiated strategies for different types of users, improve the operational capabilities of the regulatory platform, and has good application value.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A user profile-based intelligent push method for electricity-carbon collaborative delivery, characterized in that: Includes the following steps: S1. Collect and categorize users' annual historical load data and time-of-use electricity price datasets according to the season, and perform data preprocessing; S2. Classify the daily load curves and time-of-use electricity price curves of the same season, and classify the typical load curves and time-of-use electricity price curves of weekdays and weekends. Then, use the particle swarm optimization algorithm to extract the typical daily load curves and typical time-of-use electricity price curves of the season. The steps for extracting the typical weekday load curve and the typical weekend load curve in step S2 are as follows: Step 1: Assume that the user is known at the [number]th [time]. There are a total of 10 seasons If a user has a set of weekday or weekend loads during this season, then the user's weekday or weekend loads during this season are calculated based on the number of weekdays or weekends. ∈( : = (2) In the formula, For the first Users on weekdays or weekends during the season Heavenly Hourly load value, ; Step 2: Set user number Typical load on a seasonal workday or weekend for: (3) In the formula, For the first Users on weekdays or weekends during the season Typical load values ​​for a given hour; Step 3: Use the shortest distance between all coordinates of the curves as the objective function. (4) In the formula, The objective function value; in, Note that the load constraints must be met to ensure that the load remains within the specified range, i.e.: (5) In the formula, This is the minimum value within the load range; This is the maximum value within the load range; and ； Fourth step: Solve the objective function described in formula (4) using the particle swarm optimization algorithm, and obtain the optimal solution through iterative solution. As a typical load curve for the corresponding season; The steps for extracting the seasonal typical weekday time-of-use electricity price curve and the seasonal typical weekend time-of-use electricity price curve are as follows: Step 1: Assume that the user is known at the [number]th [time]. Of the seasons, there are a total of For weekdays or weekends, the user's time-of-use electricity rates for weekdays or weekends during this season are calculated as follows: ∈( )×24 : = (6) In the formula, For the first During the season, on weekdays or weekends, users... Heavenly Hourly electricity price; Step 2: Set user number The typical time-of-use electricity price curve for a seasonal weekday or weekend is as follows: (7) In the formula: For the first During the season, on weekdays or weekends, users... Hourly electricity price; Step 3: Use the shortest distance between all coordinates of the curves as the objective function. (8) In the formula: The objective function value; For the first The number of weekdays or weekend days in a season; in, The value at any given time cannot exceed the upper or lower limit of that time. (9) In the formula, for The lower limit of the time; for The upper limit of the time frame; Fourth step: Solve the objective function described in formula (8) using the particle swarm optimization algorithm, and obtain the optimal solution through iterative solution. As a typical time-of-use electricity price curve for the corresponding season; S3. Under the same season, based on the typical daily load curve and time-of-use electricity price curve, calculate the daily electricity price demand elasticity on weekdays and weekends. S4. Use the GMM algorithm to perform cluster analysis on the elasticity of electricity price demand, construct user profiles, and form a user tag library; Step S4 specifically includes the following steps: S41. Determine the number of clusters for weekday and weekend electricity price elasticities under the same season. and the sum of squared errors As an evaluation metric for determining the number of clusters: (11) In the formula, It is a cluster within a cluster; for The mean of each data point in the dataset; For a point in the cluster; S42. Assume the actual number of clusters is... The actual number of clusters is compared to and Based on the comparison results The optimal number of clusters is determined by the changes in the data. S43. Use GMM to fit the dataset to a mixture of multiple Gaussian distributions to discover the underlying structure of the data: S431. Assume that each class of data follows a normal distribution, and the probability model of the normal distribution is: (12) In the formula, The number of clusters; As weight; The probability density is a normal distribution; is a parameter of the probability density; For the sample; The probability density expression is: (13) In the formula: For the sample Standard deviation; For the sample The mean; Estimate the probability using the training data of the normal distribution probability model. The probability density function of a normal distribution and the weights of each class. ; S432, Calculate the values ​​for each data point. The probability of occurrence in a normal-like distribution is obtained by substituting the data into... Find the probability of belonging to each class from a normal distribution. : (14) In the formula: For a specific piece of data in the dataset; For the first There are several normal distributions; S433, By comparing probabilities The value of the probability value determines which cluster the sample belongs to; S5. Classification of users’ electricity price demand elasticity: Based on the constructed user profile, users are divided into price-sensitive users and price-insensitive users; S6. Push differentiated electricity purchase strategies and carbon emission reduction regulatory strategies to price-sensitive users and insensitive users respectively.

2. The method for intelligent push notification based on user profiles using electricity and carbon dioxide as described in claim 1, characterized in that: The data preprocessing described in step S1 includes handling missing values ​​and handling malformed values; Missing value handling: Missing values ​​are supplemented using the average daily load or electricity price data for the corresponding season, i.e., mean interpolation, with the following formula: (1) In the formula, This refers to the average daily load or electricity price data for the corresponding season. This refers to the number of days in the corresponding season. For the first Daily load or electricity price data; The steps for handling abnormal values ​​are as follows: Using the Laida criterion, first calculate the average value of the load data. , For the first Daily load data, and the standard deviation of the load data. ; Then, the standard deviation is calculated using Bessel's formula. and set the measurement value standard deviation ( ; When the following conditions are met: >3 At that time, the judgment For outliers, filter them and use... Substitute; Recalculate after outlier. and Then, a new round of screening is carried out to check for other abnormal data in the remaining data until all data meets the requirements.

3. The method for intelligent push notification based on user profiles for carbon dioxide recycling according to claim 2, characterized in that: In step S3, the typical load curve and time-of-use electricity price curve for the same season and corresponding date are combined to form a typical load-electricity price curve. Then, the daily electricity price demand elasticity of users is calculated using the electricity price elasticity formula, and an electricity price demand elasticity curve is formed. Furthermore, the elasticity coefficient of electricity load is set in the electricity price elasticity formula. This indicates changes in electricity prices over a certain period of time. Changes in user electricity demand caused by The percentage is expressed as follows: (10) In the formula, , These represent typical electricity prices and electricity demand, respectively.

4. The method for intelligent push notification of electricity and carbon based on user profiles according to claim 1, characterized in that: In step S5, a user electricity consumption behavior profile is generated based on the clustering results of the user's price demand elasticity, and the user is divided into price-sensitive users and price-insensitive users.

5. The method for intelligent push notification based on user profiles for carbon dioxide recycling according to claim 4, characterized in that: In step S5, the user price demand elasticity curve is updated periodically based on the user's real-time feedback on time-of-use electricity pricing, and the user type is iterated to improve the accuracy of push notifications.

Images