An ai intelligent power consumption prediction method based on power consumption behavior portrait

Abstract

The application relates to an AI intelligent power consumption prediction method based on power consumption behavior portrait, relates to the technical field of temperature control, and comprises the following steps: acquiring power consumption data of a plurality of basic users in a historical transaction year; determining the types of the basic users according to the power consumption data of the plurality of basic users in the historical transaction year, determining the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of daily total quantity corresponding to the historical transaction year of each type of user according to the power consumption data of the plurality of basic users in the historical transaction year, and determining the once-calibrated reference time-of-use power consumption of each basic user in the historical transaction year according to the power consumption data of the plurality of basic users in the historical transaction year and the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of daily total quantity corresponding to the historical transaction year of each type of user; and performing power consumption prediction in a future transaction year according to the once-calibrated reference time-of-use power consumption of each basic user in the historical transaction year, so that the accuracy of power consumption prediction is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electricity consumption prediction, and particularly relates to an AI intelligent electricity consumption prediction method based on electricity consumption behavior portrait. BACKGROUND

[0002] Stable supply of electric energy is necessary for the development of today's social economy and the normal life of the people, and accurate electricity consumption prediction can provide reliable guidance for electricity production and power supply scheduling, thereby improving the power supply quality of the power system. The time series distribution of electricity consumption data has certain statistical characteristics. First, the electricity consumption data itself is the superposition of electricity consumption of various industries. For industrial electricity consumption, although different regions have different electricity consumption, the industrial electricity consumption in a certain region usually has certain periodicity, that is, the working day electricity consumption is relatively stable, and the weekend electricity consumption is significantly reduced. For residential electricity consumption, its periodicity is relatively not obvious, and the electricity consumption is related to holidays, weather conditions, etc., and generally shows great volatility. Therefore, after superimposing the electricity consumption of industry, residents, business, government agencies and institutions, etc., the total electricity consumption in a region shows certain periodicity and certain volatility, which brings challenges to electricity consumption prediction tasks.

[0003] The existing industrial electricity decomposition algorithm generally adopts the benchmark electricity method. The benchmark electricity method usually sets the benchmark electricity based on historical data or fixed standards, and then compares the actual electricity for decomposition. This method is simple and intuitive, but lacks flexibility and is difficult to adapt to the complex changes of industrial user production electricity. The production electricity of industrial users is often affected by many factors, such as production plan, equipment state, market demand, etc. Changes in these factors may cause significant changes in electricity consumption patterns. The existing user electricity decomposition algorithm often does not fully consider the particularity of industrial user production electricity, resulting in deviation between the decomposition result and the actual electricity consumption. The production electricity of industrial users usually has periodicity, seasonality, volatility and other characteristics, which are not fully reflected in the existing decomposition algorithm.

[0004] Therefore, it is necessary to provide an AI intelligent electricity consumption prediction method based on electricity consumption behavior portrait to improve the accuracy of electricity consumption prediction. SUMMARY

[0005] The application provides an AI intelligent power consumption prediction method based on power consumption behavior portrait, comprising: acquiring power consumption data of a plurality of basic users in a historical transaction year; determining the type of the basic users according to the power consumption data of the plurality of basic users in the historical transaction year, determining the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of the total daily load of each type of user corresponding to the historical transaction year according to the power consumption data of the plurality of basic users in the historical transaction year, and determining the once-calibrated reference time-of-use power consumption of each basic user in the historical transaction year according to the power consumption data of the plurality of basic users in the historical transaction year and the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of the total daily load of each type of user corresponding to the historical transaction year; and performing power consumption prediction in a future transaction year according to the once-calibrated reference time-of-use power consumption of each basic user in the historical transaction year.

[0006] Further, the type of the basic users is determined according to the power consumption data of the plurality of basic users in the historical transaction year, comprising: determining a first power consumption type of each basic user, wherein the first power consumption type is industry, business, office or charging station; for each first power consumption type, extracting historical power consumption features of each basic user included in the first power consumption type from the power consumption data of each basic user included in the first power consumption type in the historical transaction year, and determining a second power consumption type of each basic user included in the first power consumption type according to the historical power consumption features of each basic user included in the first power consumption type.

[0007] Further, the historical power consumption features of each basic user included in the first power consumption type are extracted from the power consumption data of each basic user included in the first power consumption type in the historical transaction year, and the second power consumption type of each basic user included in the first power consumption type is determined according to the historical power consumption features of each basic user included in the first power consumption type, comprising: classifying dates to determine a plurality of date types; for each basic user, determining power consumption features of the basic user corresponding to each date type according to the power consumption data of the basic user in the historical transaction year; for any two basic users included in the first power consumption type, calculating the similarity of the two basic users according to the power consumption features of the two basic users corresponding to each date type; and dividing the plurality of basic users included in the first power consumption type into a plurality of basic user groups according to the similarity of any two basic users included in the first power consumption type by a clustering algorithm, wherein one basic user group corresponds to one second power consumption type.

[0008] Further, the air conditioner time-of-use electricity unit curve and the air conditioner proportion of daily total electricity of each type of user corresponding to the historical transaction year are determined according to the electricity data of the historical transaction year of the plurality of basic users, including: for each two-type electricity type, the basic user time-of-use electricity of the two-type electricity type is determined according to the electricity data of the historical transaction year of each basic user included in the two-type electricity type, the reference time-of-use electricity of each date type corresponding to the two-type electricity type is determined according to the basic user time-of-use electricity of the two-type electricity type, the daily air conditioner electricity of each date type corresponding to the two-type electricity type is determined based on the reference time-of-use electricity of each date type corresponding to the two-type electricity type, the air conditioner time-of-use electricity unit curve and the air conditioner proportion of daily total electricity of the two-type electricity type corresponding to the historical transaction year are determined according to the daily air conditioner electricity of each date type corresponding to the two-type electricity type.

[0009] Further, the daily air conditioner electricity of each date type corresponding to the two-type electricity type is determined based on the reference time-of-use electricity of each date type corresponding to the two-type electricity type, including: for each date type, the initial air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type is calculated according to the difference between the basic user time-of-use electricity of the two-type electricity type and the reference time-of-use electricity of the corresponding date type, the calibrated air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type is determined by calibrating the initial air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type, and the daily air conditioner electricity of the date type corresponding to the two-type electricity type is determined according to the calibrated air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type.

[0010] Further, the daily air conditioner electricity of each date type corresponding to the two-type electricity type is determined based on the reference time-of-use electricity of each date type corresponding to the two-type electricity type, including: for each date type, the initial air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type is calculated according to the difference between the basic user time-of-use electricity of the two-type electricity type and the reference time-of-use electricity of the corresponding date type, the calibrated air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type is determined by calibrating the initial air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type, and the daily air conditioner electricity of the date type corresponding to the two-type electricity type is determined according to the calibrated air conditioner time-of-use electricity of the date type corresponding to the two-type electricity type.

[0011] Further, the method further comprises: determining the air conditioner time-sharing electricity standard curve and the air conditioner proportion of the total daily electricity of the second electricity type corresponding to the historical transaction year according to the daily air conditioner electricity of each date type corresponding to the second electricity type; and determining the air conditioner time-sharing electricity standard curve according to the ratio of the calibrated air conditioner time-sharing electricity of the second electricity type corresponding to the date type to the daily air conditioner electricity of each date type corresponding to the second electricity type; and determining the air conditioner proportion of the total daily electricity of the second electricity type corresponding to the historical transaction year according to the ratio of the daily air conditioner electricity of each date type corresponding to the second electricity type to the basic electricity of each date type corresponding to the second electricity type.

[0012] Further, the method further comprises: determining the one-time calibration reference time-sharing electricity of each basic user in the historical transaction year according to the one-time calibration air conditioner time-sharing electricity of the basic user; and determining the one-time calibration reference time-sharing electricity of the basic user in the historical transaction year according to the one-time calibration air conditioner time-sharing electricity of the basic user.

[0013] Further, the method further comprises: determining the electricity data of the basic user in the future transaction year according to the one-time calibration reference time-sharing electricity of each basic user in the historical transaction year, which comprises: obtaining the daily weather information of the future transaction year; determining the air conditioner reference day from the historical transaction year according to the daily weather information of the future transaction year; determining the daily air conditioner electricity of the basic user according to the air conditioner reference day; and determining the electricity data of the basic user in the future transaction year according to the one-time calibration reference time-sharing electricity of the basic user in the historical transaction year and the daily air conditioner electricity.

[0014] Further, the method further comprises: determining the electricity data of the basic user in the future transaction year according to the one-time calibration reference time-sharing electricity of each basic user in the historical transaction year, which comprises: determining the daily electricity of the basic user according to the one-time calibration reference time-sharing electricity of the basic user in the historical transaction year and the daily air conditioner electricity; determining the monthly electricity of the basic user according to the daily electricity of the basic user; and determining the annual electricity of the basic user in the future transaction year according to the monthly electricity of the basic user.

[0015] Compared with the prior art, the AI intelligent electricity prediction method based on electricity behavior portrait provided by the present specification has at least the following beneficial effects:

[0016] 1. By obtaining and analyzing historical transaction year electricity data of multiple base users, the electricity consumption behavior and pattern of users can be more accurately understood. By determining key parameters such as user type, air conditioner time-of-use electricity unit curve and air conditioner proportion of daily total, the electricity consumption prediction model can be further refined, thereby improving the accuracy and reliability of the prediction. Accurate electricity consumption prediction helps the power company to more reasonably arrange power production and supply, avoiding the situation of power surplus or shortage;

[0017] 2. By first determining the electricity consumption type of the base user (such as industrial, commercial, office or charging station), and then further determining the second type of electricity consumption according to the historical electricity consumption characteristics, more refined classification of users can be achieved. This classification helps to more accurately understand the electricity consumption behavior and demand of different users, thereby improving the accuracy of electricity consumption prediction. By extracting and analyzing historical electricity consumption characteristics, and clustering users according to these characteristics, a prediction model more suitable for different user groups can be constructed. This model can better capture and reflect the electricity consumption rules and trends of various users, thereby improving the accuracy and reliability of the prediction. BRIEF DESCRIPTION OF DRAWINGS

[0018] The present specification will be further illustrated in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, the same numbers represent the same structures, wherein:

[0019] Figure 1 is a flowchart of an AI intelligent electricity consumption prediction method based on electricity consumption behavior portrait in an embodiment of the present application;

[0020] Figure 2 is a schematic diagram of the air conditioner electricity consumption before and after correction shown in an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings needed in the embodiment description will be briefly introduced as follows.

[0022] Figure 1 is a flowchart of an AI intelligent electricity consumption prediction method based on electricity consumption behavior portrait in an embodiment of the present application, as shown in Figure 1 , an AI intelligent electricity consumption prediction method based on electricity consumption behavior portrait can include the following steps.

[0023] Step 110, obtaining the historical electricity consumption data of multiple base users in a transaction year.

[0024] The basic user can be a user whose number of historical transaction years corresponding to the power consumption data is greater than a number threshold, for example, a user whose number of historical transaction years corresponding to the power consumption data is greater than 1 year is taken as the basic user.

[0025] In step 120, the type of the basic user is determined according to the power consumption data of the historical transaction years of the plurality of basic users, the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of the total daily amount corresponding to the historical transaction years of each type of user are determined according to the power consumption data of the historical transaction years of the plurality of basic users, and the once-calibration reference time-of-use power consumption of each basic user in the historical transaction year is determined according to the power consumption data of the historical transaction years of the plurality of basic users and the air conditioner time-of-use power consumption unit curve and the air conditioner proportion of the total daily amount corresponding to the historical transaction years of each type of user.

[0026] In some embodiments, the type of the basic user is determined according to the power consumption data of the historical transaction years of the plurality of basic users, including:

[0027] The type of the basic user is determined according to the power consumption data of the historical transaction years of the plurality of basic users, including:

[0028] For each type of power consumption type, the historical power consumption features of each basic user included in the type of power consumption type are extracted from the power consumption data of the historical transaction years of each basic user included in the type of power consumption type, and the type of power consumption type of each basic user included in the type of power consumption type is determined according to the historical power consumption features of each basic user included in the type of power consumption type.

[0029] In some embodiments, the historical power consumption features of each basic user included in the type of power consumption type are extracted from the power consumption data of the historical transaction years of each basic user included in the type of power consumption type, and the type of power consumption type of each basic user included in the type of power consumption type is determined according to the historical power consumption features of each basic user included in the type of power consumption type, including:

[0030] The dates are classified to determine a plurality of date types;

[0031] For each basic user, the power consumption features of the basic user corresponding to each date type are determined according to the power consumption data of the historical transaction years of the basic user, wherein the power consumption features can at least include the maximum daily power consumption, the average daily power consumption, the variance of the daily power consumption, etc.

[0032] For any two basic users included in the type of power consumption type, the similarity of the two basic users is calculated according to the power consumption features of the two basic users corresponding to each date type.

[0033] The plurality of basic users included in a type of electricity consumption is divided into a plurality of basic user groups according to the similarity of any two basic users included in the type of electricity consumption by a clustering algorithm (for example, a K-means algorithm, etc.), wherein one basic user group corresponds to one type of secondary electricity consumption.

[0034] Specifically, the plurality of date types can be as shown in Table 1.

[0035] Table 1

[0036]

[0037] The similarity of the two basic users can be calculated according to the following formula:

[0038]

[0039] wherein S (i,j) is the similarity of the i-th basic user and the j-th basic user, f (i,m,n) is the M-th electricity consumption feature of the i-th basic user corresponding to the n-th date type, F (j,m,n) is the M-th electricity consumption feature of the j-th basic user corresponding to the n-th date type, N is the total number of date types, M is the total number of electricity consumption features, P1 is a preset parameter, and P1 is greater than 0.

[0040] In some embodiments, according to the electricity consumption data of the historical transaction years of the plurality of basic users, the air conditioner time-of-use electricity unit curve and the air conditioner proportion of daily total amount corresponding to the historical transaction years of each type of user are determined, comprising:

[0041] For each type of secondary electricity consumption, according to the electricity consumption data of the historical transaction years of each basic user included in the type of secondary electricity consumption, the basic user time-of-use electricity of the type of secondary electricity consumption is determined, according to the basic user time-of-use electricity of the type of secondary electricity consumption, the reference time-of-use electricity of each date type corresponding to the type of secondary electricity consumption is determined, based on the reference time-of-use electricity of each date type corresponding to the type of secondary electricity consumption, the daily air conditioner electricity consumption of each date type corresponding to the type of secondary electricity consumption is determined, and according to the daily air conditioner electricity consumption of each date type corresponding to the type of secondary electricity consumption, the air conditioner time-of-use electricity unit curve and the air conditioner proportion of daily total amount corresponding to the historical transaction years of the type of secondary electricity consumption are determined.

[0042] Specifically, for the type of secondary electricity consumption of the industrial type of primary electricity consumption, the basic user time-of-use electricity of the type of secondary electricity consumption can be calculated according to the following formula:

[0043]

[0044] For the type of secondary electricity consumption of the non-industrial type of primary electricity consumption, the basic user time-of-use electricity of the type of secondary electricity consumption can be calculated according to the following formula:

[0045]

[0046] wherein Q d,t,i,基础分时用电 is the basic time-of-use electricity of the i-th second-class electricity type at t hours of d day, q d,t,u,用户分时用电 is the user time-of-use electricity of the u-th basic user included in the i-th second-class electricity type at t hours of d day, GF d,t,光伏分时发电 is the photovoltaic time-of-use electricity of the u-th basic user at t hours of d day, and U is the total number of basic users included in the i-th second-class electricity type.

[0047] For example, the historical grid time-of-use electricity of industrial basic users is summed up to obtain the daily grid time-of-use electricity data of industrial basic users from January 1, 2023 to December 31, 2023, and then the photovoltaic electricity data of the basic users is obtained according to the proportion of the annual total electricity of the basic users to the annual total electricity of the basic users, and the sum of the two parts is obtained. The daily time-of-use electricity data of industrial basic users from January 1, 2023 to December 31, 2023. Commercial and other classifications do not need to add photovoltaic electricity data, only need to perform the previous calculation process, that is, obtain the daily grid time-of-use electricity data from January 1, 2023 to December 31, 2023 as the daily time-of-use electricity data of the basic users of this classification.

[0048] In some embodiments, the reference time-of-use electricity of each date type corresponding to the second-class electricity type can be determined according to the basic user time-of-use electricity of the second-class electricity type according to the rules shown in Table 2.

[0049] Table 2

[0050]

[0051]

[0052] In some embodiments, based on the reference time-of-use electricity of each date type corresponding to the second-class electricity type, the daily air conditioning electricity of each date type corresponding to the second-class electricity type is determined, including:

[0053] For each date type, the initial air conditioning time-of-use electricity of the date type corresponding to the second-class electricity type is calculated according to the difference between the basic user time-of-use electricity of the second-class electricity type and the reference time-of-use electricity of the corresponding date type, the initial air conditioning time-of-use electricity of the date type corresponding to the second-class electricity type is calibrated to determine the calibrated air conditioning time-of-use electricity of the date type corresponding to the second-class electricity type, and the daily air conditioning electricity of the date type corresponding to the second-class electricity type is determined according to the calibrated air conditioning time-of-use electricity of the date type corresponding to the second-class electricity type.

[0054] Specifically, the initial air conditioning time-of-use electricity of the date type corresponding to the second-class electricity type can be calculated according to the following formula:

[0055] Q d,t,i,最初空调分时用电 =Q d,t,i,基础分时用电 -Q L,t,i,基准分时用电

[0056] wherein, Q d,t,i,最初空调分时用电 is the initial air conditioning time-of-use electricity of the i-th second type of electricity of d day at t hour, Q L,t,i,基准分时用电 is the reference time-of-use electricity of the i-th second type of electricity of d day at t hour.

[0057] In some embodiments, the initial air conditioning time-of-use electricity of the second type of electricity corresponding to the date type is calibrated, and the daily air conditioning electricity of the second type of electricity corresponding to the date type is determined, comprising:

[0058] According to the initial air conditioning time-of-use electricity of the second type of electricity corresponding to the date type, the daily air conditioning electricity of the second type of electricity corresponding to the date type is determined;

[0059] According to the basic user time-of-use electricity of the second type of electricity, the daily electricity of the second type of electricity corresponding to the date type is determined;

[0060] According to the basic user time-of-use electricity of the second type of electricity and the daily electricity of the second type of electricity corresponding to the date type, the basic user time-of-use electricity standard curve of the second type of electricity corresponding to the date type is calculated;

[0061] According to the initial air conditioning time-of-use electricity of the second type of electricity corresponding to the date type and the daily air conditioning electricity of the second type of electricity corresponding to the date type, the initial air conditioning electricity standard curve of the second type of electricity corresponding to the date type is determined;

[0062] According to the basic user time-of-use electricity standard curve of the second type of electricity corresponding to the date type and the initial air conditioning electricity standard curve and the initial air conditioning time-of-use electricity, the daily air conditioning electricity of the second type of electricity corresponding to the date type is determined.

[0063] Specifically, the daily electricity of the second type of electricity corresponding to the date type can be calculated according to the following formula:

[0064]

[0065] wherein, Q d,i,日用电量 is the daily electricity of the i-th second type of electricity of d day, Q d,t,i,修正空调分时用电 is the corrected electricity of the i-th second type of electricity of d day at t hour. Wherein, if Q d,t,i,最初空调分时用电 < 0, then Q d,t,i,修正空调分时用电 = 0, if Q d,t,i,最初空调分时用电 ≥ 0, then Q d,t,i,修正空调分时用电 = Q d,t,i,最初空调分时用电 .

[0066] The daily electricity consumption of the second type of electricity consumption corresponding to the date type can be calculated based on the following formula:

[0067]

[0068] Q d,i,基础用电 is the daily electricity consumption of the i-th second type of electricity consumption on day d.

[0069] The basic user time-of-use electricity unit curve and the initial air-conditioning electricity unit curve of the second type of electricity consumption corresponding to the date type can be calculated according to the following formula:

[0070]

[0071] p d,t,i,基础用户分时电量标幺曲线 is the basic user time-of-use electricity unit curve of the i-th second type of electricity consumption at t hours on day d, and p d,t,i,最初空调用电标幺曲线 is the initial air-conditioning electricity unit curve of the i-th second type of electricity consumption at t hours on day d.

[0072] If Q d,t,i,最初空调分时用电 < 0, the daily air-conditioning electricity consumption of the second type of electricity consumption corresponding to the date type is calculated according to the following formula:

[0073] Q d,t,i,一次校准空调分时用电 = Q d,t,i,修正空调分时用电 * p d,t,i,基础用户分时电量标幺曲线

[0074]

[0075] Q d,t,i,一次校准空调分时用电 is the daily air-conditioning electricity consumption of the i-th second type of electricity consumption at t hours on day d, and Q d,i,日空调用电量 is the daily air-conditioning electricity consumption of the i-th second type of electricity consumption on day d.

[0076] If Q d,t,i,最初空调分时用电 ≥ 0, the first calibrated air-conditioning time-of-use electricity of the second type of electricity consumption corresponding to the date type is calculated according to the following formula:

[0077] Q d,t,i,一次校准空调分时用电 = Q d,t,i,修正空调分时用电 * p d,t,i,最初空调用电标幺曲线

[0078]

[0079] As a preferred, for the second type of electricity consumption whose first type of electricity consumption is industry, Q d,t,i,一次校准空调分时用电 can also be second calibrated.

[0080] Specifically, the dispersion degree of the daily air-conditioning electricity of the second type of electricity of the first type of electricity of the industry type in each Celsius interval is calculated, and the daily air-conditioning electricity (primary calibration air-conditioning electricity) of each Celsius is adjusted according to the dispersion degree. The adjusted daily air-conditioning electricity is used to obtain the secondary calibration air-conditioning electricity data of the second type of electricity of the corresponding date according to the primary calibration air-conditioning electricity standard curve of the industry basic user of the corresponding date.

[0081] The dispersion degree of the daily air-conditioning electricity of the second type of electricity of the first type of electricity of the industry type in each Celsius interval can be calculated according to the following formula:

[0082]

[0083] Wherein, μ L,T,i is the average daily air-conditioning electricity of the i-th second type of electricity of the first type of electricity of the industry type corresponding to the L date type and T temperature, σ L,T,i is the standard deviation of the daily air-conditioning electricity of the i-th second type of electricity of the first type of electricity of the industry type corresponding to the L date type and T temperature, D is the number of days included in the L date type, and cv L,T,i is the dispersion degree of the daily air-conditioning electricity of the i-th second type of electricity of the first type of electricity of the industry type corresponding to the L date type and T temperature in each Celsius interval, and cv L,i is the dispersion degree of the daily air-conditioning electricity of the i-th second type of electricity of the first type of electricity of the industry type corresponding to the L date type.

[0084] When cv L,i is greater than the preset threshold value, the secondary calibration can be performed according to the following formula:

[0085]

[0086] Q d,t,二次校准空调分时用电 = Q d,i,日空调用电量 * p d,t,一次校准空调用电标幺曲线

[0087] Wherein, p d,t,一次校准空调用电标幺曲线 is the primary calibration air-conditioning electricity standard value of the i-th second type of electricity at the t-th time of the d day, Q d,t,i,一次校准空调分时用电 is the primary calibration air-conditioning electricity of the i-th second type of electricity at the t-th time of the d day, Q d,i,一次校准空调用电 is the primary calibration air-conditioning electricity of the i-th second type of electricity of the d day, and Q d,t,二次校准空调分时用电 is the secondary calibration air-conditioning electricity of the i-th second type of electricity at the t-th time of the d day.

[0088] Figure 2 is a schematic diagram of the corrected air-conditioning electricity before and after correction in an embodiment of the present application, as Figure 2As shown, the air-conditioning electricity before correction presents irregular changes with temperature, because the change of production electricity leads to the change of total electricity, and the traditional method cannot accurately identify this situation. The air-conditioning electricity after correction according to the above method increases with the increase of temperature, which accords with the understanding of the change of temperature adjustment load, and the real production load can be accurately calculated according to the corrected air-conditioning load.

[0089] In some embodiments, according to the daily air-conditioning electricity of each date type corresponding to the second type of electricity, the air-conditioning time-of-use electricity standard curve and the air-conditioning proportion of daily total electricity of the second type of electricity corresponding to the historical transaction year are determined, comprising:

[0090] According to the ratio of the calibrated air-conditioning time-of-use electricity of the date type corresponding to the second type of electricity to the daily air-conditioning electricity of each date type corresponding to the second type of electricity, the air-conditioning time-of-use electricity standard curve of the second type of electricity corresponding to the historical transaction year is determined.

[0091] According to the ratio of the daily air-conditioning electricity of each date type corresponding to the second type of electricity to the basic electricity of each date type corresponding to the second type of electricity, the air-conditioning proportion of daily total electricity of the second type of electricity corresponding to the historical transaction year is determined.

[0092] Specifically, the air-conditioning time-of-use electricity standard curve and the air-conditioning proportion of daily total electricity of the second type of electricity corresponding to the historical transaction year can be determined according to the following formula:

[0093]

[0094] Wherein, p d,t,i,空调分时用电标幺曲线 is the air-conditioning time-of-use electricity standard value of the i-th second type of electricity at the t-th time of the d-th day, Q d,y,i,最终空调分时用电 is the final air-conditioning time-of-use electricity of the i-th second type of electricity at the t-th time of the d-th day, Q d,t,i,最终空调分时用电 may be the primary calibrated air-conditioning time-of-use electricity or the secondary calibrated air-conditioning time-of-use electricity of the date type corresponding to the second type of electricity, r d,i,空调占日总量比例 is the air-conditioning proportion of daily total electricity of the i-th second type of electricity at the t-th time of the d-th day.

[0095] In some embodiments, according to the historical transaction year electricity data of a plurality of basic users and the air-conditioning time-of-use electricity standard curve and the air-conditioning proportion of daily total electricity of each type of user corresponding to the historical transaction year, the primary calibrated reference time-of-use electricity of each basic user in the historical transaction year is determined, comprising:

[0096] For each basic user, based on the historical transaction year electricity data of the basic user, the time-sharing real electricity of the basic user is determined, according to the historical transaction year air conditioner time-sharing electricity unit curve and air conditioner proportion of total daily amount corresponding to the second type of electricity of the basic user, the initial air conditioner electricity of the user and the initial air conditioner time-sharing electricity of the user of the basic user are determined, according to the initial air conditioner electricity of the user and the initial air conditioner time-sharing electricity of the user of the basic user, the first calibration air conditioner time-sharing electricity of the user of the basic user is determined, and according to the first calibration air conditioner time-sharing electricity of the user of the basic user, the first calibration reference time-sharing electricity of the basic user in the historical transaction year is determined.

[0097] Specifically, the first calibration reference time-sharing electricity of the basic user in the historical transaction year can be determined according to the following formula:

[0098]

[0099] q d,t,u,用户最初基准分时用电 = q d,t,u,用户分时真实用电 -q d,t,u,用户最初空调分时用电

[0100] q d,u,用户最初空调用电 = q d,u,用户真实用电 *r d,i,空调占日总量比例

[0101] q d,t,u,用户最初空调分时用电 = q d,u,用户最初空调用电 *p d,t,i,空调分时用电标幺曲线

[0102] If q d,t,u,用户最初基准分时用电 < 0:

[0103] q d,t,u,用户一次校准空调分时用电 = q d,u,用户最初空调用电 *q d,t,u,用户分时真实用电标幺曲线

[0104] If q d,t,u,用户最初基准分时用电 ≥ 0:

[0105] q d,t,u,用户一次校准空调分时用电 = q d,u,用户最初空调用电

[0106] q d,t,u,用户一次校准基准分时用电 = q d,t,u,用户分时真实用电 -q d,t,u,用户一次校准空调分时用电

[0107] Wherein, q d,u,用户真实用电 is the real electricity of the u-th basic user on the d-th day, q d,u,用户最初空调用电 is the initial air conditioner electricity of the u-th basic user on the d-th day, q d,t,u,用户最初基准分时用电 is the initial reference time-sharing electricity of the u-th basic user on the t-th hour of the d-th day, q d,t,u,用户最初空调分时用电标幺曲线 is the initial air conditioner time-sharing electricity unit value of the u-th basic user on the t-th hour of the d-th day, q d,t,u,用户最初空调分时用电a user initial air conditioner split-time electricity of the u-th basic user at the t-th time of the d-th day, q d,u,用户最初空调用电 a user initial air conditioner split-time electricity of the u-th basic user at the d-th day, q d,t,u,用户分时真实用电标幺曲线 a user split-time real electricity standard curve of the u-th basic user at the t-th time of the d-th day, q d,t,u,用户分时真实用电 a user split-time real electricity of the u-th basic user at the t-th time of the d-th day, q d,u,用户真实用电 a user split-time real electricity of the u-th basic user at the d-th day, q d,t,u,用户一次校准空调分时用电 a user split-time real electricity of the u-th basic user at the t-th time of the d-th day, q

[0108] Step 130, according to the once-calibrated basic split-time electricity of each basic user in the historical transaction year, performing electricity prediction in the future transaction year.

[0109] In some embodiments, step 130 specifically comprises:

[0110] performing electricity prediction in the future transaction year according to the once-calibrated basic split-time electricity of each basic user in the historical transaction year by an electricity prediction model, wherein the electricity prediction model can be a convolutional neural network model.

[0111] In some embodiments, performing electricity prediction in the future transaction year according to the once-calibrated basic split-time electricity of each basic user in the historical transaction year by an electricity prediction model comprises:

[0112] determining daily electricity consumption of the basic user day by day according to the once-calibrated basic split-time electricity and daily air conditioner electricity consumption of the basic user in the historical transaction year by the electricity prediction model;

[0113] determining monthly electricity consumption of the basic user month by month according to the daily electricity consumption of the basic user by the electricity prediction model;

[0114] determining annual electricity consumption of the basic user in the future transaction year according to the monthly electricity consumption of the basic user by the electricity prediction model.

[0115] Finally, it should be understood that the embodiments described herein are merely intended to illustrate the principles of the embodiments described herein. Other variations can also be within the scope of the present description. Therefore, as an example but not limitation, alternative configurations of the embodiments described herein can be considered consistent with the teachings of the present description. Accordingly, the embodiments of the present description are not limited to the embodiments explicitly introduced and described in the present description.

Claims

1. An AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling, characterized in that, include: Obtain electricity consumption data for multiple basic users over historical transaction years; Based on the electricity consumption data of multiple basic users in the historical transaction years, the types of basic users are determined. Based on the electricity consumption data of multiple basic users in the historical transaction years, the per-unit curve of air conditioning time-of-use electricity consumption and the proportion of air conditioning to the daily total for each type of user in the corresponding historical transaction years are determined. Based on the electricity consumption data of multiple basic users in the historical transaction years and the per-unit curve of air conditioning time-of-use electricity consumption and the proportion of air conditioning to the daily total for each type of user in the corresponding historical transaction years, the calibration benchmark time-of-use electricity consumption of each basic user in the historical transaction year is determined. Among them, the types of basic users include multiple Class I electricity consumption types, and each Class I electricity consumption type includes at least one Class II electricity consumption type. Based on the time-of-use electricity consumption of each basic user in a historical trading year, electricity consumption forecasts for future trading years are made. Calculate the daily electricity consumption for the corresponding date type for the second type of electricity consumption using the following formula: ； in, This represents the daily electricity consumption of the i-th type of Class II electricity consumption on day d. For the electricity consumption of the i-th type of Class II electricity consumption in hour t of day d after correction; Among them, if ,but ,like ,but ; The daily electricity consumption for the corresponding date type of the second type of electricity consumption is calculated based on the following formula: ； in, This refers to the daily electricity consumption of the i-th type of Class II electricity consumption on day d. The basic user time-of-use electricity per-unit curve and the initial air conditioning electricity per-unit curve for the corresponding date type of the second type of electricity consumption are calculated using the following formulas: ； ； in, The per-unit curve of time-of-use electricity consumption for the i-th type of Class II electricity consumption in hour t on day d is shown. The initial per-unit curve of air conditioning electricity consumption for the i-th type of Class II electricity consumption in hour t on day d; like The daily air conditioning electricity consumption for the corresponding date type of the second type of electricity consumption is calculated according to the following formula: ； ； in, This refers to the daily air conditioning electricity consumption of the i-th type of Class II electricity consumption in hour t of day d. This refers to the daily air conditioning electricity consumption for the i-th type of Class II electricity consumption on day d. like The time-of-use electricity consumption for the first calibration of the air conditioner for the corresponding date type of the second type of electricity consumption is calculated according to the following formula: ； 。 2. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling as described in claim 1, characterized in that, Based on the electricity consumption data of multiple basic users over historical transaction years, the types of basic users are determined, including: Identify a type of electricity consumption for each basic user, wherein the type of electricity consumption is industrial, commercial, office, or charging station. For each type of electricity consumption, historical electricity consumption characteristics of each basic user included in the first type of electricity consumption are extracted from the electricity consumption data of each basic user in the historical transaction year. Based on the historical electricity consumption characteristics of each basic user included in the first type of electricity consumption, the second type of electricity consumption of each basic user included in the first type of electricity consumption is determined.

3. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling as described in claim 2, characterized in that, From the historical transaction year's electricity consumption data of each basic user included in the first type of electricity consumption, historical electricity consumption characteristics of each basic user included in the first type of electricity consumption are extracted. Based on the historical electricity consumption characteristics of each basic user included in the first type of electricity consumption, a second type of electricity consumption for each basic user included in the first type of electricity consumption is determined, including: Categorize dates and identify multiple date types; For each basic user, the electricity consumption characteristics of the basic user for each date type are determined based on the electricity consumption data of the basic user's historical transaction years. For any two basic users included in the aforementioned electricity consumption type, the similarity between the two basic users is calculated based on their electricity consumption characteristics for each date type. Based on the similarity between any two basic users included in the first type of electricity consumption, a clustering algorithm is used to divide the multiple basic users included in the first type of electricity consumption into multiple basic user groups, wherein one basic user group corresponds to one second type of electricity consumption.

4. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling as described in claim 2 or 3, characterized in that, Based on the electricity consumption data of multiple basic users over historical trading years, the per-unit curve for air conditioning time-of-use electricity consumption and the proportion of air conditioning in the daily total are determined for each type of user in the corresponding historical trading year, including: For each type of electricity consumption, the time-of-use electricity consumption of the basic users of the type of electricity consumption is determined based on the electricity consumption data of each basic user in the historical transaction year. Based on the time-of-use electricity consumption of the basic users of the type of electricity consumption, the benchmark time-of-use electricity consumption for each date type of the type of electricity consumption is determined. Based on the benchmark time-of-use electricity consumption for each date type of the type of electricity consumption, the daily air conditioning electricity consumption for each date type of the type of electricity consumption is determined. Based on the daily air conditioning electricity consumption for each date type of the type of electricity consumption, the per-unit curve of air conditioning time-of-use electricity consumption and the proportion of air conditioning to the daily total for the historical transaction year of the type of electricity consumption are determined.

5. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling according to claim 3, characterized in that, Based on the benchmark time-of-use electricity consumption for each date type corresponding to the two types of electricity consumption, the daily air conditioning electricity consumption for each date type corresponding to the two types of electricity consumption is determined, including: For each date type, the initial air conditioning time-of-use electricity consumption for the corresponding date type is calculated based on the difference between the basic user time-of-use electricity consumption of the two types of electricity consumption and the benchmark time-of-use electricity consumption of the corresponding date type. The initial air conditioning time-of-use electricity consumption for the corresponding date type of the two types of electricity consumption is calibrated to determine the calibrated air conditioning time-of-use electricity consumption for the corresponding date type of the two types of electricity consumption. Based on the calibrated air conditioning time-of-use electricity consumption for the corresponding date type of the two types of electricity consumption, the daily air conditioning electricity consumption for the corresponding date type of the two types of electricity consumption is determined.

6. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling as described in claim 5, characterized in that, The initial time-of-use electricity consumption of air conditioners for the corresponding date type of the two types of electricity consumption is calibrated to determine the daily electricity consumption of air conditioners for the corresponding date type of the two types of electricity consumption, including: Based on the initial time-of-use electricity consumption of air conditioning for the corresponding date type of the two types of electricity consumption, determine the daily electricity consumption of air conditioning for the corresponding date type of the two types of electricity consumption; Based on the time-of-use electricity consumption of the basic users for the two types of electricity consumption, determine the daily electricity consumption for the corresponding date type for the two types of electricity consumption; Based on the time-of-use electricity consumption of the basic users of the two types of electricity consumption and the daily electricity consumption of the corresponding date type of the two types of electricity consumption, calculate the per-unit curve of the time-of-use electricity consumption of the basic users of the two types of electricity consumption for the corresponding date type; Based on the initial time-of-use air conditioning electricity consumption and daily air conditioning electricity consumption of the corresponding date type for the two types of electricity consumption, determine the per-unit curve of the initial air conditioning electricity consumption of the corresponding date type for the two types of electricity consumption; Based on the per-unit curve of basic user time-of-use electricity consumption and the initial per-unit curve of air conditioning electricity consumption and the initial time-of-use electricity consumption of air conditioning for the corresponding date type of the two types of electricity consumption, the daily air conditioning electricity consumption for the corresponding date type of the two types of electricity consumption is determined.

7. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling as described in claim 5, characterized in that, Based on the daily air conditioning electricity consumption for each date type corresponding to the two types of electricity consumption, determine the per-unit curve of air conditioning time-of-use electricity consumption and the proportion of air conditioning in the daily total for the corresponding historical transaction year for the two types of electricity consumption, including: Based on the ratio of the calibrated air conditioner time-of-use electricity consumption for the corresponding date type of the two types of electricity consumption to the daily air conditioner electricity consumption for each date type of the two types of electricity consumption, the per-unit curve of air conditioner time-of-use electricity consumption for the corresponding historical transaction year of the two types of electricity consumption is determined; Based on the ratio of the daily air conditioning electricity consumption for each date type corresponding to the two types of electricity consumption to the basic electricity consumption for each date type corresponding to the two types of electricity consumption, the proportion of air conditioning in the total daily amount for the corresponding historical transaction year of the two types of electricity consumption is determined.

8. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling according to any one of claims 1-3, characterized in that, Based on the electricity consumption data of multiple basic users in historical trading years and the per-unit curve of air conditioning time-of-use electricity consumption and the proportion of air conditioning in the daily total for each type of user in the corresponding historical trading years, the calibration benchmark time-of-use electricity consumption for each basic user in the historical trading year is determined, including: For each basic user, based on the user's historical transaction year electricity consumption data, the user's actual time-of-use electricity consumption is determined. Based on the user's second-class electricity consumption type corresponding to the historical transaction year's air conditioning time-of-use electricity consumption per unit and the proportion of air conditioning to the daily total, the user's initial air conditioning electricity consumption and initial air conditioning time-of-use electricity consumption are determined. Based on the user's initial air conditioning electricity consumption and initial air conditioning time-of-use electricity consumption, the user's primary calibration air conditioning time-of-use electricity consumption is determined. Based on the user's primary calibration air conditioning time-of-use electricity consumption, the user's primary calibration benchmark time-of-use electricity consumption in the historical transaction year is determined.

9. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling according to any one of claims 1-3, characterized in that, Based on the time-of-use electricity consumption of each basic user in a historical trading year, electricity consumption forecasts for future trading years are made, including: The electricity consumption forecasting model uses the time-of-use electricity consumption of each basic user based on a calibration benchmark in a historical trading year to predict the electricity consumption of future trading years.

10. The AI-powered intelligent electricity consumption prediction method based on electricity consumption behavior profiling according to claim 9, characterized in that, The electricity consumption forecasting model predicts the electricity consumption for future trading years based on the time-of-use electricity consumption of each basic user in a historical trading year, using a calibration benchmark. This includes: The daily electricity consumption of basic users is determined day by day by using the electricity consumption prediction model based on the time-of-use electricity consumption and daily air conditioning electricity consumption of basic users in the historical transaction year calibration benchmark. The monthly electricity consumption of basic users is determined month by month based on their daily electricity consumption using an electricity consumption forecasting model. The electricity consumption forecasting model determines the annual electricity consumption of basic users in future trading years based on their monthly electricity consumption.

Images