Household electricity load prediction method and device

Abstract

The application provides a household electricity load prediction method and device, and the method comprises the following steps: constructing a household electricity load model; obtaining a load calculation value of a to-be-predicted unit time period according to the household electricity load model; and correcting the load calculation value of the to-be-predicted unit time period through a neural network to obtain a load prediction value of the to-be-predicted unit time period. The application can conveniently and accurately realize the prediction of the household electricity load, has low cost, and has a wide application range.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, specifically to a method and device for predicting household electricity load. Background Technology

[0002] Load forecasting is a crucial aspect of energy management, providing key data for its implementation. Load forecasting technology is primarily used in provincial and municipal power grids or large industrial parks, applied in scenarios such as power generation planning. This level of load forecasting involves a wide range of areas and exhibits aggregated load types. Current load forecasting technologies can be broadly categorized into four types: naive methods, classical linear methods, machine learning methods, and deep learning methods. All of these methods heavily rely on historical data acquisition, especially deep learning methods, where model training requires data from the past year or even several years.

[0003] For residential electricity load forecasting, especially for single households, there are two main challenges: firstly, the storage capacity for historical electricity data is very limited; secondly, load usage is random and highly uncertain. Therefore, how to conveniently and accurately forecast household electricity load has become a pressing technical problem that needs to be solved. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides a method and device for predicting household electricity load, which can conveniently and accurately predict household electricity load, and has low cost and wide applicability.

[0005] The technical solution adopted in this invention is as follows:

[0006] A method for predicting household electricity load includes the following steps: constructing a household electricity load model; obtaining the calculated load value for a unit time period to be predicted based on the household electricity load model; and correcting the calculated load value for the unit time period to be predicted using a neural network to obtain the predicted load value for the unit time period.

[0007] The household electricity load model includes a long-term stable load component, a medium-term fluctuating load component, and a short-term random load component.

[0008] The long-term stable load component is constructed based on objective factors affecting the load of self-consuming equipment, the medium-term fluctuating load component is constructed based on meteorological and seasonal factors, and the short-term random load component is constructed based on historical load calculation values ​​and historical load forecast values.

[0009] The objective factors affecting the load of self-consuming electrical equipment include the size of the house, the number of people living in the house, and the geographical location of the house. The meteorological factors include temperature, humidity, radiation, and weather type.

[0010] The long-term stable load component is:

[0011] L long =ω1*ω2*ω3

[0012] Among them, L long The long-term stable load components are represented by ω1, ω2, and ω3, which are the electricity load coefficients related to the area of ​​the house, the number of people living in the house, and the geographical location of the house, respectively.

[0013] The medium-term fluctuating load component is:

[0014] L mid =(L mid，temp +L mid，hum +L min，irr )*j(ws)

[0015] Among them, L mid L represents the medium-term fluctuating load component. mid,temp L mid,hum L mid,irr These are the electrical load sub-components related to the temperature, the electrical load sub-component related to the humidity, and the electrical load sub-component related to the radiation, respectively, where j(ws) is a weather and seasonal correction factor.

[0016] The short-term random load component is:

[0017]

[0018] Among them, L short L represents the short-term random load component. n This represents the calculated load value over a time period of n. This represents the load forecast value for a time period n, where n = t-1, t-2, or t-3, and μ is the mean of the historical load calculation value sequence. θ1 and θ2 are coefficients.

[0019] The method of correcting the load calculation value of the unit time period to be predicted by means of a neural network includes: constructing a BP neural network; obtaining the error between the load prediction value and the load calculation value of the previous unit time period; updating the connection weights and thresholds of the BP neural network using the error; and substituting the load calculation value of the unit time period to be predicted into the updated BP neural network to obtain the load prediction value of the unit time period to be predicted.

[0020] A household electricity load forecasting device includes: a construction module for constructing a household electricity load model; an acquisition module for acquiring a calculated load value for a unit time period to be predicted based on the household electricity load model; and a correction module for correcting the calculated load value for the unit time period to be predicted using a neural network to obtain a predicted load value for the unit time period.

[0021] The correction module is specifically used for: constructing a BP neural network; obtaining the error between the load forecast value and the calculated load value of the previous unit time period of the unit time period to be predicted; updating the connection weights and thresholds of the BP neural network using the error; and substituting the calculated load value of the unit time period to be predicted into the updated BP neural network to obtain the load forecast value of the unit time period to be predicted.

[0022] The beneficial effects of this invention are:

[0023] This invention establishes a household electricity load model and dynamically corrects the model's calculation results. It eliminates the need to store large amounts of historical data as training samples and is not limited by geographical location. As a result, it can conveniently and accurately predict household electricity load at a low cost and with a wide range of applications. Attached Figure Description

[0024] Figure 1 This is a flowchart of a household electricity load prediction method according to an embodiment of the present invention;

[0025] Figure 2 This is a block diagram of a household electricity load prediction device according to an embodiment of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] like Figure 1 As shown, the household electricity load prediction method of this invention includes the following steps:

[0028] S1, Construct a household electricity load model.

[0029] The household electricity load model in this embodiment of the invention is a mathematical model, which includes multiple load components.

[0030] In one embodiment of the present invention, the household electricity load model may include three components: a long-term stable load component, a medium-term fluctuating load component, and a short-term random load component.

[0031] The household electricity load model is as follows:

[0032] L t =L long +L mid +L short

[0033] Among them, L long L represents the long-term stable load component. mid L represents the medium-term fluctuating load component. short L represents the short-term random load component. t The household electricity load during time period t is represented by L. long L mid L short Summation calculation yields that, in this embodiment of the invention, t represents the unit time period to be predicted, i.e., L. t This represents the calculated load value for the unit time period to be predicted.

[0034] In one specific embodiment of the present invention, a day can be used as a unit of time.

[0035] Long-term stable loads are less affected by users' subjective behavior, including self-consuming electrical appliances such as refrigerators and sockets. Therefore, long-term stable load components can be constructed based on objective factors affecting the load of self-consuming electrical appliances. Objective factors affecting the load of self-consuming electrical appliances include the size of the household, the number of people living in the household, and the geographical location of the house. In one embodiment of the present invention, the long-term stable load component is:

[0036] L long =ω1*ω2*ω3

[0037] Among them, ω1, ω2, and ω3 are the electricity load coefficients related to the area of ​​the house, the number of people living in the house, and the geographical location of the house, respectively. These coefficients can be obtained by summarizing or fitting experimental data or survey data.

[0038] In a specific embodiment of the present invention, the relationship between the electricity load factor ω1 and the house area S is shown in Table 1.

[0039] Table 1

[0040]

[0041]

[0042] In a specific embodiment of the present invention, the relationship between the electricity load factor ω2 and the number of household residents nor is as follows:

[0043]

[0044] In a specific embodiment of the present invention, the relationship between the electricity load factor ω3 and the geographical location of the house is shown in Table 2.

[0045] Table 2

[0046]

[0047] Medium-term fluctuating loads are highly correlated with weather and user behavior, including typical appliances such as air conditioners, electric heaters, and lighting. Therefore, medium-term fluctuating load components can be constructed based on meteorological and seasonal factors. Meteorological factors include temperature, humidity, radiation, and weather type. In one embodiment of the present invention, electricity load sub-components corresponding to temperature, humidity, and radiation can be obtained, and the sum of the above electricity load sub-components can be corrected based on weather type and seasonal factors. Specifically, the medium-term fluctuating load components are:

[0048] L mid =(L mid，temp +L mid，hum +L mid，irr )*j(ws)

[0049] Among them, L mid,temp L mid,hum L mid,irr These are the electrical load sub-components related to temperature, humidity, and radiation, respectively, with j(ws) representing the weather and seasonal correction factor. Each component and correction factor can also be obtained by summarizing or fitting experimental or survey data.

[0050] In a specific embodiment of the present invention, the electrical load subcomponent related to temperature (Temp) is:

[0051] L mid，temp =0.0482Temp 2 0.371Temp - 5.997

[0052] In a specific embodiment of the present invention, the electrical load sub-component related to humidity (Hum) is:

[0053] L mid，hum = -0.332Hum + 3.034

[0054] In a specific embodiment of the present invention, the electrical load subcomponent related to the radiation quantity Irr is:

[0055] Lmid，irr =1.528Irr-1.144

[0056] In a specific embodiment of the present invention, the weather and season correction factors are:

[0057] j(ws)=1 / c pq

[0058] Among them, c pq c represents the regression constant term when the month is p and the weather type is q. pq The values ​​are shown in Table 3.

[0059] Table 3

[0060] 3-5(spring) 1.493 1.689 1.949 3.571 6.536 6-8(summer) 1.405 1.443 1.859 2.985 4.184 9-11(autumn) 1.443 1.629 2.020 4.219 6.289 12-2(winter) 1.733 1.908 2.625 4.651 7.092

[0061] Short-term random loads are unconventional and highly susceptible to special events such as user leave on weekdays, equipment failures, and power outages. Therefore, they can be described using the Box-Jenkins time series method. In one embodiment of this invention, the short-term random load component is constructed based on historical load calculations and historical load forecasts. The short-term random load component is as follows:

[0062]

[0063] Among them, L n This represents the calculated load value over a time period of n. This represents the load forecast value for a time period n, where n = t-1, t-2, or t-3, t-1 represents the unit time period preceding t, t-2 represents the unit time period preceding t-1, and t-3 represents the unit time period preceding t-2. μ is the mean of the historical load calculation value series. θ1 and θ2 are coefficients. Each of these coefficients can be given an initial value and then updated as the historical load calculation value sequence changes.

[0064] S2, obtain the load calculation value for the unit time period to be predicted based on the household electricity load model.

[0065] S3 uses a neural network to correct the calculated load value for the unit time period to be predicted, so as to obtain the predicted load value for the unit time period.

[0066] In one embodiment of the present invention, a BP neural network can be constructed to obtain the load forecast value of the previous unit time period of the unit time period to be predicted. With load calculation value L t-1 The error between the values ​​is used to update the connection weights and thresholds (bias) of the BP neural network, and finally the load value L for the unit time period to be predicted is calculated. tSubstitute the values ​​into the updated BP neural network to obtain the load forecast for the unit time period to be predicted.

[0067] In other words, the embodiments of the present invention can combine the calculated and predicted load values ​​of the previous unit time period to dynamically correct the load value of the unit time period to be predicted, thereby obtaining a more accurate load prediction value.

[0068] Specifically, the BP neural network can be initialized first, defining the number of input layer nodes as 3, the number of hidden layer nodes as 2, and the number of output layer nodes as 1. The input is {x1, x2, x3}, and the output is Y. The weights from the input layer to the hidden layer are ω. ij The weights from the hidden layer to the output layer are ω. jk The threshold from the input layer to the hidden layer is a. j The threshold from the hidden layer to the output layer is b. k The excitation function is Given a learning rate η, where i = 1, 2, 3, j = 1, 2, k = 1.

[0069] Then the hidden layer output H is performed. j Calculation:

[0070]

[0071] Output layer output O k Calculation:

[0072]

[0073] Furthermore, the calculated load value L for the previous unit time period t-1 (Including the long-term stable load component, medium-term fluctuating load component, and short-term random load component L from the previous unit time period) t-1,long L t-1,mid L t-1,short ) and load forecast Update the input / output sequence:

[0074]

[0075] Next, we will calculate the error:

[0076]

[0077] Perform weight updates:

[0078]

[0079] Perform threshold update:

[0080]

[0081] Finally, calculate the updated output of the BP neural network:

[0082]

[0083] The final load forecast value It can serve as an important data basis for family energy management.

[0084] The household electricity load prediction method according to embodiments of the present invention establishes a household electricity load model and dynamically corrects the calculation results of the model. It does not require storing a large amount of historical data as training samples and is not limited by geographical location. Therefore, it can conveniently and accurately predict household electricity load, and has low cost and wide applicability.

[0085] Corresponding to the household electricity load prediction method in the above embodiments, the present invention also proposes a household electricity load prediction device.

[0086] like Figure 2 As shown, the household electricity load forecasting device of this embodiment includes a construction module 10, an acquisition module 20, and a correction module 30. The construction module 10 is used to construct a household electricity load model; the acquisition module 20 is used to acquire the calculated load value for a given time period based on the household electricity load model; and the correction module 30 is used to correct the calculated load value for the given time period using a neural network to obtain the predicted load value for that time period.

[0087] The household electricity load model in this embodiment of the invention is a mathematical model, which includes multiple load components.

[0088] In one embodiment of the present invention, the household electricity load model may include three components: a long-term stable load component, a medium-term fluctuating load component, and a short-term random load component.

[0089] The household electricity load model is as follows:

[0090] L t =L long +L mid +L short

[0091] Among them, L long L represents the long-term stable load component. mid L represents the medium-term fluctuating load component. short L represents the short-term random load component. t The household electricity load during time period t is represented by L. long L mid L short Summation calculation yields that, in this embodiment of the invention, t represents the unit time period to be predicted, i.e., L.t This represents the calculated load value for the unit time period to be predicted.

[0092] In one specific embodiment of the present invention, a day can be used as a unit of time.

[0093] Long-term stable loads are less affected by users' subjective behavior, including self-consuming electrical appliances such as refrigerators and sockets. Therefore, long-term stable load components can be constructed based on objective factors affecting the load of self-consuming electrical appliances. Objective factors affecting the load of self-consuming electrical appliances include the size of the household, the number of people living in the household, and the geographical location of the house. In one embodiment of the present invention, the long-term stable load component is:

[0094] L long =ω1*ω2*ω3

[0095] Among them, ω1, ω2, and ω3 are the electricity load coefficients related to the area of ​​the house, the number of people living in the house, and the geographical location of the house, respectively. These coefficients can be obtained by summarizing or fitting experimental data or survey data.

[0096] In a specific embodiment of the present invention, the relationship between the electricity load factor ω1 and the house area S is shown in Table 1.

[0097] In a specific embodiment of the present invention, the relationship between the electricity load factor ω2 and the number of household residents nor is as follows:

[0098]

[0099] In a specific embodiment of the present invention, the relationship between the electricity load factor ω3 and the geographical location of the house is shown in Table 2.

[0100] Medium-term fluctuating loads are highly correlated with weather and user behavior, including typical appliances such as air conditioners, electric heaters, and lighting. Therefore, medium-term fluctuating load components can be constructed based on meteorological and seasonal factors. Meteorological factors include temperature, humidity, radiation, and weather type. In one embodiment of the present invention, electricity load sub-components corresponding to temperature, humidity, and radiation can be obtained, and the sum of the above electricity load sub-components can be corrected based on weather type and seasonal factors. Specifically, the medium-term fluctuating load components are:

[0101] L mid =(L mid，temp +L mid，hum +L mid，irr )*j(ws)

[0102] Among them, L mid,temp L mid,hum L mid,irrThese are the electrical load sub-components related to temperature, humidity, and radiation, respectively, with j(ws) representing the weather and seasonal correction factor. Each component and correction factor can also be obtained by summarizing or fitting experimental or survey data.

[0103] In a specific embodiment of the present invention, the electrical load subcomponent related to temperature (Temp) is:

[0104] L mid，temp =0.0482Temp 2 -0.371Temp-5.997

[0105] In a specific embodiment of the present invention, the electrical load sub-component related to humidity (Hum) is:

[0106] L mid，hum = -0.332Hum + 3.034

[0107] In a specific embodiment of the present invention, the electrical load subcomponent related to the radiation quantity Irr is:

[0108] L mid，irr =1.528Irr-1.144

[0109] In a specific embodiment of the present invention, the weather and season correction factors are:

[0110] j(ws)=1 / c pq

[0111] Among them, c pq c represents the regression constant term when the month is p and the weather type is q. pq The values ​​are shown in Table 3.

[0112] Short-term random loads are unconventional and highly susceptible to special events such as user leave on weekdays, equipment failures, and power outages. Therefore, they can be described using the Box-Jenkins time series method. In one embodiment of this invention, the short-term random load component is constructed based on historical load calculations and historical load forecasts. The short-term random load component is as follows:

[0113]

[0114] Among them, L n This represents the calculated load value over a time period of n. This represents the load forecast value for a time period n, where n = t-1, t-2, or t-3, t-1 represents the unit time period preceding t, t-2 represents the unit time period preceding t-1, and t-3 represents the unit time period preceding t-2. μ is the mean of the historical load calculation value series. θ1 and θ2 are coefficients. Each of these coefficients can be given an initial value and then updated as the historical load calculation value sequence changes.

[0115] In one embodiment of the present invention, the correction module 30 may specifically construct a BP neural network to obtain the load forecast value of the previous unit time period of the unit time period to be predicted. With load calculation value L t-1 The error between the values ​​is calculated, and the connection weights and thresholds (bias) of the BP neural network are updated using the error. Finally, the load value L for the unit time period to be predicted is calculated. t Substitute the values ​​into the updated BP neural network to obtain the load forecast for the unit time period to be predicted.

[0116] In other words, the embodiments of the present invention can combine the calculated and predicted load values ​​of the previous unit time period to dynamically correct the load value of the unit time period to be predicted, thereby obtaining a more accurate load prediction value.

[0117] Specifically, the correction process of correction module 30 is as follows:

[0118] First, we can initialize the BP neural network, defining the input layer as having 3 nodes, the hidden layer as having 2 nodes, and the output layer as having 1 node. The input is {x1, x2, x3}, and the output is Y. The weights from the input layer to the hidden layer are ω. ij The weights from the hidden layer to the output layer are ω. jk The threshold from the input layer to the hidden layer is a. j The threshold from the hidden layer to the output layer is b. k The excitation function is Given a learning rate η, where i = 1, 2, 3, j = 1, 2, k = 1.

[0119] Then the hidden layer output H is performed. j Calculation:

[0120]

[0121] Output layer output O k Calculation:

[0122]

[0123] Furthermore, the calculated load value L for the previous unit time period t-1 (Including the long-term stable load component, medium-term fluctuating load component, and short-term random load component L from the previous unit time period) t-1,long L t-1,mid L t-1,short ) and load forecast Update the input / output sequence:

[0124]

[0125] Next, we will calculate the error:

[0126]

[0127] Perform weight updates:

[0128]

[0129] Perform threshold update:

[0130]

[0131] Finally, calculate the updated output of the BP neural network:

[0132]

[0133] The final load forecast value It can serve as an important data basis for family energy management.

[0134] The household electricity load prediction device according to embodiments of the present invention establishes a household electricity load model and dynamically corrects the calculation results of the model. It does not require storing a large amount of historical data as training samples and is not limited by geographical location. Therefore, it can conveniently and accurately predict household electricity load, and has low cost and wide applicability.

[0135] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. "A plurality of" means two or more, unless otherwise explicitly specified.

[0136] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0137] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0138] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0139] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0140] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0141] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0142] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0143] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0144] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for predicting household electricity load, characterized in that, Includes the following steps: Construct a household electricity load model; The calculated load value for the unit time period to be predicted is obtained based on the household electricity load model. The calculated load value for the unit time period to be predicted is corrected by a neural network to obtain the predicted load value for the unit time period. The load calculation value for the unit time period to be predicted is corrected using a neural network, specifically including: constructing a BP neural network; obtaining the error between the load prediction value and the load calculation value of the previous unit time period; updating the connection weights and threshold of the BP neural network using the error; and substituting the load calculation value of the unit time period to be predicted into the updated BP neural network to obtain the load prediction value for the unit time period to be predicted. The household electricity load model includes a long-term stable load component, a medium-term fluctuating load component, and a short-term random load component. The long-term stable load component is: Among them, L long The long-term stable load components are represented by ω1, ω2, and ω3, which are respectively the electricity load coefficients related to the house area, the number of people living in the house, and the geographical location of the house. The medium-term fluctuating load component is: Among them, L mid L represents the medium-term fluctuating load component. mid,temp L mid,hum L mid,irr These are the electrical load sub-components related to the temperature, the electrical load sub-component related to the humidity, and the electrical load sub-component related to the radiation, respectively, where j(ws) is a weather and seasonal correction factor. The short-term random load component is: Among them, L short L represents the short-term random load component. n This represents the calculated load value over a time period of n. This represents the load forecast value for a time period n, where n = t-1, t-2, or t-3, and μ is the mean of the historical load calculation value sequence.

1.

2. θ1 and θ2 are coefficients.

2. A household electricity load prediction device, characterized in that, include: The construction module is used to build a household electricity load model; The acquisition module is used to acquire the load calculation value of the unit time period to be predicted based on the household electricity load model; The correction module is used to correct the calculated load value for the unit time period to be predicted using a neural network, so as to obtain the predicted load value for the unit time period. The correction module is specifically used for: constructing a BP neural network; obtaining the error between the load forecast value and the load calculation value of the previous unit time period of the unit time period to be predicted; and updating the connection weights and thresholds of the BP neural network using the error. The calculated load value for the unit time period to be predicted is substituted into the updated BP neural network to obtain the predicted load value for the unit time period to be predicted. The household electricity load model includes a long-term stable load component, a medium-term fluctuating load component, and a short-term random load component. The long-term stable load component is: Among them, L long The long-term stable load components are represented by ω1, ω2, and ω3, which are respectively the electricity load coefficients related to the house area, the number of people living in the house, and the geographical location of the house. The medium-term fluctuating load component is: Among them, L mid L represents the medium-term fluctuating load component. mid,temp L mid,hum L mid,irr These are the electrical load sub-components related to the temperature, the electrical load sub-component related to the humidity, and the electrical load sub-component related to the radiation, respectively, where j(ws) is a weather and seasonal correction factor. The short-term random load component is: Among them, L short L represents the short-term random load component. n This represents the calculated load value over a time period of n. This represents the load forecast value for a time period n, where n = t-1, t-2, or t-3, and μ is the mean of the historical load calculation value sequence.

1.

2. θ1 and θ2 are coefficients.

Images