CN116109016A - Energy consumption prediction model training and energy consumption prediction method and device - Google Patents
Energy consumption prediction model training and energy consumption prediction method and device Download PDFInfo
- Publication number
- CN116109016A CN116109016A CN202310379038.3A CN202310379038A CN116109016A CN 116109016 A CN116109016 A CN 116109016A CN 202310379038 A CN202310379038 A CN 202310379038A CN 116109016 A CN116109016 A CN 116109016A
- Authority
- CN
- China
- Prior art keywords
- energy consumption
- communities
- community
- prediction model
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005265 energy consumption Methods 0.000 title claims abstract description 217
- 238000000034 method Methods 0.000 title claims abstract description 81
- 238000012549 training Methods 0.000 title claims abstract description 66
- 239000011159 matrix material Substances 0.000 claims abstract description 98
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000002372 labelling Methods 0.000 claims abstract description 18
- 238000003062 neural network model Methods 0.000 claims description 28
- 238000004590 computer program Methods 0.000 claims description 19
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 6
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 18
- 230000006870 function Effects 0.000 description 16
- 238000012512 characterization method Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 5
- 239000007789 gas Substances 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 239000002737 fuel gas Substances 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Mathematical Optimization (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Computational Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Primary Health Care (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Algebra (AREA)
- Educational Administration (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The specification relates to the technical field of artificial intelligence, in particular to an energy consumption prediction model training method and device. The energy consumption prediction model training method comprises the steps of carrying out labeling processing on a plurality of pieces of user energy consumption information collected by each community in a plurality of communities, determining a label corresponding to sample data of each community, wherein the sample data is determined by the plurality of pieces of user energy consumption information; determining an adjacency matrix according to the similarity between every two communities; and training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model. By using the embodiment of the specification, the relevance among communities is considered in the energy consumption prediction process, and the accuracy of the energy consumption prediction result is improved.
Description
Technical Field
The specification relates to the technical field of artificial intelligence, in particular to an energy consumption prediction model training method and device.
Background
When the energy consumption is predicted for each community, only the historical energy consumption condition of users in the community is considered at present. Specifically, based on a statistical method (such as multiple linear regression and autoregressive moving average model), carrying out regression treatment on the historical energy consumption to obtain a model for predicting the future resource consumption; based on machine learning and deep learning models (for example, a long-term and short-term memory neural network model is used), model training is carried out by taking historical energy consumption as a training sample, and a model for predicting future resource consumption is obtained. When the future resource consumption is predicted based on a statistical method, the method is simple in model and high in use efficiency, but only a linear relation can be captured, a predicted scene is single, and the energy consumption conditions of a plurality of time steps and a long time in the future are difficult to predict. When future resource consumption is predicted based on a machine learning and deep learning model, the relevance between communities is not considered, and prediction is performed only according to the gas consumption condition of a single community, so that the accuracy of a prediction result is low.
How to accurately predict a plurality of time steps and the energy consumption condition for a long time in the future on the basis of considering the relevance between communities is a problem to be solved in the prior art.
Disclosure of Invention
In order to solve the problems in the prior art, the embodiment of the specification provides an energy consumption prediction model training and energy consumption prediction method and device, so that the relevance among communities is considered in the energy consumption prediction process, and the accuracy of an energy consumption prediction result is improved.
In order to solve the technical problems, the specific technical scheme in the specification is as follows:
in one aspect, embodiments of the present disclosure provide a method for training an energy usage prediction model, comprising,
labeling a plurality of pieces of user energy consumption information acquired by each community in a plurality of communities, and determining a label corresponding to sample data of each community, wherein the sample data is determined by the plurality of pieces of user energy consumption information;
determining an adjacency matrix according to the similarity between every two communities; and
and training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model.
Further, the labeling processing is performed on the plurality of user energy consumption information collected for each community in the plurality of communities, and determining the label corresponding to the sample data of each community further includes:
For each community, determining a plurality of users included in the community respectively;
calculating an average value of energy consumption data corresponding to each user, wherein the user energy consumption information comprises the energy consumption data; and
and taking the average value as the label corresponding to the sample data.
Further, the determining a adjacency matrix based on the similarity between each two communities further comprises,
determining the position information of each community respectively;
based on the position information, respectively determining first similarity between every two communities to obtain a first sub-adjacency matrix;
based on the sample data corresponding to each two communities, respectively determining a second similarity between each two communities to obtain a second sub-adjacency matrix; and
the adjacency matrix is determined based on the first sub-adjacency matrix and the second sub-adjacency matrix.
Further, the determining a first similarity between each two communities based on the location information, respectively, to obtain a first sub-adjacency matrix further includes,
based on the position information of every two communities, respectively determining Euclidean distance values;
Judging whether the Euclidean distance value is smaller than a distance threshold value or not;
under the condition that the Euclidean distance value is smaller than a distance threshold value, determining that a first similarity between two communities corresponding to the Euclidean distance value is a first numerical value;
under the condition that the Euclidean distance value is not smaller than a distance threshold value, determining that a first similarity between two communities corresponding to the Euclidean distance value is a second numerical value; and
and constructing the first sub-adjacency matrix based on a node identifier corresponding to a community, the first value and the second value, wherein the node identifier is a unique identifier corresponding to a node in a graph included in the preset energy consumption prediction model, and the node corresponds to the community.
Further, the determining a second similarity between each two communities based on the sample data corresponding to each two communities, respectively, further includes,
wherein the said
Characterizing node identities corresponding to said communities, respectively, said +.>
Characterizing said second similarity, said ++>
Characterizing sample matrices determined from the sample data corresponding to the communities, respectively, the
Characterizing a matrix similarity score, and said ++ >
The total number of the plurality of communities is characterized.
Further, the preset energy consumption prediction model comprises a preset graph neural network model and a preset self-attention model, the preset energy consumption prediction model is trained based on the sample data, the label and the adjacent matrix, and the trained energy consumption prediction model further comprises,
processing each sample data by using the preset graph neural network model to obtain a corresponding predicted energy consumption sequence;
processing the predicted energy consumption sequence by utilizing the preset self-attention model to obtain corresponding predicted energy consumption data; and
and training the preset graph neural network model and the preset self-attention model by utilizing the difference between the predicted energy consumption data and the label corresponding to the sample data to obtain a trained graph neural network predicted model and a trained self-attention model so as to form the trained energy consumption predicted model.
Further, the user energy usage information includes energy usage data including at least one of gas usage data, water usage data, and electricity usage data, and acquisition time information.
In another aspect, embodiments of the present disclosure also provide a method of predicting energy usage, comprising,
determining a corresponding target trained energy consumption prediction model according to the received community identification to be predicted; and
processing the energy consumption information set of the user to be predicted corresponding to the community identification to be predicted by utilizing the energy consumption prediction model after target training to obtain target energy consumption data corresponding to the community identification to be predicted,
the target trained energy consumption prediction model is obtained by training the preset energy consumption prediction model by adopting any one of the methods.
On the other hand, the embodiment of the specification also provides an energy consumption prediction model training device, which comprises,
the labeling unit is used for labeling the plurality of pieces of user energy consumption information acquired by each community in the plurality of communities, determining a label corresponding to sample data of each community, wherein the sample data is determined by the plurality of pieces of user energy consumption information;
a first determining unit, configured to determine an adjacency matrix according to a similarity between every two communities; and
The training unit is used for training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model.
In another aspect, embodiments of the present disclosure also provide an energy usage prediction apparatus, comprising,
the second determining unit is used for determining a corresponding target trained energy consumption prediction model according to the received community identification to be predicted; and
a processing unit for processing the energy consumption information set of the user to be predicted corresponding to the community identification to be predicted by using the energy consumption prediction model after target training to obtain target energy consumption data corresponding to the community identification to be predicted,
the target trained energy consumption prediction model is obtained by training the device aiming at the preset energy consumption prediction model.
In another aspect, embodiments of the present disclosure further provide a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the method described above when executing the computer program.
In another aspect, embodiments of the present disclosure also provide a computer-readable storage medium having stored thereon computer instructions that, when executed by a processor, perform the above-described method.
In another aspect, the present description embodiments also provide a computer program product comprising a computer program/instruction which, when executed by a processor, implements a method.
When the embodiment of the specification is utilized, when the trained energy consumption prediction model for energy consumption prediction is trained, labeling processing is carried out on a plurality of pieces of user energy consumption information collected by each community in a plurality of communities, and labels corresponding to sample data of each community are determined, wherein the sample data is determined by the plurality of pieces of user energy consumption information; determining an adjacency matrix according to the similarity between every two communities; and training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model. Therefore, based on the adjacency matrix in the graph neural network, the correlation between communities is considered in the process of energy consumption prediction, and the accuracy of the energy consumption prediction result is improved.
Drawings
In order to more clearly illustrate the embodiments of the present description or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present description, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of an implementation system of an energy consumption prediction model training method according to an embodiment of the present disclosure;
FIG. 2 is a flowchart of an energy consumption prediction model training method according to an embodiment of the present disclosure;
FIG. 3 is a flow chart illustrating a method of energy consumption prediction according to an embodiment of the present disclosure;
FIG. 4 is a flowchart of a labeling method according to an embodiment of the present disclosure;
FIG. 5A is a schematic diagram of an energy consumption prediction model training method according to another embodiment of the present disclosure;
FIG. 5B is a flowchart illustrating a method for determining an adjacency matrix according to an embodiment of the present disclosure;
FIG. 5C is a flowchart of a first sub-adjacency matrix determination method according to an embodiment of the present disclosure;
FIG. 5D is a schematic diagram of an energy consumption prediction model training method according to another embodiment of the present disclosure;
FIG. 5E is a graph showing a predicted energy consumption result according to an embodiment of the present disclosure;
FIG. 5F is a graph showing a predicted energy consumption according to another embodiment of the present disclosure;
FIG. 6A is a schematic diagram illustrating a training device for energy consumption prediction model according to an embodiment of the present disclosure;
FIG. 6B is a schematic diagram illustrating a structure of an energy consumption prediction apparatus according to an embodiment of the present disclosure;
fig. 7 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure.
[ reference numerals description ]
110. An electronic device;
120. a database;
130. the energy consumption prediction model after target training;
140. a user terminal;
150. target energy consumption data;
160. community information to be predicted;
501. a plurality of first user energy usage information;
502. a plurality of second user energy usage information;
503. sample data;
504. a label;
505. an adjacency matrix;
511. presetting a graph neural network model;
512. presetting a self-attention model;
513. a loss function;
521. predicting an energy consumption sequence;
522. predicting energy consumption data;
531. the trained graph neural network model;
532. A trained self-attention model;
611. a labeling unit;
612. a first determination unit;
613. a training unit;
621. a second determination unit;
622. a processing unit;
702. a computer device;
704. a processing device;
706. storing the resource;
708. a driving mechanism;
710. an input/output module;
712. an input device;
714. an output device;
716. a presentation device;
718. a graphical user interface;
720. a network interface;
722. a communication link;
724. a communication bus.
Detailed Description
The technical solutions of the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is apparent that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.
It should be noted that the terms "first," "second," and the like in the description and the claims of the specification and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the present description described herein may be capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or device that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or device.
In the technical scheme of the specification, the related user energy consumption information and the information corresponding to communities are collected, stored, used, processed, transmitted, provided, disclosed, applied and the like, which meet the requirements of related laws and regulations, and necessary security measures are adopted without violating the public order.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
Fig. 1 is a schematic diagram of an implementation system of an energy consumption prediction model training method according to an embodiment of the present disclosure, which may include an electronic device 110 and a database 120.
The electronic device 110 may access the database 120 through a network, for example. The database 120 may store user energy usage information for a plurality of communities.
In one embodiment, the electronic device 110 may read the user energy consumption information of a plurality of communities from the database 120, and perform a labeling process based on the user energy consumption information of the plurality of communities, so as to obtain labels corresponding to each community. And determining an adjacency matrix according to the similarity between every two communities. And training the preset energy consumption prediction model by taking the adjacency matrix, sample data determined by the user energy consumption information of the communities and the labels as samples to obtain the trained energy consumption prediction model. The preset energy consumption prediction model may include, for example, a preset graph neural network model and a preset self-attention model, where the preset graph neural network model is used to process sample data obtained from user energy consumption information of multiple communities and an adjacent matrix to obtain a predicted energy consumption sequence corresponding to each community; and processing the predicted energy consumption sequence by using a preset self-attention model to obtain predicted energy consumption data. The electronic device 110 may compare the obtained predicted energy usage data with the corresponding tag, and train the preset graph neural network model and the preset self-attention model according to the comparison result.
In an embodiment, the application scenario may further include the target trained energy usage prediction model 130, the user terminal 140, the target energy usage data 150, and the community information to be predicted 160, where the community information to be predicted may include, for example, a community identifier to be predicted, and the community information to be predicted may further include, for example, a set of user energy usage information to be predicted. The user terminal 140 is communicatively coupled to the electronic device 110 via a network. For example, the user terminal 140 may obtain the target trained energy usage prediction model 130 corresponding to the community identifier to be predicted from the electronic device 110, and process the set of user energy usage information to be predicted based on the obtained target trained energy usage prediction model 130, to obtain the target energy usage data 150 corresponding to the set of user energy usage information to be predicted.
The method for training the energy consumption prediction model provided in the present specification may be generally performed by the electronic device 110, or may be performed by a server or the like communicatively connected to the electronic device 110. The energy usage prediction model application method (energy usage prediction method) provided in the present specification may be executed by the user terminal 140 or the electronic device 110. Accordingly, the energy consumption prediction model training apparatus provided in the present specification may be generally provided in the electronic device 110, or may be provided in a server communicatively connected to the electronic device 110. The energy consumption prediction model application device (energy consumption prediction device) provided in the present specification may be provided in the user terminal 140 or the electronic device 110.
It should be understood that the number and types of electronic devices, databases, community information to be predicted, and user terminals in fig. 1 are merely illustrative. There may be any number and type of electronic devices, databases, community information to be predicted, and user terminals, as desired for implementation.
FIG. 2 is a flowchart of an energy consumption prediction model training method according to an embodiment of the present disclosure. The training process of the energy usage prediction model is described in this figure, but may include more or fewer operational steps based on routine or non-inventive labor. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When a system or apparatus product in practice is executed, it may be executed sequentially or in parallel according to the method shown in the embodiments or the drawings. As shown in fig. 2, the method may include:
s210, labeling the information of the energy consumption of a plurality of users collected by each community in the communities, and determining a label corresponding to sample data of each community;
s220, determining an adjacency matrix according to the similarity between every two communities;
and S230, training a preset energy consumption prediction model based on the sample data, the labels and the adjacency matrix to obtain a trained energy consumption prediction model.
When the embodiment of the specification is utilized, when the trained energy consumption prediction model for energy consumption prediction is trained, labeling processing is carried out on a plurality of pieces of user energy consumption information collected by each community in a plurality of communities, and labels corresponding to sample data of each community are determined, wherein the sample data is determined by the plurality of pieces of user energy consumption information; determining an adjacency matrix according to the similarity between every two communities; and training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model. Currently, the energy usage between communities with relevance (similar features) is also highly relevant. Therefore, the accuracy of energy consumption prediction can be improved by considering the relevance among communities in the process of energy consumption prediction. Therefore, based on the adjacency matrix in the graph neural network, the correlation between communities is considered in the process of energy consumption prediction, and the accuracy of the energy consumption prediction result is improved.
According to one embodiment of the present description, each community includes a plurality of households, each household being a user. Each user has corresponding energy consumption information every day. And respectively acquiring daily energy consumption information of each user aiming at each community, and summarizing the acquired user energy consumption into consumption sets which are respectively associated with corresponding community identifications.
And carrying out sampling processing on the consumption sets associated with each community identifier to obtain corresponding sample data. Specifically, for each community, at least one user identifier is selected randomly from a plurality of user identifiers included in the community. And determining the user energy consumption information corresponding to each user identification as candidate information from the consumption set for each day. And respectively preprocessing the candidate information to obtain sample data. The preprocessing may include, for example, data format normalization processing, data complement processing, outlier rejection processing, and the like.
For example, determining the corresponding label for the sample data of each community may be determining all energy usage data included in the usage set corresponding to the community identifier for each day, calculating a mean value of the energy usage data, and taking the mean value data as the label.
For example, community a includes three users x, y and z, and on day F, the user energy usage information corresponding to the three users x, y and z includes (a, first information), (b, second information) and (c, third information), and the user energy data corresponding to the three users x, y and z includes a, b and c. In the case where the randomly determined user identifications are y and z, the sample data corresponding to community a on day F are (b, second information) and (c, third information), and the labels are the average of a, b and c.
Determining the adjacency matrix according to the similarity between every two communities may be, for example, determining the similarity between every two communities based on the difference between labels corresponding to every two communities, and determining the adjacency matrix based on the similarity. Specifically, for each day, differences between labels corresponding to every two communities are respectively determined, the differences are summarized as target differences, and the similarity between every two communities is determined based on the average value of the target differences. Specifically, in the case that the target difference value is smaller than the gap threshold, the similarity between the two communities is determined to be a first value, otherwise, the similarity between the two communities is determined to be a second value, for example, the first value is 1, and the second value is a value smaller than 1, for example, may be 0. An adjacency matrix corresponding to the plurality of communities is constructed based on the node corresponding to each community, the first value and the second value.
It should be noted that the similarity between every two communities can be determined by the distance between communities. For example, the distance between each two communities is determined according to the position information of the communities. And comparing the distance with a distance threshold, and determining the similarity between the two communities corresponding to the distance to be a first value under the condition that the distance is smaller than the distance threshold, otherwise, determining the similarity between the two communities corresponding to the distance to be a second value. The embodiments of the present disclosure do not limit the similarity between two communities, but are merely similarity between two communities in one or more dimensions. The present embodiment merely defines determining the adjacency matrix based on the similarity.
After determining the sample data, the label and the adjacency matrix, training the preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model. Specifically, the sample data and the adjacent matrix are input into a preset energy consumption prediction model to obtain predicted energy consumption data. And performing difference processing on the predicted energy consumption data and the corresponding labels by using a loss function to obtain a loss function value, training a preset energy consumption prediction model based on the loss function value until the obtained loss function value or the training times meet preset conditions, and determining the last preset energy consumption prediction model as a trained energy consumption prediction model for energy consumption prediction.
According to another embodiment of the present specification, the user energy usage information includes energy usage data and acquisition time information, and the energy usage data includes at least one of gas usage data, water usage data, and electricity usage data.
The acquisition time is, for example, in units of one day.
Fig. 3 is a flowchart of a method for predicting energy consumption according to an embodiment of the present disclosure. In this figure a predictive process for energy usage is described, but more or fewer operational steps may be included based on routine or non-inventive labor. As shown in fig. 3, the method may include:
S310, determining a corresponding target trained energy consumption prediction model according to the received community identification to be predicted;
s320, processing the to-be-predicted user energy consumption information set corresponding to the to-be-predicted community identification by using the energy consumption prediction model after target training to obtain target energy consumption data corresponding to the to-be-predicted community identification.
By utilizing the embodiment of the specification, training is performed in advance on a preset energy consumption prediction model to obtain a trained energy consumption prediction model, and the trained energy consumption prediction model is associated with a sample community identifier corresponding to sample data for training the trained energy consumption prediction model. The specific method for training the preset energy consumption prediction model to obtain the trained energy consumption prediction model may be as shown in fig. 2.
And when the energy consumption is predicted for the community, receiving the community identification to be predicted, and determining a target sample community which is consistent and matched with the community identification to be predicted from a plurality of sample communities. And taking the trained energy consumption prediction model associated with the target sample community as a target trained energy consumption prediction model to predict. The trained energy prediction model includes a weighted graph network obtained after training.
When the user needs to predict the energy consumption aiming at the community to be predicted, the user energy consumption information set to be predicted can be input through the user terminal, and the user energy consumption information to be predicted can be, for example, the historical energy consumption information set of the community. When the user needs to forecast the energy consumption of the community to be forecasted, the forecast date can be input. The historical energy amount information set includes energy amount information corresponding to a plurality of dates.
And under the condition that the user does not input the energy consumption information set of the user to be predicted, acquiring the corresponding energy consumption information set of the user to be predicted from the historical consumption information base based on the community identification to be predicted.
And processing the to-be-predicted user energy consumption information set by using the target trained energy consumption prediction model to obtain target energy consumption data, and sending the target energy consumption data to a user terminal for the user to review.
Processing the to-be-predicted user energy consumption information set by using the target trained energy consumption prediction model to obtain target energy consumption data, for example, processing the to-be-predicted user energy consumption information set by using a target trained graph neural network model to obtain a target predicted energy consumption sequence; and processing the target predicted energy consumption sequence by using a preset self-attention model after target training to obtain the target energy consumption data.
Fig. 4 is a flowchart of a labeling processing method according to an embodiment of the present disclosure. The labeling process is depicted in this figure, but may include more or fewer operational steps based on conventional or non-inventive labor. As shown in fig. 4, the method may include:
s411, respectively determining a plurality of users included in each community;
s412, calculating an average value of the energy consumption data corresponding to each user;
s413, the average value is set as a label corresponding to the sample data.
According to another embodiment of the present description, for each community, the users included in the community are determined separately. User energy usage information corresponding to each user per day is determined. For a certain day, energy usage data corresponding to each user is acquired, and an average value of the plurality of energy usage data is determined. Meanwhile, for the day, sample data corresponding to the community is determined, and the average value is used as a label corresponding to the sample data.
For example, community a includes three users x, y and z, and on day 1, the user energy usage information corresponding to the three users x, y and z includes (a 1, first information), (b 1, second information) and (c 1, third information), and the user energy data corresponding to the three users x, y and z includes a1, b1 and c1; on day 2, the user energy usage information corresponding to the three users x, y and z includes (a 2, fourth information), (b 2, fifth information) and (c 2, sixth information), and the user energy data corresponding to the three users x, y and z includes a2, b2 and c2; on day 3, the user energy usage information corresponding to the three users of x, y and z includes (a 3, seventh information), (b 3, eighth information) and (c 3, ninth information), and the user energy data corresponding to the three users of x, y and z includes a3, b3 and c3. In the case that the randomly determined user identifications are y and z, sample data corresponding to the community a on day 1 are (b 1, second information) and (c 1, third information), and the labels are average values of a1, b1 and c1; sample data corresponding to community a on day 2 are (b 2, fifth information) and (c 2, sixth information), and the label is the average of a2, b2 and c2; sample data corresponding to community a on day 3 are (b 3, eighth information) and (c 3, ninth information), and the label is the average of a3, b3 and c3.
In determining the average value, the number of days used was related to the predicted frequency. For example, when prediction is required once every other week, it is necessary to determine that the average value of the energy usage data for one week is the corresponding label.
FIG. 5A is a schematic diagram of an energy consumption prediction model training method according to another embodiment of the present disclosure; FIG. 5B is a flowchart illustrating a method for determining an adjacency matrix according to an embodiment of the present disclosure; FIG. 5C is a flowchart of a first sub-adjacency matrix determination method according to an embodiment of the present disclosure; FIG. 5D is a schematic diagram of an energy consumption prediction model training method according to another embodiment of the present disclosure; FIG. 5E is a graph showing a predicted energy consumption result according to an embodiment of the present disclosure; fig. 5F is a graph showing a predicted energy consumption result according to another embodiment of the present disclosure. The determination of the adjacency matrix and the first sub-adjacency matrix is described in this fig. 5B and 5C, but may include more or fewer operational steps based on conventional or non-inventive labor.
According to another embodiment of the present disclosure, the preset energy usage prediction model includes a preset graph neural network model and a preset self-attention model, training the preset energy usage prediction model based on sample data, a label and an adjacency matrix, and obtaining the trained energy usage prediction model includes: processing each sample data by using a preset graph neural network model to obtain a corresponding predicted energy consumption sequence; processing the predicted energy consumption sequence by using a preset self-attention model to obtain corresponding predicted energy consumption data; and training the preset graph neural network model and the preset self-attention model by utilizing the difference between the predicted energy consumption data and the labels corresponding to the sample data to obtain a trained graph neural network predicted model and a trained self-attention model so as to form a trained energy consumption predicted model.
As shown in fig. 5A, for example, a plurality of first user energy usage information 501 is collected from community a, and a plurality of second user energy usage information 502 is collected from community B. Based on the plurality of first user energy usage information 501 and the plurality of second user energy usage information 502, sample data 503 and a tag 504 corresponding to each sample data 503 are determined. Specifically, a method for determining the sample data 503 and the label 504 corresponding to each sample data 503 based on the plurality of first user energy usage information 501 and the plurality of second user energy usage information 502 is shown in fig. 4, and is not described herein.
Based on sample data determined from a plurality of user energy usage information corresponding to a plurality of communities, an adjacency matrix 505 is determined. Specifically, based on sample data corresponding to each two communities, a second similarity between each two communities is determined, respectively, to obtain the adjacency matrix 505. Determining the second similarity between each two communities based on the sample data corresponding to each two communities may be shown in the following equations (1) and (2), for example.
The adjacency matrix 505 and the sample data 503 are input into a pre-map neural network model 511 to obtain a predicted energy usage sequence 521. The predicted energy usage sequence 521 is processed using a predetermined self-attention model 512 (transducer model) to obtain predicted energy usage data 522.
The predicted energy usage data 522 and corresponding labels 504 are processed using the loss function 513 to obtain loss function values. Training is performed on the preset graph neural network model 511 and the preset self-attention model 512 based on the loss function value, and a trained graph neural network model 531 and a trained self-attention model 532 which satisfy preset conditions are obtained for energy consumption data prediction.
The loss function 513 can be shown, for example, as in the following equations (3) and (4).
Wherein,,
characterization of the loss function value->
Characterization of total duration (total days) for training,>
for predicting frequency (in days), +.>
The number of steps of the output is characterized, and (2) >
Characterization of the total number of days (e.g., G days of user energy usage information is collected, T is G) corresponding to the sample data>
Characterizing a second step size>
Characterization of preset graph neural network model 511 and preset self-attention model 512,/for example>
Characterization of predicted energy usage data,/->
Characterizing the start date,/->
Characterized in->
Predicted energy consumption data output by the step model, < +.>
Characterized in->
Step (3) labels.
Specifically, the processing procedure of the sample data by the preset map neural network model 511 and the preset self-attention model 512 is shown in fig. 5D. As shown in the first layer, user energy consumption information is collected for a plurality of users included in each community. Sample data and tags are generated. And determining that nodes in the graph neural network correspond to community identifications, namely each community identification has a node corresponding to the community identification, and taking sample data and labels obtained from the user energy consumption information of the community as sample data and labels corresponding to the nodes. Based on the similarity between every two communities, a corresponding adjacency matrix is determined. The sample data and the adjacency matrix are then input into a predetermined graph neural network model (GRU) to obtain a predicted energy usage sequence, and the predicted energy usage sequence is input into a self-attention model (comprising a self-attention layer, such as a transducer layer) to obtain predicted energy usage data.
Specifically, the preset map neural network model includes an update formula such as the following formula (5).
After determining the output characteristics of each node identifier at each time node, inputting the output characteristics into the preset graph neural network model again, wherein each iteration of the preset graph neural network model needs the hidden state of the last step
Simultaneously receiving input sample data of the current time step +.>
Then, a new predicted energy consumption containing history information is outputted, as shown in the following formulas (6) and (7).
Due to the fact that graphs and adjacent matrixes (representing the association degree among communities) corresponding to the graphs in the graph neural network model, the association among communities is considered in the process of energy consumption prediction. Therefore, the energy consumption is predicted based on the trained graph neural network model, and the accuracy of the energy consumption prediction result is improved. The graph is, for example, a topological graph, the topological graph comprises a plurality of nodes and connection relations between every two nodes, each node is associated with a corresponding node identifier, the node identifier is a unique identifier corresponding to a node in the graph included in a preset energy consumption prediction model, and the node corresponds to a community. The adjacency matrix characterizes the connection relationship between every two nodes.
It should be noted that fig. 5A only shows that user energy usage information is collected from two communities to obtain sample data, which is only used for exemplary purposes and is not limited to the present description. In practical applications, the number of communities need only be greater than or equal to two.
As shown in fig. 5B, the method for determining the adjacency matrix according to the similarity between every two communities may include:
s521, position information of each community is respectively determined;
s522, based on the position information, determining a first similarity between every two communities respectively to obtain a first sub-adjacency matrix;
s523, respectively determining a second similarity between every two communities based on sample data corresponding to every two communities to obtain a second sub-adjacency matrix;
s524, determining an adjacency matrix based on the first sub-adjacency matrix and the second sub-adjacency matrix.
According to another embodiment of the present disclosure, the location information of each community may be, for example, longitude and latitude of the community, which may identify the location of the community.
Based on the obtained position information, a first similarity between every two communities is determined, and based on the value of the first similarity, a first sub-adjacency matrix is determined. The first sub-adjacency matrix corresponds to nodes of a graph in the graph neural network. For example, if the first similarity between the first community and the second community is 1 and the first similarity between the first community and the third community is 0, in the first sub-adjacency matrix, the corresponding value is 1 with the row where the node of the first community is located and the column where the node of the second community is located; and the corresponding numerical value is 0 with the row where the node of the first community is located and the column where the node of the third community is located.
The specific determination of the second similarity between each two communities is determined by the above formula (1) and formula (2), respectively, based on sample data corresponding to each two communities. Based on the second similarity, a corresponding second sub-adjacency matrix is determined, the second sub-adjacency matrix corresponding to nodes of the graph in the graph neural network.
The determining the adjacency matrix based on the first sub-adjacency matrix and the second sub-adjacency matrix may specifically be determining a hadamard product of the first sub-adjacency matrix and the second sub-adjacency matrix as the adjacency matrix. Each data in the adjacency matrix is a product of a first adjacency value and a second adjacency value at corresponding positions in the first sub-adjacency matrix and the second sub-adjacency matrix.
Determining the adjacency matrix may also include, for example, the following equation (8).
As shown in fig. 5C, the method for determining the first similarity between every two communities based on the location information may include:
S5221, determining Euclidean distance values based on the position information of every two communities;
s5222, judging whether the Euclidean distance value is smaller than a distance threshold value;
s5223, determining a first similarity between two communities corresponding to the Euclidean distance value as a first numerical value under the condition that the Euclidean distance value is smaller than a distance threshold value;
s5224, under the condition that the Euclidean distance value is not smaller than the distance threshold value, determining that the first similarity between two communities corresponding to the Euclidean distance value is a second numerical value;
s5225, constructing a first sub-adjacency matrix based on the node identification corresponding to the community, the first value and the second value.
According to another embodiment of the present specification, euclidean distance values between two communities are determined based on the location information of each two communities, respectively. And comparing each Euclidean distance value with a distance threshold value, and judging whether the Euclidean distance is smaller than the distance threshold value.
And under the condition that the Euclidean distance value is smaller than the distance threshold value, determining the first similarity between two communities corresponding to the Euclidean distance value as a first numerical value. And under the condition that the Euclidean distance value is not smaller than the distance threshold value, determining that the second similarity between two communities corresponding to the Euclidean distance value is a second numerical value.
Based on the node identifier, the first value and the second value corresponding to the community, a specific method for constructing the first sub-adjacency matrix is shown in fig. 5B, which is not described herein.
It should be noted that, based on the method of the embodiment of the present disclosure, the set of to-be-predicted user energy consumption information corresponding to the to-be-predicted community identifier is processed to obtain corresponding target energy consumption data, and the target energy consumption data is compared with the actual value, as shown in fig. 5E and 5F, where the abscissa in fig. 5E and 5F is the prediction time, and the ordinate is the target energy consumption data (target gas consumption data) corresponding to each prediction time. As can be seen from fig. 5E and fig. 5F, the method of the embodiment of the present disclosure has higher accuracy in predicting the fuel gas consumption.
Fig. 6A is a schematic structural diagram of an energy consumption prediction model training device according to an embodiment of the present disclosure. As shown in fig. 6A, including,
a labeling unit 611, configured to perform labeling processing on a plurality of pieces of user energy usage information collected by each of a plurality of communities, determine a label corresponding to sample data of each community, where the sample data is determined by the plurality of pieces of user energy usage information;
A first determining unit 612, configured to determine an adjacency matrix according to the similarity between every two communities;
the training unit 613 is configured to train the preset energy usage prediction model based on the sample data, the label and the adjacency matrix, and obtain a trained energy usage prediction model.
Since the principle of the device for solving the problem is similar to that of the method, the implementation of the device can be referred to the implementation of the method, and the repetition is omitted.
Fig. 6B is a schematic structural diagram of an energy consumption prediction device according to an embodiment of the present disclosure. As shown in fig. 6B, including,
a second determining unit 621, configured to determine a corresponding target trained energy consumption prediction model according to the received community identifier to be predicted; and
the processing unit 622 is configured to process the to-be-predicted user energy consumption information set corresponding to the to-be-predicted community identifier by using the energy consumption prediction model after the target training, so as to obtain target energy consumption data corresponding to the to-be-predicted community identifier.
It should be noted that the energy consumption prediction model after the target training is obtained by training the device shown in fig. 6A for the preset energy consumption prediction model.
Since the principle of the device for solving the problem is similar to that of the method, the implementation of the device can be referred to the implementation of the method, and the repetition is omitted.
Fig. 7 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure, where an apparatus in the present disclosure may be the computer device in the present embodiment, and perform the method of the present disclosure. The computer device 702 may include one or more processing devices 704, such as one or more Central Processing Units (CPUs), each of which may implement one or more hardware threads. The computer device 702 may also include any storage resources 706 for storing any kind of information, such as code, settings, data, etc. For example, and without limitation, storage resources 706 may include any one or more of the following combinations: any type of RAM, any type of ROM, flash memory devices, hard disks, optical disks, etc. More generally, any storage resource may store information using any technology. Further, any storage resource may provide volatile or non-volatile retention of information. Further, any storage resources may represent fixed or removable components of computer device 702. In one case, the computer device 702 can perform any of the operations of the associated instructions when the processing device 704 executes the associated instructions stored in any storage resource or combination of storage resources. The computer device 702 also includes one or more drive mechanisms 708, such as a hard disk drive mechanism, an optical disk drive mechanism, and the like, for interacting with any storage resources.
The computer device 702 may also include an input/output module 710 (I/O) for receiving various inputs (via an input device 712) and for providing various outputs (via an output device 714). One particular output mechanism may include a presentation device 716 and an associated Graphical User Interface (GUI) 718. In other embodiments, input/output module 710 (I/O), input device 712, and output device 714 may not be included as just one computer device in a network. The computer device 702 can also include one or more network interfaces 720 for exchanging data with other devices via one or more communication links 722. One or more communication buses 724 couple the above-described components together.
Communication link 722 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the internet), a point-to-point connection, etc., or any combination thereof. Communication link 722 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.
The embodiments of the present specification also provide a computer readable storage medium storing a computer program which, when executed by a processor, implements the above method.
The present description also provides a computer program product comprising a computer program which, when executed by a processor, implements the above method.
It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present description is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the specification. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing detailed description of the embodiments has been presented for purposes of illustration and description, and it should be understood that the foregoing is by way of example only, and is not intended to limit the scope of the invention.
Claims (13)
1. The energy consumption prediction model training method is characterized by comprising the following steps of:
labeling a plurality of pieces of user energy consumption information acquired by each community in a plurality of communities, and determining a label corresponding to sample data of each community, wherein the sample data is determined by the plurality of pieces of user energy consumption information;
determining an adjacency matrix according to the similarity between every two communities; and
and training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model.
2. The method of claim 1, wherein the tagging the plurality of user energy usage information collected for each of the plurality of communities, and determining the tag corresponding to the sample data for each of the communities comprises:
for each community, determining a plurality of users included in the community respectively;
calculating an average value of energy consumption data corresponding to each user, wherein the user energy consumption information comprises the energy consumption data; and
and taking the average value as the label corresponding to the sample data.
3. The method of claim 1, wherein determining an adjacency matrix based on the similarity between each two communities comprises:
determining the position information of each community respectively;
based on the position information, respectively determining first similarity between every two communities to obtain a first sub-adjacency matrix;
based on the sample data corresponding to each two communities, respectively determining a second similarity between each two communities to obtain a second sub-adjacency matrix; and
the adjacency matrix is determined based on the first sub-adjacency matrix and the second sub-adjacency matrix.
4. The method of claim 3, wherein determining a first similarity between each two communities based on the location information to obtain a first sub-adjacency matrix comprises:
based on the position information of every two communities, respectively determining Euclidean distance values;
judging whether the Euclidean distance value is smaller than a distance threshold value or not;
under the condition that the Euclidean distance value is smaller than a distance threshold value, determining that a first similarity between two communities corresponding to the Euclidean distance value is a first numerical value;
Under the condition that the Euclidean distance value is not smaller than a distance threshold value, determining that a first similarity between two communities corresponding to the Euclidean distance value is a second numerical value; and
and constructing the first sub-adjacency matrix based on a node identifier corresponding to a community, the first value and the second value, wherein the node identifier is a unique identifier corresponding to a node in a graph included in the preset energy consumption prediction model, and the node corresponds to the community.
5. A method according to claim 3, wherein said determining a second similarity between each two of said communities based on said sample data corresponding to each two of said communities, respectively, comprises:
6. The method of claim 1, wherein the pre-set energy usage prediction model comprises a pre-set graph neural network model and a pre-set self-attention model, wherein training the pre-set energy usage prediction model based on the sample data, the label, and the adjacency matrix, the obtaining the trained energy usage prediction model comprises:
Processing each sample data by using the preset graph neural network model to obtain a corresponding predicted energy consumption sequence;
processing the predicted energy consumption sequence by utilizing the preset self-attention model to obtain corresponding predicted energy consumption data; and
and training the preset graph neural network model and the preset self-attention model by utilizing the difference between the predicted energy consumption data and the label corresponding to the sample data to obtain a trained graph neural network predicted model and a trained self-attention model so as to form the trained energy consumption predicted model.
7. The method of claim 1, wherein the user energy usage information comprises energy usage data and acquisition time information, the energy usage data comprising at least one of gas usage data, water usage data, and electrical usage data.
8. An energy usage prediction method, comprising:
determining a corresponding target trained energy consumption prediction model according to the received community identification to be predicted; and
processing the energy consumption information set of the user to be predicted corresponding to the community identification to be predicted by utilizing the energy consumption prediction model after target training to obtain target energy consumption data corresponding to the community identification to be predicted,
The target trained energy consumption prediction model is obtained by training the preset energy consumption prediction model by adopting the method of any one of claims 1-6.
9. An energy usage prediction model training device, comprising:
the labeling unit is used for labeling the plurality of pieces of user energy consumption information acquired by each community in the plurality of communities, determining a label corresponding to sample data of each community, wherein the sample data is determined by the plurality of pieces of user energy consumption information;
a first determining unit, configured to determine an adjacency matrix according to a similarity between every two communities; and
the training unit is used for training a preset energy consumption prediction model based on the sample data, the label and the adjacency matrix to obtain a trained energy consumption prediction model.
10. An energy consumption prediction apparatus, comprising:
the second determining unit is used for determining a corresponding target trained energy consumption prediction model according to the received community identification to be predicted; and
a processing unit for processing the energy consumption information set of the user to be predicted corresponding to the community identification to be predicted by using the energy consumption prediction model after target training to obtain target energy consumption data corresponding to the community identification to be predicted,
The target trained energy consumption prediction model is obtained by training the device according to claim 9 for the preset energy consumption prediction model.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any of the preceding claims 1-8 when executing the computer program.
12. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, performs the method of any of the preceding claims 1-8.
13. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the method according to any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310379038.3A CN116109016A (en) | 2023-04-11 | 2023-04-11 | Energy consumption prediction model training and energy consumption prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310379038.3A CN116109016A (en) | 2023-04-11 | 2023-04-11 | Energy consumption prediction model training and energy consumption prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116109016A true CN116109016A (en) | 2023-05-12 |
Family
ID=86256473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310379038.3A Pending CN116109016A (en) | 2023-04-11 | 2023-04-11 | Energy consumption prediction model training and energy consumption prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116109016A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105335800A (en) * | 2015-11-19 | 2016-02-17 | 国网天津市电力公司 | Method for forecasting electricity consumption of power consumers based on joint learning |
| US20170060109A1 (en) * | 2013-06-06 | 2017-03-02 | International Business Machines Corporation | Managing time-substitutable electricity usage using dynamic controls |
| CN115049169A (en) * | 2022-08-16 | 2022-09-13 | 国网湖北省电力有限公司信息通信公司 | Regional power consumption prediction method, system and medium based on combination of frequency domain and spatial domain |
-
2023
- 2023-04-11 CN CN202310379038.3A patent/CN116109016A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170060109A1 (en) * | 2013-06-06 | 2017-03-02 | International Business Machines Corporation | Managing time-substitutable electricity usage using dynamic controls |
| CN105335800A (en) * | 2015-11-19 | 2016-02-17 | 国网天津市电力公司 | Method for forecasting electricity consumption of power consumers based on joint learning |
| CN115049169A (en) * | 2022-08-16 | 2022-09-13 | 国网湖北省电力有限公司信息通信公司 | Regional power consumption prediction method, system and medium based on combination of frequency domain and spatial domain |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Verenich et al. | Survey and cross-benchmark comparison of remaining time prediction methods in business process monitoring | |
| US20220131770A1 (en) | System and method for predicting and reducing subscriber churn | |
| CN113660225A (en) | Network attack event prediction method, system, device and medium based on time sequence point | |
| CN117041017B (en) | Intelligent operation and maintenance management method and system for data center | |
| CN110647456A (en) | Fault prediction method, system and related device of storage equipment | |
| CN108683658B (en) | Industrial control network flow abnormity identification method based on multi-RBM network construction reference model | |
| CN111797942A (en) | User information classification method and device, computer equipment and storage medium | |
| CN115859191A (en) | Fault diagnosis method and device, computer readable storage medium and computer equipment | |
| CN117113159A (en) | Deep learning-based power consumer side load classification method and system | |
| CN114238764A (en) | Course recommendation method, device and equipment based on recurrent neural network | |
| CN114266288A (en) | Network element detection method and related device | |
| CN111325255B (en) | Specific crowd delineating method and device, electronic equipment and storage medium | |
| CN112232985A (en) | Power distribution and utilization data monitoring method and device for ubiquitous power Internet of things | |
| CN112115994A (en) | Training method and device of image recognition model, server and storage medium | |
| CN116109016A (en) | Energy consumption prediction model training and energy consumption prediction method and device | |
| CN116484016A (en) | Time sequence knowledge graph reasoning method and system based on automatic maintenance of time sequence path | |
| CN113239272B (en) | Intention prediction method and intention prediction device of network management and control system | |
| CN110309948A (en) | Complete vehicle logistics order forecast method and device, logistics system and computer-readable medium | |
| CN113285977B (en) | Network maintenance method and system based on block chain and big data | |
| JP2008187612A (en) | Method and device for structuring traffic analytic model, structuring program, and storage medium thereof | |
| CN113313615A (en) | Method and device for quantitatively grading and grading enterprise judicial risks | |
| CN113139332A (en) | Automatic model construction method, device and equipment | |
| CN114417817B (en) | Session information cutting method and device | |
| CN116975300B (en) | Information mining method and system based on big data set | |
| CN115907770B (en) | Ethernet phishing fraud identification and early warning method based on time sequence feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230512 |