CN113780679B - Load prediction method and device based on ubiquitous power Internet of things - Google Patents
Load prediction method and device based on ubiquitous power Internet of things Download PDFInfo
- Publication number
- CN113780679B CN113780679B CN202111138775.1A CN202111138775A CN113780679B CN 113780679 B CN113780679 B CN 113780679B CN 202111138775 A CN202111138775 A CN 202111138775A CN 113780679 B CN113780679 B CN 113780679B
- Authority
- CN
- China
- Prior art keywords
- data
- long
- production element
- load prediction
- short
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000004519 manufacturing process Methods 0.000 claims abstract description 166
- 230000015654 memory Effects 0.000 claims abstract description 53
- 238000000605 extraction Methods 0.000 claims abstract description 51
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 239000000284 extract Substances 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 21
- 230000004913 activation Effects 0.000 claims description 20
- 238000006243 chemical reaction Methods 0.000 claims description 20
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 230000006399 behavior Effects 0.000 claims description 8
- 230000005611 electricity Effects 0.000 claims description 8
- 230000007787 long-term memory Effects 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 230000006403 short-term memory Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000006855 networking Effects 0.000 claims 1
- 230000007246 mechanism Effects 0.000 abstract description 10
- 235000008694 Humulus lupulus Nutrition 0.000 abstract 1
- 230000008569 process Effects 0.000 description 9
- 238000012549 training Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000010219 correlation analysis Methods 0.000 description 2
- 230000008034 disappearance Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000007087 memory ability Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000005612 types of electricity Effects 0.000 description 1
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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Primary Health Care (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Water Supply & Treatment (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a load prediction method and device based on ubiquitous power internet of things, wherein the method comprises the following steps: respectively acquiring data sets of a target user and other users in the ubiquitous power Internet of things range, and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data; constructing a feature extraction model based on a depth residual error network, wherein the feature extraction model extracts associated features between a first data set and electric data of a target user; and constructing a load prediction model based on the long-short-term memory network, and performing load prediction according to a second data set by the load prediction model to obtain load data of the target user. According to the load prediction method and the load prediction device, the uncertainty of load prediction can be reduced through the load prediction model based on the attention mechanism and the long-short-period memory network connected by long-short hops, and the load prediction accuracy of a target user can be improved.
Description
Technical Field
The invention relates to the technical field of power systems, in particular to a load prediction method and device based on ubiquitous power internet of things.
Background
The 5G technology provides technical guarantee for massive, high-frequency, low-time-delay and low-energy-consumption communication of data in the background of ubiquitous electric power Internet of things, and is based on the fact that a data acquisition system and a data communication system are gradually installed for each device, each user and even each electric equipment in each user of a power grid in the present stage. Meanwhile, in recent years, blowout type development of big data technology, deep learning method, optimization control technology, cloud computing, edge computing and other technologies also provides technical support for data mining.
When load prediction is performed on electric data of a target user, some of the prior art schemes utilize mathematical modeling or artificial intelligence technology to improve the accuracy of load prediction, and some schemes improve the accuracy of load prediction by introducing new influencing factors, such as common data types of electricity price, weather, temperature, humidity and the like.
In the load prediction model in the prior art, due to the fact that the dimension of the considered influencing factors is low, input information is incomplete, complete data features are difficult to cover, high uncertainty exists in the solution, the accuracy of the predicted power grid load is insufficient, and the operation management and the scheduling planning of the power grid are not facilitated.
Disclosure of Invention
The invention aims to provide a load prediction method and device based on ubiquitous power Internet of things, which are used for solving the technical problem that in the prior art, the accuracy of a load prediction result of a target user is insufficient.
The aim of the invention can be achieved by the following technical scheme:
A load prediction method based on ubiquitous power Internet of things comprises the following steps:
Respectively acquiring data sets of a target user and other users in the ubiquitous power Internet of things range, and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
Constructing a feature extraction model based on a depth residual error network, wherein the feature extraction model extracts associated features between a first data set and electric data of a target user; wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
And constructing a load prediction model based on a long-short-term memory network, wherein the load prediction model predicts the load according to a second data set to obtain the load data of the target user, the second data set comprises the first data set, the association characteristic, the residual production element data and the residual production element category, and the residual production element data and the residual production element category are determined according to the production element data.
Optionally, before constructing the load prediction model based on the long-term and short-term memory network, the method further comprises:
And calculating the similarity among all production elements according to the production element data, removing the production elements with the similarity exceeding a preset threshold, taking the data of other production elements as the residual production element data, and clustering the residual production element data to obtain the residual production element category.
Optionally, calculating the similarity between all production elements according to the production element data includes:
Calculating the similarity between the two production elements by using a formula ;
Wherein I (X, Y) is the similarity between the production elements X, Y, X, Y is the production element, X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
Optionally, the inter-user association data includes:
network topology association relation, user service association degree and electricity consumption behavior similarity;
The user service association degree represents the association degree of business between users, and the electricity consumption behavior similarity is obtained by clustering the electrical data between the users.
Optionally, the feature extraction model extracting the correlation feature between the first data set and the electrical data of the target user includes:
Arranging data in the first data set into a multi-channel matrix form, extracting shallow layer association features through a convolution layer and an activation function layer, extracting deep layer association features through a plurality of residual blocks, and carrying out weighted summation on the shallow layer association features and the deep layer association features through a first global long jump connection to obtain association features; the first global long jump connection spans several residual blocks and one convolutional layer.
Optionally, the load prediction model performs load prediction according to a second data set to obtain load data of the target user, which includes:
And arranging the data in the second data set into a multi-channel matrix form, performing nonlinear conversion through a convolution layer and an activation function layer to obtain a nonlinear conversion result, performing characteristic learning and transmission through a plurality of long-short-period memory network layers, and outputting load data of the target user through a batch normalization layer, the activation function layer and a full connection layer.
Optionally, the learning and transferring of the features through the plurality of long-short term memory network layers includes:
the long-period memory network basic unit group reserves learned knowledge or learns new knowledge through local short jump connection;
The long-short-period memory network layer reserves learned knowledge or learns new knowledge through local long-jump connection;
Superposing the nonlinear conversion result and the learning result of the long-period memory network layer through a second global long-jump connection;
The local short jump connection spans a group of long-short-term memory network basic unit groups, the local long jump connection spans one long-short-term memory network layer, the second global long jump connection spans a plurality of long-short-term memory network layers, and the long-short-term memory network layer comprises a plurality of long-short-term memory network basic unit groups.
The invention also provides a load prediction device based on the ubiquitous power internet of things, which comprises:
The data set acquisition module is used for respectively acquiring data sets of a target user and other users in the ubiquitous power internet of things range and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
the feature extraction module is used for constructing a feature extraction model based on a depth residual error network, and the feature extraction model is used for extracting association features between the first data set and the electric data of the target user; wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
The load prediction module is used for constructing a load prediction model based on a long-term and short-term memory network, the load prediction model carries out load prediction according to a second data set to obtain load data of the target user, the second data set comprises the first data set, the association characteristic, residual production element data and residual production element types, and the residual production element data and the residual production element types are determined according to the production element data.
Optionally, the method further comprises:
And the production element data processing module is used for calculating the similarity among all production elements according to the production element data, removing the production elements with the similarity exceeding a preset threshold, taking the data of other production elements as the residual production element data, and clustering the residual production element data to obtain the residual production element category.
Optionally, the production element data processing module calculates the similarity between all production elements according to the production element data, including:
calculating the similarity between the two production elements by using a formula ;
Wherein I (X, Y) is the similarity between the production elements X, Y, X, Y is the production element, X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
The invention provides a load prediction method and device based on ubiquitous power internet of things, wherein the method comprises the following steps: respectively acquiring data sets of a target user and other users in the ubiquitous power Internet of things range, and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data; constructing a feature extraction model based on a depth residual error network, wherein the feature extraction model extracts associated features between a first data set and electric data of a target user; wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data; and constructing a load prediction model based on a long-short-term memory network, wherein the load prediction model predicts the load according to a second data set to obtain the load data of the target user, the second data set comprises the first data set, the association characteristic, the residual production element data and the residual production element category, and the residual production element data and the residual production element category are determined according to the production element data.
In view of the above, the technical scheme of the invention has the following beneficial effects:
(1) According to the load prediction method, load prediction is carried out on the target user according to various data of a plurality of users in the range of the ubiquitous Internet of things, potential characteristics among various data in the local Internet of things are fully excavated by utilizing high-dimensional characteristics, and uncertainty of load prediction can be reduced;
(2) The influence of other users on the target user is fully mined through the inter-user association data, and a feature extraction model is constructed to extract association features between the first data set and the target user electrical data, so that the load prediction accuracy of the target user can be improved;
(3) The influence of the production elements on the load is fully excavated through the production element data in the user, and the classification of the production element data can improve the knowledge extraction of the production element data, thereby being beneficial to improving the efficiency of load prediction;
(4) By the load prediction model based on the long-short-term memory network, knowledge learning and memory capacity of data in different time periods can be improved, and load prediction can be performed on a target user more accurately.
Drawings
FIG. 1 is a flow chart of a load prediction method of the present invention;
FIG. 2 is a schematic structural diagram of a feature extraction model based on a depth residual network according to the present invention;
FIG. 3 is a schematic diagram of a load prediction model based on a long-short term memory network according to the present invention;
FIG. 4 is a schematic diagram of the LSTM basic cell of the invention;
fig. 5 is a schematic diagram of the load predicting apparatus according to the present invention.
Detailed Description
The embodiment of the invention provides a load prediction method and device based on ubiquitous power Internet of things, which are used for solving the technical problem that in the prior art, the accuracy of a load prediction result of a target user is insufficient.
In order that the invention may be readily understood, a more complete description of the invention will be rendered by reference to the appended drawings. Preferred embodiments of the present invention are shown in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Referring to fig. 1, an embodiment of a load prediction method based on ubiquitous power internet of things of the present invention includes:
S1: respectively acquiring data sets of a target user and other users in the ubiquitous power Internet of things range, and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
S2: constructing a feature extraction model based on a depth residual error network, wherein the feature extraction model extracts associated features between a first data set and electric data of a target user; wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
s3: and constructing a load prediction model based on a long-short-term memory network, wherein the load prediction model predicts the load according to a second data set to obtain the load data of the target user, the second data set comprises the first data set, the association characteristic, the residual production element data and the residual production element category, and the residual production element data and the residual production element category are determined according to the production element data.
In this embodiment, there are a plurality of range dividing methods of the ubiquitous power internet of things, which mainly include the following five methods: the same geographic location, the same administrative plate, the same business district, the same attribute building group, the same platform district.
In step S1, users in the ubiquitous power internet of things range are divided into target users and other users, the target users are targets for load prediction, data sets of the target users and other users in the ubiquitous power internet of things range are respectively acquired, the data sets comprise electric data, inter-user related data, production element data and load related data in the users, and the acquired data sets are preprocessed.
Wherein, the electrical data mainly includes: active power, reactive power, electrical energy, frequency, voltage, current, harmonics, open-close state, etc.; the inter-user association data mainly comprises: network topology association relationship, user service association degree and electricity consumption behavior similarity; the user internal production factor data includes: the system comprises a user investment condition, a user profit and loss condition, a user internal object acquisition condition, a user internal object use condition, a user internal object transfer condition, a user internal human resource allocation condition, a user main service completion condition, power utilization conditions of different electric appliances in the user and the like; the load-related data refers to other data related to the load of the user, and mainly includes: weather, temperature, humidity, electricity price, working day, holiday, etc.
The user service association degree represents the association degree of business between users. For example, when a service relationship exists between users A, B, the number of times of service passing between users A, B and the importance degree of each time of service passing are weighted and summed to obtain the quasi-association degree C B of the user B to the user A; the same method can obtain the quasi-association degree of other users except the user B to the user A, and further obtain the sum C sum of the quasi-association degree of other users to the user A; and calculating the ratio of the quasi-association degree C B and the sum of the quasi-association degrees C sum, and taking the ratio as the service association degree of the user B to the user A.
The electricity consumption behavior similarity is obtained by analyzing the electrical data of the user. Firstly, obtaining a clustering center curve of the electrical data of the users by a clustering method, then calculating the distance of the clustering center curve of the users, and measuring the similarity degree of the user behaviors by using the distance, wherein the smaller the distance is, the more similar the electricity utilization behaviors among the users are.
Specifically, in this embodiment, a K-means clustering method is used, firstly, the number of clusters K 1 is given, then K data are randomly selected as cluster centers, the remaining data are distributed to the cluster centers closest to each other, and each cluster center represents a cluster; each time a piece of data is allocated, the clustering center recalculates the clustering center according to the existing data in the corresponding clusters, and the calculated target is to enable the square sum of the distances from other data in the clusters to the clustering center to be minimum; the convergence condition of the k-means method is that the cluster center is no longer changed, i.e. the sum of squares of the distances of the individual data to the corresponding cluster center is minimal. In this embodiment, the cosine distance may be used to calculate the distances of the cluster center curves of different users.
In this embodiment, since the types of production elements in the power industry are multiple and the data amount is large, the similarity between all the production elements is calculated according to the production element data, the production elements with the similarity exceeding the preset threshold are removed, the data of other production elements are used as the remaining production element data, and the remaining production element data are clustered to obtain the remaining production element type.
Specifically, the formula is first used: Calculating the similarity between two production elements;
wherein X, Y is the similarity between the production elements X, Y, I (X, Y), X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
Then, carrying out correlation analysis by using a mutual information method, and removing production elements with higher similarity to reduce information redundancy; because linear or nonlinear correlation exists among production elements in the industry, the mutual information method can capture any type of correlation, so that the embodiment utilizes the mutual information method to perform correlation analysis on the production elements in the industry, removes the production elements with similarity exceeding a preset threshold, retains data of the remaining production elements, and utilizes a clustering algorithm to cluster the data of the remaining production elements to obtain the categories of the remaining production elements.
It should be noted that, in the step S3, the K-means method may be used to perform clustering, the number of clusters is set to K 2, and the clustering process is the same as that of the K-means method in the step S1, which is not described herein.
Referring to fig. 2, in step S2, a feature extraction model based on a depth residual network is constructed, a first data set is used as input data of the feature extraction model, and electric data of a target user is used as output data of the feature extraction model, so as to train the feature extraction model. Wherein the first data set includes electrical data of other users, inter-user association data, and load-related data.
Specifically, the input data of the feature extraction model is arranged in a multi-channel matrix form according to types, each channel corresponds to one type of data, and the input data is rearranged in a matrix form according to the sequence of the preceding columns and the following columns in a time sequence form. It should be noted that, the type refers to data of a further layer spanned by the input data, for example, the input data is "load related data", and then the type specifically refers to weather, temperature, humidity, and the like in the load related data.
In this embodiment, when the feature extraction model is trained, shallow associated features are extracted through a convolution layer and an activation function layer of the depth residual network, and in a preferred embodiment, the number of layers of the convolution layer and the activation function layer is 1-3; the inside of each residual block comprises a convolution layer, an activation function layer and a batch normalization layer, the first end and the last end of each residual block are provided with short jump connection, and deep features are extracted through n residual blocks; the first global long jump connection spans n residual blocks and one convolution layer, and the shallow layer features are weighted and summed with the deep layer features through the first global long jump connection to obtain associated features; and the associated features output the electric data of the target user through the batch normalization layer and the activation function layer.
It should be noted that the first global long jump connection and the short jump connection are both used to avoid gradient disappearance and improve the learning ability of the feature extraction model. The number n of residual blocks is the super parameter of the feature extraction model, is set manually before model training, and is adjusted manually according to the multiple iteration training results. In each iteration of the feature extraction model training process, the convolution kernel in the convolution operation is a parameter of the feature extraction model, after each iteration training, the output target user electrical data and the obtained target user data are compared to obtain an error between the output target user electrical data and the obtained target user data, and the convolution kernel in the convolution operation is updated according to error feedback, so that the error is finally minimized, and the feature extraction model is trained.
And (3) extracting shallow association features and deep association features between input data and target user electrical data by using the trained feature extraction model, and taking the obtained shallow association features and deep association features as one of the inputs of the load prediction model in the step (S3).
Similarly, the output data of the feature extraction model is arranged in a multi-channel matrix by type, each channel corresponding to one type of electrical data. The electric data of the target user is output by the characteristic extraction model, and the type is the active power, reactive power, electric energy, frequency, voltage, current, harmonic wave, switching state and the like spanned by the output.
It should be noted that the shallow association feature represents a shallow association relationship between input data (i.e., the first data set) and output data (i.e., the electrical data of the target user) of the feature extraction model, and the deep association feature represents a deep association relationship between the input data and the output data of the feature extraction model.
It should be noted that, the deeper the network layer number of the neural network model, the more knowledge is learned from the input data, so the shallow or deep association relationship is only related to the network layer number. In the feature extraction model of the embodiment, the shallow network consists of 1-3 convolution layers and an activation function layer, the middle output of the last layer of the shallow network is a shallow association feature, the deep network consists of a plurality of residual blocks and one convolution layer, and the middle output of the last layer of the deep network is a deep association feature.
Referring to fig. 3, in step S3, a load prediction model based on long-short term memory network (long-Short Term Memory, LSTM) is constructed, and input data of the load prediction model is a second data set, where the second data set includes: the input data of the feature extraction model is a first data set, shallow layer associated features and deep layer associated features extracted by the feature extraction model, residual production element data and residual production element categories. Specifically, similarly to the feature extraction model in step S2, the input data of the load prediction model in step S3 is arranged in a multi-channel matrix form, each channel corresponds to one type of data, and the data in each channel is rearranged in a matrix form in the order of the preceding and following columns.
The output data of the load prediction model in step S3 is the load data of the target user, and the arrangement form of the output data is a time sequence form.
In this embodiment, the process of training the load prediction model is:
firstly, utilizing a shallow network to perform nonlinear conversion, and converting input data into a nonlinear space so as to more effectively form a mapping with output data;
Then, constructing a deep network comprising a plurality of LSTM basic unit groups and three jump connections, wherein the jump connections reserve the intermediate process learning result to the following layer times across multiple layers, and the effect is that the learning result of the jump connection starting point can be reserved even if the partial learning of the jump connection starting point and the end point is not available, thereby solving the problem that the prediction precision is difficult to be improved due to the fact that the LSTM basic unit groups are stacked in the traditional method;
And finally, carrying out knowledge reservation and fusion by using the all-situation-length jump connection, and finally forming a mapping with the output.
It should be noted that m and p m in fig. 3 are hyper-parameters of the load prediction model, and are manually set before training and adjusted according to the training results of multiple iterations. The parameters to be adjusted in the training process are weights of parameters in the forgetting gate, the input gate and the output gate in the lstm basic unit and a convolution kernel in convolution operation, and after each iteration training, the parameters are updated in a feedback mode according to the output load prediction result and the error of the actual load, and finally the error is minimized.
Specifically, when the load prediction model uses a shallow network to perform nonlinear conversion, the nonlinear conversion of input data is realized by using a plurality of convolution layers and activation function layers which form the shallow network, and in a preferred embodiment, the shallow network is composed of 1-3 convolution layers and activation function layers, and the nonlinear conversion of input data is realized by using 1-3 convolution layers and activation function layers. Because the input data types are multiple and the feature dimension is high, the load prediction model utilizes the attention mechanism to realize effective screening and weighting treatment of important features and non-important features, and the weight size is determined by the contribution size of the features to the output, namely the gradient size of the features to the output.
It should be noted that, for the load prediction model, each layer of input may be referred to as a feature of data, the input data is referred to as an input feature, the data of the intermediate process is referred to as an intermediate feature or a potential feature, and a plurality of local features are also included in the input feature and the potential feature. Therefore, in this embodiment, the features processed by the attention mechanism not only refer to the shallow associated features and the deep associated features extracted by the feature extraction model, but also include all the input features and the potential features, and in this embodiment, the attention mechanism not only performs weighting processing on the input features and the potential features, but also performs weighting processing on local features inside the input features and the potential features.
Specifically, after nonlinear conversion, the embodiment constructs a deep network including a plurality of LSTM basic unit groups and three kinds of hopping connections (a second global long hopping connection, a local long hopping connection, and a local short hopping connection), and realizes feature learning and transfer through m LSTM layers.
Further, the starting point and the end point of the second global long jump connection span m LSTM layers, nonlinear conversion results of the shallow network and learning results of the m LSTM layers can be overlapped, and the overlapped intermediate results are ensured not to lose shallow characteristics due to deepening of network layers;
Further, the starting point and the end point of the local long jump connection span 1 LSTM layer, so that each LSTM layer is ensured not to cause adverse effect on the subsequent network, i.e. the learned knowledge or the learned new knowledge can be reserved. Each LSTM layer includes p m LSTM basic cell groups, which are divided into groups, i.e., each LSTM layer includes groups of LSTM basic cells.
Further, the starting point and the end point of the local short-hop connection span each LSTM basic unit group, so that the part covered by the local short-hop connection can keep learned knowledge or learn new knowledge. Each LSTM basic unit group comprises an LSTM basic unit layer, an attention mechanism layer, an activation function layer and a Dropout layer.
It should be noted that, in the load prediction model in this embodiment, a convolution operation is performed after m LSTM layers, which is aimed at mapping the learning results of m LSTM layers to the nonlinear space after nonlinear conversion of the shallow network, so that the nonlinear conversion results of the shallow network and the learning results of m LSTM layers are overlapped, and it is ensured that the intermediate result obtained after overlapping will not lose the shallow characteristics due to deepening of the network layer.
It should be noted that fig. 2 and fig. 3 are similar in overall structure, but the core is different (fig. 2 is a residual block, fig. 3 is an LSTM layer), resulting in different functions (fig. 2 is feature extraction, fig. 3 is load prediction). In the neural network, adding the convolution layer is equivalent to performing nonlinear conversion, so that the front part of fig. 2 performs nonlinear conversion to obtain shallow features, and the front part of fig. 3 performs nonlinear conversion to convert input data into nonlinear space so as to form effective mapping with output data, thereby improving the prediction capability of the conventional LSTM method.
Since the LSTM basic unit has a certain memory capacity through the cooperation of the forgetting gate, the input gate and the output gate, and can learn data of a certain period of time, the LSTM basic unit is widely used for load prediction. Conventional approaches implement load prediction by either a limited number of LSTM primitives or stacking multiple LSTM primitive groups (deepening hierarchy). However, as the hierarchy deepens and the data period becomes longer, the improvement of prediction accuracy encounters a bottleneck for two reasons: firstly, the LSTM basic unit can forget the previous part when processing data with too long period; secondly, as the hierarchy is deepened, the problem of gradient disappearance occurs, namely, knowledge learned by the front part of the network is difficult to store in the rear part of the network.
Referring to fig. 4, the LSTM basic unit provided in this embodiment specifically includes a forgetting gate, an input gate, and an output gate, where x (t) is input data at time t; s (t-1) and S (t) are respectively a t-1 and a t-moment state memory unit; h (t-1) and h (t) are respectively the intermediate output of the hidden layers at the time t-1 and the time t; sigma is a Sigmoid nonlinear activation function; Is a nonlinear activation function; the addition is the inner product operation; the forgetting gate determines a reserved part and a forgetting part of the state memory unit at the time t-1; the input gate and the forget gate jointly determine the update of the state memory unit at the moment t; the output gate determines the intermediate output of the hidden layer at the time t.
It should be noted that, in the load prediction model in this embodiment, each LSTM layer introduces a local long-jump connection and a local short-jump connection, so that knowledge learning and memory capability of data in different periods can be provided, and each LSTM layer introduces an attention mechanism so that importance judgment can be performed on data in different periods affecting a prediction result and different weights can be given. In the embodiment, a load prediction model of a long-short-time memory network based on an attention mechanism and three jump connections is built, an LSTM basic unit is introduced, excellent learning ability of the LSTM basic unit for data in a certain period is reserved, and three jump connection modes are introduced, so that knowledge can be reserved or learned to new knowledge when the network is deepened, and finally, the memory ability and the learning ability of the network are improved, and data in a longer period can be learned.
The LSTM basic unit group in the embodiment comprises an attention mechanism layer, and the constructed long-short-term memory network comprises a plurality of LSTM basic unit groups, so that the attention mechanism is added at a plurality of positions in the whole load prediction process, and the load prediction model can realize effective screening and weighting processing of important features and non-important features by using the attention mechanism.
The load prediction method based on the ubiquitous power Internet of things is provided, load prediction is carried out on a target user according to various data of a plurality of users in the range of the ubiquitous Internet of things, potential features among various data in the local Internet of things are fully excavated by utilizing high-dimensional features, and uncertainty of load prediction can be reduced; the influence of other users on the target user is fully mined through the inter-user association data, and a feature extraction model is constructed to extract association features between the first data set and the target user electrical data, so that the load prediction accuracy of the target user can be improved; the influence of the production elements on the load is fully excavated through the production element data in the user, and the classification of the production element data can improve the knowledge extraction of the production element data, thereby being beneficial to improving the efficiency of load prediction; by the load prediction model based on the long-short-term memory network, knowledge learning and memory capacity of data in different time periods can be improved, and load prediction can be performed on a target user more accurately.
Referring to fig. 5, the invention further provides an embodiment of a load prediction device based on ubiquitous power internet of things, which comprises:
the data set acquisition module 11 is used for respectively acquiring data sets of a target user and other users in the ubiquitous power internet of things range and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
A feature extraction module 12, configured to construct a feature extraction model based on a depth residual network, where the feature extraction model extracts a correlation feature between the first data set and the electrical data of the target user; wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
The load prediction module 13 is configured to construct a load prediction model based on a long-short-term memory network, where the load prediction model predicts a load according to a second data set to obtain load data of the target user, the second data set includes the first data set, the association feature, remaining production element data, and a remaining production element category, and the remaining production element data and the remaining production element category are determined according to the production element data.
Optionally, the method further comprises:
And the production element data processing module is used for calculating the similarity among all production elements according to the production element data, removing the production elements with the similarity exceeding a preset threshold, taking the data of other production elements as the residual production element data, and clustering the residual production element data to obtain the residual production element category.
Optionally, the production element data processing module calculates the similarity between all production elements according to the production element data, including:
calculating the similarity between the two production elements by using a formula ;
Wherein I (X, Y) is the similarity between the production elements X, Y, X, Y is the production element, X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
The load prediction device based on the ubiquitous power Internet of things is provided, load prediction is carried out on a target user according to various data of a plurality of users in the range of the ubiquitous Internet of things, potential features among various data in the local Internet of things are fully excavated by utilizing high-dimensional features, and uncertainty of load prediction can be reduced; the influence of other users on the target user is fully mined through the inter-user association data, and a feature extraction model is constructed to extract association features between the first data set and the target user electrical data, so that the load prediction accuracy of the target user can be improved; the influence of the production elements on the load is fully excavated through the production element data in the user, and the classification of the production element data can improve the knowledge extraction of the production element data, thereby being beneficial to improving the efficiency of load prediction; by the load prediction model based on the long-short-term memory network, knowledge learning and memory capacity of data in different time periods can be improved, and load prediction can be performed on a target user more accurately.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
In the embodiments provided in the present application, it should be understood that the disclosed system, system and method may be implemented in other manners. For example, the system embodiments described above are merely illustrative, e.g., the division of the elements is merely a logical functional division, and there may be additional divisions when actually implemented, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interface, system or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. The load prediction method based on the ubiquitous power internet of things is characterized by comprising the following steps of:
Respectively acquiring data sets of a target user and other users in the ubiquitous power Internet of things range, and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
Constructing a feature extraction model based on a depth residual network, wherein the feature extraction model extracts associated features between a first data set and electric data of a target user, and the feature extraction model comprises the following steps:
arranging data in the first data set into a multi-channel matrix form, extracting shallow layer association features through a convolution layer and an activation function layer, extracting deep layer association features through a plurality of residual blocks, and carrying out weighted summation on the shallow layer association features and the deep layer association features through a first global long jump connection to obtain association features; the first global long jump connection spans several residual blocks and one convolutional layer;
wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
Constructing a load prediction model based on a long-term and short-term memory network, wherein the load prediction model carries out load prediction according to a second data set to obtain load data of the target user, and the load prediction model comprises the following steps:
The data in the second data set are arranged into a multi-channel matrix form, nonlinear conversion is carried out through a convolution layer and an activation function layer to obtain a nonlinear conversion result, characteristic learning and transmission are carried out through a plurality of long-period and short-period memory network layers, and load data of the target user are output through a batch normalization layer, an activation function layer and a full connection layer;
the learning and transferring of the features through the plurality of long-short-term memory network layers comprises:
the long-period memory network basic unit group reserves learned knowledge or learns new knowledge through local short jump connection;
The long-short-period memory network layer reserves learned knowledge or learns new knowledge through local long-jump connection;
Superposing the nonlinear conversion result and the learning result of the long-period memory network layer through a second global long-jump connection;
The local short jump connection spans a group of long-short-term memory network basic unit groups, the local long jump connection spans one long-short-term memory network layer, the second global long jump connection spans a plurality of long-short-term memory network layers, and the long-short-term memory network layer comprises a plurality of groups of long-short-term memory network basic unit groups;
The second data set includes the first data set, the associated features, remaining production element data, and remaining production element categories, the remaining production element data and the remaining production element categories being determined from the production element data.
2. The ubiquitous power internet of things-based load prediction method according to claim 1, further comprising, before constructing the long-term and short-term memory network-based load prediction model:
And calculating the similarity among all production elements according to the production element data, removing the production elements with the similarity exceeding a preset threshold, taking the data of other production elements as the residual production element data, and clustering the residual production element data to obtain the residual production element category.
3. The ubiquitous power internet of things-based load prediction method according to claim 2, wherein calculating the similarity between all production elements according to the production element data comprises:
Calculating the similarity between the two production elements by using a formula ;
Wherein I (X, Y) is the similarity between the production elements X, Y, X, Y is the production element, X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
4. The ubiquitous power internet of things-based load prediction method according to claim 1, wherein the inter-user association data comprises:
network topology association relation, user service association degree and electricity consumption behavior similarity;
The user service association degree represents the association degree of business between users, and the electricity consumption behavior similarity is obtained by clustering the electrical data between the users.
5. Load prediction device based on ubiquitous electric power thing networking, characterized by comprising:
The data set acquisition module is used for respectively acquiring data sets of a target user and other users in the ubiquitous power internet of things range and preprocessing the data sets; wherein the data set comprises electrical data, inter-user association data, production element data inside users and load related data;
The feature extraction module is used for constructing a feature extraction model based on a depth residual error network, the feature extraction model is used for extracting association features between a first data set and electric data of a target user, and the feature extraction module comprises the following components:
arranging data in the first data set into a multi-channel matrix form, extracting shallow layer association features through a convolution layer and an activation function layer, extracting deep layer association features through a plurality of residual blocks, and carrying out weighted summation on the shallow layer association features and the deep layer association features through a first global long jump connection to obtain association features; the first global long jump connection spans several residual blocks and one convolutional layer;
wherein the first data set includes electrical data of other users, the inter-user association data, and the load-related data;
The load prediction module is used for constructing a load prediction model based on a long-term and short-term memory network, the load prediction model carries out load prediction according to a second data set to obtain load data of the target user, and the load prediction module comprises the following components:
The data in the second data set are arranged into a multi-channel matrix form, nonlinear conversion is carried out through a convolution layer and an activation function layer to obtain a nonlinear conversion result, characteristic learning and transmission are carried out through a plurality of long-period and short-period memory network layers, and load data of the target user are output through a batch normalization layer, an activation function layer and a full connection layer;
the learning and transferring of the features through the plurality of long-short-term memory network layers comprises:
the long-period memory network basic unit group reserves learned knowledge or learns new knowledge through local short jump connection;
The long-short-period memory network layer reserves learned knowledge or learns new knowledge through local long-jump connection;
Superposing the nonlinear conversion result and the learning result of the long-period memory network layer through a second global long-jump connection;
The local short jump connection spans a group of long-short-term memory network basic unit groups, the local long jump connection spans one long-short-term memory network layer, the second global long jump connection spans a plurality of long-short-term memory network layers, and the long-short-term memory network layer comprises a plurality of groups of long-short-term memory network basic unit groups;
The second data set includes the first data set, the associated features, remaining production element data, and remaining production element categories, the remaining production element data and the remaining production element categories being determined from the production element data.
6. The ubiquitous power internet of things-based load prediction device of claim 5, further comprising:
And the production element data processing module is used for calculating the similarity among all production elements according to the production element data, removing the production elements with the similarity exceeding a preset threshold, taking the data of other production elements as the residual production element data, and clustering the residual production element data to obtain the residual production element category.
7. The ubiquitous power internet of things-based load prediction device according to claim 5, wherein the production element data processing module calculates similarities between all production elements according to the production element data comprises:
Calculating the similarity between the two production elements by using a formula ;
Wherein I (X, Y) is the similarity between the production elements X, Y, X, Y is the production element, X is the data of the production element X, Y is the data of the production element Y, p (X) is the probability of X occurring in all events containing the production element X, p (Y) is the probability of Y occurring in all events containing the production element Y, and p (X, Y) is the probability of X and Y occurring simultaneously in all events.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111138775.1A CN113780679B (en) | 2021-09-27 | 2021-09-27 | Load prediction method and device based on ubiquitous power Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111138775.1A CN113780679B (en) | 2021-09-27 | 2021-09-27 | Load prediction method and device based on ubiquitous power Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113780679A CN113780679A (en) | 2021-12-10 |
| CN113780679B true CN113780679B (en) | 2024-04-16 |
Family
ID=78853955
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111138775.1A Active CN113780679B (en) | 2021-09-27 | 2021-09-27 | Load prediction method and device based on ubiquitous power Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113780679B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114548510A (en) * | 2022-01-18 | 2022-05-27 | 国网江西省电力有限公司南昌供电分公司 | Power load prediction method and system based on hybrid attention mechanism |
| CN114707765A (en) * | 2022-05-10 | 2022-07-05 | 浙江大学 | Dynamic weighted aggregation-based federated learning load prediction method |
| CN117134470B (en) * | 2023-10-26 | 2024-02-13 | 珠海市嘉德电能科技有限公司 | Sleep control method and related device of BMS battery management system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299401A (en) * | 2018-07-12 | 2019-02-01 | 中国海洋大学 | Metropolitan area space-time stream Predicting Technique based on deep learning model LSTM-ResNet |
| CN110889545A (en) * | 2019-11-20 | 2020-03-17 | 国网重庆市电力公司电力科学研究院 | Power load prediction method and device and readable storage medium |
| CN111104599A (en) * | 2019-12-23 | 2020-05-05 | 北京百度网讯科技有限公司 | Method and apparatus for outputting information |
| CN112200031A (en) * | 2020-09-27 | 2021-01-08 | 上海眼控科技股份有限公司 | Network model training method and equipment for generating image corresponding word description |
| CN112364975A (en) * | 2020-10-14 | 2021-02-12 | 山东大学 | Terminal operation state prediction method and system based on graph neural network |
-
2021
- 2021-09-27 CN CN202111138775.1A patent/CN113780679B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299401A (en) * | 2018-07-12 | 2019-02-01 | 中国海洋大学 | Metropolitan area space-time stream Predicting Technique based on deep learning model LSTM-ResNet |
| CN110889545A (en) * | 2019-11-20 | 2020-03-17 | 国网重庆市电力公司电力科学研究院 | Power load prediction method and device and readable storage medium |
| CN111104599A (en) * | 2019-12-23 | 2020-05-05 | 北京百度网讯科技有限公司 | Method and apparatus for outputting information |
| CN112200031A (en) * | 2020-09-27 | 2021-01-08 | 上海眼控科技股份有限公司 | Network model training method and equipment for generating image corresponding word description |
| CN112364975A (en) * | 2020-10-14 | 2021-02-12 | 山东大学 | Terminal operation state prediction method and system based on graph neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113780679A (en) | 2021-12-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| Dong et al. | Hourly energy consumption prediction of an office building based on ensemble learning and energy consumption pattern classification | |
| CN113780679B (en) | Load prediction method and device based on ubiquitous power Internet of things | |
| Chen et al. | Short-term load forecasting with deep residual networks | |
| Fan et al. | Development of prediction models for next-day building energy consumption and peak power demand using data mining techniques | |
| Wang et al. | A novel combined model based on hybrid optimization algorithm for electrical load forecasting | |
| Zeng et al. | A learning framework based on weighted knowledge transfer for holiday load forecasting | |
| Dou et al. | Hybrid model for renewable energy and loads prediction based on data mining and variational mode decomposition | |
| CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
| Li et al. | Building’s hourly electrical load prediction based on data clustering and ensemble learning strategy | |
| Hu et al. | Short-term photovoltaic power prediction based on similar days and improved SOA-DBN model | |
| Tian et al. | An adaptive ensemble predictive strategy for multiple scale electrical energy usages forecasting | |
| CN113591368A (en) | Comprehensive energy system multi-energy load prediction method and system | |
| Zhang et al. | Load Prediction Based on Hybrid Model of VMD‐mRMR‐BPNN‐LSSVM | |
| Chen et al. | MultiCycleNet: multiple cycles self-boosted neural network for short-term electric household load forecasting | |
| Bi et al. | Self-adaptive teaching-learning-based optimizer with improved RBF and sparse autoencoder for high-dimensional problems | |
| CN115329899A (en) | Clustering equivalent model construction method, system, equipment and storage medium | |
| Nightingale et al. | Effect of clustering in federated learning on non-IID electricity consumption prediction | |
| Li | Energy consumption forecasting with deep learning | |
| Yong et al. | Short-term building load forecasting based on similar day selection and LSTM network | |
| Prashanthi et al. | A comparative study of the performance of machine learning based load forecasting methods | |
| Liu et al. | A novel deep learning-based evolutionary model with potential attention and memory decay-enhancement strategy for short-term wind power point-interval forecasting | |
| CN114066068A (en) | Short-term power load prediction method, device, equipment and storage medium | |
| Liu et al. | Short-term power load forecasting via recurrent neural network with similar day selection | |
| Wang et al. | Short-term Load Forecasting Based On Variational Mode Decomposition And Chaotic Grey Wolf Optimization Improved Random Forest Algorithm |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |