CN109636059A - Electric heating distribution transformer load forecasting method and device - Google Patents
Electric heating distribution transformer load forecasting method and device Download PDFInfo
- Publication number
- CN109636059A CN109636059A CN201811584211.9A CN201811584211A CN109636059A CN 109636059 A CN109636059 A CN 109636059A CN 201811584211 A CN201811584211 A CN 201811584211A CN 109636059 A CN109636059 A CN 109636059A
- Authority
- CN
- China
- Prior art keywords
- data
- sample data
- predicted
- weather
- time
- 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
- 238000005485 electric heating Methods 0.000 title claims abstract description 73
- 238000013277 forecasting method Methods 0.000 title claims abstract description 15
- 238000013528 artificial neural network Methods 0.000 claims abstract description 54
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000012549 training Methods 0.000 claims abstract description 35
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 19
- 238000010801 machine learning Methods 0.000 claims abstract description 12
- 230000035772 mutation Effects 0.000 claims description 39
- 238000012545 processing Methods 0.000 claims description 22
- 238000010438 heat treatment Methods 0.000 claims description 10
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 claims description 9
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 claims description 9
- 230000005611 electricity Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 8
- 239000010410 layer Substances 0.000 description 20
- 238000010586 diagram Methods 0.000 description 14
- 230000006870 function Effects 0.000 description 9
- 238000005457 optimization Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 210000002569 neuron Anatomy 0.000 description 5
- 239000003245 coal Substances 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 239000002245 particle Substances 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 235000006508 Nelumbo nucifera Nutrition 0.000 description 2
- 240000002853 Nelumbo nucifera Species 0.000 description 2
- 235000006510 Nelumbo pentapetala Nutrition 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 239000011229 interlayer Substances 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 239000012528 membrane Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- QHGUCRYDKWKLMG-UHFFFAOYSA-N octopamine Chemical compound NCC(O)C1=CC=C(O)C=C1 QHGUCRYDKWKLMG-UHFFFAOYSA-N 0.000 description 1
- 229960001576 octopamine Drugs 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000002344 surface layer Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001131 transforming effect 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
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of electric heating distribution transformer load forecasting method and devices.Wherein, this method comprises: acquiring the sample data in the time to be predicted corresponding order history period, using combination forecasting, sample data in the order history period is predicted to obtain the load prediction data of the electric heating distribution transformer of time to be predicted, wherein, combination forecasting is the combination of integrated empirical mode decomposition EEMD prediction model and backpropagation BP neural network prediction model, wherein, EEMD prediction model and BP neural network prediction model are obtained using multiple groups training data by machine learning training, every group of training data in multiple groups training data include: historical data in historical time and after the historical time section electric heating distribution transformer of predicted time load prediction data.The present invention solves the technical problem that electric heating distribution transformer load prediction accuracy is not high in the related technology.
Description
Technical field
The present invention relates to Electric control fields, in particular to a kind of electric heating distribution transformer load forecasting method
And device.
Background technique
The power prediction of science is the guarantee of operation of power networks.Distribution network construction, operation, planning etc. all with transformation in power distribution network
The load condition of device is related.Therefore, distribution transformer load forecast is to ensureing that distribution transformer is safe and reliable, economic fortune
Row is significant.Winter cold is very long, and a big chunk energy all consumes in heating.In recent years, " coal changes electricity " war is proposed
Slightly, coal-burning boiler is substituted using electric heater or scattered coal is warmed oneself.And with the continuous popularization of " coal changes electricity " project, winter power load
It is significantly increased, rural area each family winter daily load value increases by 5~10 times afterwards for rough estimate " coal changes electricity ".Meeting pair is significantly increased in load
Mutually band ability and high-voltage fence reliability etc. brings tremendous influence between low-voltage network power supply capacity and reliability, electric substation.
In the related technology there are mainly two types of power-system short-term load forecasting methods, one is regression analyses, time sequence
The traditional prediction methods such as column method, input quantity only have load value sequence, and model is simple, and calculating speed is very fast, but can not be to examining
Consider the load prediction modeling of multifactor impact;One is using neural network, support vector machine as the machine learning method of representative,
Handling multi-factor problem (as considered meteorologic factor problem) has stronger study and analog capability, can be preferably to power load
Lotus short-term forecast, but its accuracy rate and speed are still unable to satisfy the short term of the distribution transformer of the electric heating equipment containing high proportion
Prediction requires.
For above-mentioned problem, currently no effective solution has been proposed.
Summary of the invention
The embodiment of the invention provides a kind of electric heating distribution transformer load forecasting method and devices, at least to solve phase
The not high technical problem of electric heating distribution transformer load prediction accuracy in the technology of pass.
According to an aspect of an embodiment of the present invention, a kind of electric heating distribution transformer load forecasting method is provided, is wrapped
It includes: acquiring the sample data in the time to be predicted corresponding order history period;Using combination forecasting, to described predetermined
Sample data in historical time section is predicted to obtain the load of the electric heating distribution transformer of the time to be predicted
Prediction data, wherein the combination forecasting is integrated empirical mode decomposition EEMD prediction model and backpropagation BP nerve
The combination of Network Prediction Model, wherein the EEMD prediction model and the BP neural network prediction model are to be instructed using multiple groups
Practice what data were obtained by machine learning training, every group of training data in the multiple groups training data includes: historical time
Interior historical data and after the historical time section electric heating distribution transformer of predicted time load prediction data.
Optionally, the sample data packet in the time to be predicted corresponding order history period is acquired
It includes: according to the Changes in weather index of target date, acquiring in the time to be predicted corresponding order history period
The sample data, wherein the sample data is divided into normal weather sample data and mutation according to the Changes in weather index
Weather sample data.
Optionally, in the Changes in weather index according to the target date, it is corresponding to acquire the time to be predicted
Before the sample data in the order history period, further includes: in the following manner, determine that the Changes in weather refers to
Number E:Wherein, P1, P2At predetermined number time point in the afternoon cumulative for the load at morning target date,
Value.
Optionally, using the combination forecasting, the sample data in the order history period is carried out
It includes: in the sample that prediction, which obtains the load prediction data of the electric heating distribution transformer of the time to be predicted,
In the case that data are normal weather sample data, the corresponding mean daily temperature of the normal weather sample data and date are obtained
Type;The corresponding mean daily temperature of the normal weather sample data and date type based on acquisition, it is pre- using the combination
Model is surveyed, the normal weather sample data in the order history period is predicted to obtain the time to be predicted
The electric heating distribution transformer the load prediction data, wherein the normal weather sample data is corresponding normal
Weather includes at least one of: fine day, and at the cloudy day, the rainy day, snowy day, the date type includes: working day or day off;
And/or in the case where the sample data is mutation weather sample data, it is corresponding to obtain the mutation weather sample data
Daily minimum temperature and Changes in weather index;The corresponding Daily minimum temperature of the mutation weather sample data and weather based on acquisition
Variability index, using the combination forecasting, to the mutation weather sample data in the order history period into
Row prediction obtains the load prediction data of the electric heating distribution transformer of the time to be predicted.
Optionally, use the combination forecasting, to the sample data in the order history period into
Row prediction obtains after the load prediction data of the electric heating distribution transformer of the time to be predicted, further includes:
Error processing is carried out to the obtained load prediction data, obtains error result, wherein the error result include with down toward
It is one of few: root-mean-square error RMSE, mean absolute percentage error MAPE, absolute percent error AE, maximum relative error emax。
According to another aspect of an embodiment of the present invention, a kind of electric heating distribution transformer load prediction device is additionally provided,
It include: acquisition module, for acquiring the sample data in the time to be predicted corresponding order history period;Module is obtained, is used
In using combination forecasting, the sample data in the order history period is predicted to obtain the time to be predicted
The electric heating distribution transformer load prediction data, wherein the combination forecasting is integrated empirical mode decomposition
The combination of EEMD prediction model and backpropagation BP neural network prediction model, wherein the EEMD prediction model and the BP
Neural network prediction model is obtained using multiple groups training data by machine learning training, in the multiple groups training data
Every group of training data includes: historical data in historical time and the electricity of predicted time is adopted after the historical time section
The load prediction data of warm distribution transformer.
Optionally, the acquisition module includes: acquisition unit, for the Changes in weather index according to target date, acquisition
The sample data in time to be predicted corresponding order history period, wherein the sample data according to
The Changes in weather index is divided into normal weather sample data and mutation weather sample data.
Optionally, the acquisition module further include: determination unit, for becoming according to the weather of the target date
Change index, before acquiring the sample data in time to be predicted corresponding order history period, by with
Under type determines the Changes in weather index E:Wherein, P1, P2For morning target date, afternoon
The load accumulated value at predetermined number time point.
Optionally, the module that obtains includes: first processing units, for being normal weather sample in the sample data
In the case where data, the corresponding mean daily temperature of the normal weather sample data and date type are obtained;Institute based on acquisition
The corresponding mean daily temperature of normal weather sample data and date type are stated, using the combination forecasting, to described predetermined
The normal weather sample data in historical time section is predicted to obtain the electric heating distribution of the time to be predicted
The load prediction data of transformer, wherein the corresponding normal weather of the normal weather sample data include it is following at least
One of: fine day, at the cloudy day, the rainy day, snowy day, the date type includes: working day or day off;And/or second processing list
Member, for it is corresponding to obtain the mutation weather sample data in the case where the sample data is mutation weather sample data
Daily minimum temperature and Changes in weather index;The corresponding Daily minimum temperature of the mutation weather sample data and day based on acquisition
Gas variability index, using the combination forecasting, to the mutation weather sample data in the order history period
It is predicted to obtain the load prediction data of the electric heating distribution transformer of the time to be predicted.
Optionally, described device further include: processing module, for using the combination forecasting, to described predetermined
The sample data in historical time section is predicted to obtain the electric heating distribution transformer of the time to be predicted
After the load prediction data, Error processing is carried out to the obtained load prediction data, obtains error result, wherein
The error result includes at least one of: root-mean-square error RMSE, and mean absolute percentage error MAPE, absolute percentage misses
Poor AE, maximum relative error emax。
In embodiments of the present invention, using integrated empirical mode decomposition EEMD prediction model and backpropagation BP neural network
The mode of prediction model combination by acquiring the sample data in the time to be predicted corresponding order history period, and uses
Said combination prediction model has reached and has been predicted the sample data in the order history period to obtain the electricity of time to be predicted
The purpose of the load prediction data of heating distribution transformer, so that the technical effect for effectively reducing load prediction error is realized,
And then solves the technical problem that electric heating distribution transformer load prediction accuracy is not high in the related technology.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of electric heating distribution transformer load forecasting method according to an embodiment of the present invention;
Fig. 2 is a kind of structure chart of BP neural network model according to an embodiment of the present invention;
Fig. 3 is a kind of flow chart of BP neural network prediction according to an embodiment of the present invention;
Fig. 4 is a kind of flow chart of EEMD-BP neural network ensemble prediction technique of preferred embodiment according to the present invention;
Fig. 5 is the sample original value of preferred embodiment according to the present invention and the schematic diagram of EEMD decomposition amount;
Fig. 6 is the prediction curve and actual value of the mutation lower three kinds of prediction techniques of weather of preferred embodiment according to the present invention
The schematic diagram of curve;
Fig. 7 is the structural schematic diagram of electric heating distribution transformer load prediction device according to an embodiment of the present invention;
Fig. 8 is the structure of the acquisition module 72 of electric heating distribution transformer load prediction device according to an embodiment of the present invention
Schematic diagram;
Fig. 9 is the optimization of the acquisition module 72 of electric heating distribution transformer load prediction device according to an embodiment of the present invention
Structural schematic diagram;
Figure 10 is that electric heating distribution transformer load prediction device according to an embodiment of the present invention obtains the knot of module 74
Structure schematic diagram;
Figure 11 is the optimization structural schematic diagram of electric heating distribution transformer load prediction device according to an embodiment of the present invention.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
The model that the present invention protects all should belong in member's every other embodiment obtained without making creative work
It encloses.
It should be noted that description and claims of this specification and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to the embodiment of the present invention described herein can in addition to illustrating herein or
Sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, for example, the process, method, system, product or equipment for containing a series of steps or units are not necessarily limited to
Step or unit those of is clearly listed, but may include be not clearly listed or for these process, methods, product
Or other step or units that equipment is intrinsic.
In Electric control field, BP (Back Propagation) neural network prediction is because its plasticity is strong, structure is simple
The advantages of be widely used.In the related technology using neural network, support vector machine as the machine learning method of representative
Power-system short-term load forecasting method, processing multi-factor problem (as consider meteorologic factor problem) have it is stronger study and
Analog capability, can be better to Short-Term Load Forecasting.Specifically, such as by establishing sending out without irradiation level for BP neural network
Electricity short-period forecast model, while using self-organizing feature map SOM (Self-Organizing feature Map) to cloud amount
Forecast information carries out weather pattern clustering recognition, avoids the overfitting problem of single type neural network;For another example, pass through first
Learning vector quantizations LVQ (Learning Vector Quantization) neural network classifies to weather pattern, then
It is predicted using BP neural network, using Genetic Algorithms (Genetic Algorithm) optimization BP nerve during prediction
The weight and threshold values of network improve the training speed of prediction network;Also such as, GA and particle swarm algorithm PSO will be used respectively
The BP neural network of (Particle Swarm Optimization) optimization compares, and knows PSO-BP algorithm compared to GA-BP
The model optimization effect of algorithm is more preferable, and prediction error is lower, and applicability is stronger;For another example propose that a kind of parallel film calculates PMC
The electric load short-term groupings prediction technique of (Parallel Membrane Computing), will be outside linear regression model (LRM), trend
The support vector machines for pushing away model, updated gray correlation analysis and particle group optimizing parameter is respectively put into underlying membrane prediction, is then input to
Optimized weight coefficient combination in the film of surface layer at times.
According to embodiments of the present invention, a kind of embodiment of the method for electric heating distribution transformer load prediction is provided, is needed
Illustrate, step shown in the flowchart of the accompanying drawings can be in a computer system such as a set of computer executable instructions
It executes, although also, logical order is shown in flow charts, and it in some cases, can be to be different from herein suitable
Sequence executes shown or described step.
Fig. 1 is the flow chart of electric heating distribution transformer load forecasting method according to an embodiment of the present invention, such as Fig. 1 institute
Show, this method comprises the following steps:
Step S102 acquires the sample data in the time to be predicted corresponding order history period;
Step S104, using combination forecasting, to the sample data in the order history period predicted to obtain to
The load prediction data of the electric heating distribution transformer of predicted time, wherein combination forecasting is integrated empirical mode decomposition
The combination of EEMD prediction model and backpropagation BP neural network prediction model, wherein EEMD prediction model and BP neural network
Prediction model is obtained using multiple groups training data by machine learning training, every group of training data in multiple groups training data
Include: historical data in historical time and after the historical time section electric heating distribution transformer of predicted time it is negative
Lotus prediction data.
In embodiments of the present invention, using integrated empirical mode decomposition EEMD prediction model and backpropagation BP neural network
The mode of prediction model combination by acquiring the sample data in the time to be predicted corresponding order history period, and uses
Said combination prediction model has reached and has been predicted the sample data in the order history period to obtain the electricity of time to be predicted
The purpose of the load prediction data of heating distribution transformer, so that the technical effect for effectively reducing load prediction error is realized,
And then solves the technical problem that electric heating distribution transformer load prediction accuracy is not high in the related technology.
In embodiments of the present invention, above-mentioned EEMD prediction model and BP neural network prediction model pass through machine learning training
It obtains, multiple groups training data is trained by the way of machine learning, wherein above-mentioned every group of training data include:
Historical data in historical time and after the historical time section electric heating distribution transformer of predicted time load prediction
Data.Wherein, the historical data in historical time may include the sample data in the order history period.It is obtained after training
EEMD prediction model and BP neural network prediction model, can accurately be predicted according to the historical data in historical time at this
The load prediction data of the electric heating distribution transformer of predicted time after historical time section.Therefore, by machine learning training
Combination forecasting can be predicted to obtain the electric heating of time to be predicted according to the sample data in the order history period and match
The load prediction data of piezoelectric transformer, substantially increases precision of prediction.
Wherein, integrate empirical mode decomposition EEMD (Ensemble Empirical Mode Decomposition) be through
The improved method for testing mode decomposition EMD (Empirical Mode Decomposition), EMD are mentioned by Norden E.Huang
Non-stationary signal, can be resolved into the feature modular function of different frequency by a kind of signal processing method of circulation screening removing out
Component IMF (Intrinsic Mode Function).But EMD is easy to appear modal overlap problem.EEMD changes as EMD's
Into method, by the way that white noise is added in original signal and repeatedly carries out EMD decomposition, to eliminate the modal overlap problem of EMD.It is logical
It crosses and load sequence is decomposed with EEMD, then each component is predicted using different dynamic neural networks, or using EEMD points
The algorithm that solution, GA-BP are predicted carries out forecasting wind speed, achieves good prediction effect.
Preferably to illustrate the utilization of EEMD in a preferred embodiment of the invention, simply it is situated between to EEMD method first
It continues.The daily load sequence of non-stationary is resolved into the component sequence of 4 groups of frequencies from low to high with EEMD method by the embodiment of the present invention
Column and 1 group of residual components sequence.Optionally, EEMD decomposition step is as follows:
(1) setting EMD always decomposes number M, and wherein M can take 100;Initialize EMD experiment number m, white noise acoustic amplitude system
Number k is 1;
(2) it executes the m times EMD to decompose: random Gaussian white noise sequence l being added in former daily load sequence x (t)m(t),
Obtain sequence x to be processedm(t), i.e.,
K takes and 0.5 is advisable in formula;
(3) by xm(t) EMD processing is carried out, 4 IMF component c are obtainedm.i(t) (i=1,2,3,4) and a residual components rm
(t);
(4) if m < M, return step (2), m=m+1;If m=M, carry out in next step, having obtained M group IMF component at this time
And residual components;
(5) M group IMF component and residual components mean value are calculated, obtains the decomposition result of final EEMD, i.e., 4 groups of frequencies are by low
Supreme vector sequence and 1 group of residual components sequence:
Above-mentioned EEMD method is mainly that the white noise signal of random addition normal distribution carries out multiple EMD decomposition, so as to divide
Average value after solution can eliminate influence of noise and EMD modal overlap problem.
Meanwhile said combination prediction model is the combination of EEMD method Yu above-mentioned BP neural network, i.e., is situated between in addition to above-mentioned
The EEMD method to continue, additionally uses above-mentioned BP neural network, wherein the advantages of BP neural network be to extract by learning and
The non-linear relation output and input is approached, training only needs input data, do not need to do complex hypothesis to input data.BP
Neural network has an input layer, one or more hidden layers and an output layer, and every layer has multiple neurons, interlayer neuron
Realize power connection, same layer neuron is connectionless.Fig. 2 is a kind of structure of BP neural network model according to an embodiment of the present invention
Figure, as shown in Fig. 2, input layer, hidden layer and the output layer number of plies are 1 in the BP neural network model, neuron number difference
For n, m, l;xi(i=1,2 ..., n), yj(j=1,2 ..., m), ok(k=1,2 ..., l) it is input layer, hidden layer, output layer
Output vector;dkFor desired output;Weigh vij, power wjkFor interlayer neuron weight;αk、θjFor output layer, hidden layer threshold value.
For hidden layer, have
yj=f (netj)
For output layer, have
ok=f (netk)
In above two formula, transforming function transformation function f (x) is unipolarity Sigmoid function
When network output O and desired output D is not waited, there are output error E, are defined as follows:
It further spreads out to hidden layer
Find out that error E is the function about weight w, v by formula (8).In order to learn to reduce error, the adjustment amount of weight every time
Should be directly proportional to the decline of the gradient of error, i.e.,
Wherein η is learning rate, is the network parameter of BP neural network;Define error letter
NumberFor
Then the final expression formula of BP neural network weighed value adjusting amount is
Δwjk=η (dk-ok)ok(1-ok)yj (13)
Similarly, BP neural network adjusting thresholds amount is
In conjunction with above-mentioned calculation formula, Fig. 3 is a kind of flow chart of BP neural network prediction according to an embodiment of the present invention, should
The pre- flow gauge of BP neural network is as shown in Figure 3.
Electric heating based on integrated empirical mode decomposition EEMD and BP neural network algorithm combination in the embodiment of the present invention is matched
Piezoelectric transformer load forecasting method, it is also contemplated that the environmental conditions such as heating day weather pattern, heating degree/day are to resident's heating row
For influence.Preferably, acquiring the sample data in the time to be predicted corresponding order history period may include: according to pre-
The Changes in weather index fixed the date acquires the sample data in the time to be predicted corresponding order history period, wherein sample
Data are divided into normal weather sample data and mutation weather sample data according to Changes in weather index.Wherein, normal weather includes
Fine day, cloudy day, snowy day, rainy day, corresponding Changes in weather index are less than certain predetermined threshold.And when weather variability index is big
When being equal to above-mentioned predetermined threshold, then mutation weather is identified as.Because different weather is to electric heating distribution transformer loading effects
Difference, therefore by the above-mentioned acquisition to different weather sample data, the accuracy of load prediction is improved, varying environment condition is made
Influence prediction of the lower resident's heating behavior to power grid more refines.
When being calculated for the Changes in weather index of target date, it is preferred that in the Changes in weather according to target date
Index can also include: by with lower section before acquiring the sample data in the time to be predicted corresponding order history period
Formula determines Changes in weather index E:Wherein, P1, P2When for morning target date, predetermined number in afternoon
Between the load accumulated value put.Wherein, above-mentioned Changes in weather index predetermined threshold can be set to 30%, and then when E >=30% is prominent
Restore a reactionary rule gas, otherwise is normal weather.Sample is divided into normal weather sample and mutation day gas sample according to the size of the daily E of sample
This.Since load and atmospheric temperature variation all have certain Retarder theory, related data can be every 1 hour record one
It is secondary, while above-mentioned predetermined number can be 12.
Preferably, using combination forecasting, the sample data in the order history period is predicted to obtain to pre-
It is normal weather sample data that the load prediction data for surveying the electric heating distribution transformer of time, which may include: in sample data,
In the case of, obtain the corresponding mean daily temperature of normal weather sample data and date type;Normal weather sample based on acquisition
The corresponding mean daily temperature of data and date type, using combination forecasting, to the normal weather in the order history period
Sample data is predicted to obtain the load prediction data of the electric heating distribution transformer of time to be predicted, wherein normal weather
The corresponding normal weather of sample data includes at least one of: fine day, and at the cloudy day, the rainy day, snowy day, date type includes: work
Day or day off;And/or in the case where sample data is mutation weather sample data, obtain mutation weather sample data
Corresponding Daily minimum temperature and Changes in weather index;The corresponding Daily minimum temperature of mutation weather sample data and day based on acquisition
Gas variability index is predicted to obtain using combination forecasting to the mutation weather sample data in the order history period
The load prediction data of the electric heating distribution transformer of time to be predicted.Not because of work daily power consumption and rest daily power consumption
Together, therefore working day or day off are distinguished in above-mentioned date type.Temperature change is obvious when simultaneously because of improper weather, therefore
Daily minimum temperature, the progress load prediction of Changes in weather index are also contemplated for mutation weather sample data, and in normal weather
Only consider that mean daily temperature, date type can be predicted in sample data.The setting further improves different weather shape
Prediction precision under state.Meanwhile above-mentioned load data, temperature data, Changes in weather index cannot directly enter combined prediction
The input layer of BP neural network in model is inputted again after being normalized using mapminmax function, and day
The quantizing method of phase type can be with are as follows: working day takes 0, and day off takes 1.
By the embodiment of the present invention and all preferred embodiments, it is preferred that using combination forecasting, gone through to predetermined
Sample data in the history period predicted to obtain the electric heating distribution transformer of time to be predicted load prediction data it
Afterwards, it can also include: that Error processing is carried out to obtained load prediction data, obtain error result, wherein error result can be with
Including at least one of: root-mean-square error RMSE, mean absolute percentage error MAPE, absolute percent error AE, maximum phase
To error emax.By above-mentioned Error processing, error analysis can be carried out to the result of prediction, meanwhile, by above-mentioned several mistakes
The selection of poor index, can be clear judge to match using the electric heating of the embodiment of the present invention and all preferred embodiments
What piezoelectric transformer load forecasting method carried out predicts whether precisely.
Based on the above embodiment and preferred embodiment additionally provides a kind of EEMD-BP in a preferred embodiment of the invention
Neural network ensemble prediction technique, Fig. 4 are a kind of EEMD-BP neural network ensemble predictions of preferred embodiment according to the present invention
The flow chart of method, as shown in figure 4, this method mainly contains the following steps:
(1) it determines Changes in weather index and carries out data normalized, for example, public using above-mentioned Changes in weather index E
Opening determining weather is that normal weather is still mutated weather, and place is normalized to sampled data using mapminmax function
Reason;
(2) according to Changes in weather index by sample be divided into normal weather sample and mutation weather sample, choose 15 days fine days,
Cloudy day, snowy day, rainy day and mutation weather sample are used as check sample on the 15th day wherein being used as within first 14 days training sample;
(3) sample data is decomposed and combined using EEMD method: daily 24 × 1 load data carries out EEMD points
Solution, obtains the IMF component and 1 group of residual components of 4 groups of different frequencies, by taking on January 5th, 2017 (fine) as an example, Fig. 5 is according to this hair
The sample original value of bright preferred embodiment and the schematic diagram of EEMD decomposition amount, wherein 5 components after being decomposed using EEMD are big
It is small as shown in Figure 5;By 14 days daily by EEMD IMF1, IMF2, IMF3, IMF4 decomposed and residual components combination, 5 groups are obtained
24 × 14 data;
(4) BP neural network is predicted: being considered mean daily temperature, date type when data are predicted under normal weather, is dashed forward
Become synoptic model and also considers Daily minimum temperature, Changes in weather index;
(5) sum and carry out error analysis: the following index of utilization judges prediction effect, root-mean-square error RMSE, is averaged absolutely
To percentage error MAPE, absolute percent error AE, maximum relative error emax
Wherein: Yi' and YiRespectively indicate i-th of predicted value and actual value.
For example, in embodiments of the present invention, a kind of tool of EEMD-BP neural network ensemble prediction technique is additionally provided
Body application, for example, load data and region gas with Beijing electric heating distribution transformer in a certain historical time section
Image data carries out above-mentioned EEMD-BP Application of Neural Network analysis.Wherein, BP neural network input layer, hidden layer and output layer mind
It is respectively 27,10,24 through first number, parameter value is chosen as shown in table 1.
Prediction result is analyzed under table 1BP neural network parameter value (one) normal weather
According to the EEMD-BP prediction model that above-mentioned preferred embodiment is established, respectively to fine day, cloudy day, rainy day and snowy day
Load predicted.And BP neural network is respectively adopted, after EMD-BP neural network is predicted in contrast, comparing result
As shown in 2-table of table 5:
The fine day load prediction error statistics of 23 kinds of methods of table
The cloudy load prediction error statistics of 33 kinds of methods of table
The rainy day load prediction error statistics of 43 kinds of methods of table
The snowy day load prediction error statistics of 53 kinds of methods of table
By table 2- table 5 it can be seen that compared to neural net prediction methods such as BP, EMD-BP, EEMD-BP is predicted average exhausted
It is relatively small to percentage error MAPE, illustrate that EEMD-BP prediction precision is higher;Root-mean-square error RMSE, maximum are opposite accidentally
Poor emaxIt is relatively small, i.e., it is more acurrate to the mutability prediction of initial data;AE is also relatively small, illustrates EEMD-BP prediction whole
Initial data is bonded on body.
(2) prediction result under weather is mutated to analyze
According to the EEMD-BP prediction model that above-mentioned preferred embodiment is established, predicted in the case where being mutated weather, and use
BP neural network, EMD-BP neural network predicted after in contrast, comparing result is as shown in table 6:
The mutation weather load of 63 kinds of methods of table predicts error statistics
Fig. 6 is the prediction curve and actual value of the mutation lower three kinds of prediction techniques of weather of preferred embodiment according to the present invention
The schematic diagram of curve, three kinds of prediction techniques and actual value comparable situation are as shown in Figure 6.
As can be seen from Table 6, the mutation lower 4 kinds of errors of weather are universal larger compared with normal weather error.EEMD-BP's
Mean absolute percentage error MAPE is 14.33%, and precision of prediction is higher than other two kinds of models, illustrates EEMD-BP in certain journey
It can be improved the accuracy of load prediction under mutation weather on degree.The maximum relative error e of EEMD-BPmaxIt is 23.25%, occurs
Moment is 4:00, and actual value experienced primary larger decline at this time, illustrates that EEMD-BP is more quasi- to the mutability prediction of initial data
Really.
And then the EEMD-BP neural network ensemble that can be seen that preferred embodiment of the invention design from table 1-6, Fig. 6 is pre-
Survey method can preferably predict that the variation tendency of load, rate of precision increase compared with BP neural network, EMD-BP neural network.
Since prediction model considers the influence of heating degree/day and weather condition to resident's heating behavior in method, therefore answered according to above-mentioned
With analysis shows, EEMD-BP neural network ensemble prediction technique can effectively realize the distribution of the electric heating equipment containing high proportion
The short-term load forecasting of transformer, and effectively reduce load prediction error.
According to embodiments of the present invention, a kind of device of electric heating distribution transformer load prediction is additionally provided, Fig. 7 is basis
The structural schematic diagram of the electric heating distribution transformer load prediction device of the embodiment of the present invention, as shown in fig. 7, the device includes:
Acquisition module 72 obtains module 74.The electric heating distribution transformer load prediction device is illustrated below.
Acquisition module 72, for acquiring the sample data in the time to be predicted corresponding order history period;
Module 74 is obtained, above-mentioned acquisition module 72 is connected to, for using combination forecasting, to the order history period
Interior sample data is predicted to obtain the load prediction data of the electric heating distribution transformer of time to be predicted, wherein combination
Prediction model is the combination of integrated empirical mode decomposition EEMD prediction model and backpropagation BP neural network prediction model,
In, EEMD prediction model and BP neural network prediction model are obtained using multiple groups training data by machine learning training,
Every group of training data in multiple groups training data include: historical data in historical time and after the historical time section it is pre-
Survey the load prediction data of the electric heating distribution transformer of time.
Fig. 8 is the structure of the acquisition module 72 of electric heating distribution transformer load prediction device according to an embodiment of the present invention
Schematic diagram, as shown in figure 8, the acquisition module 72 includes: acquisition unit 82.The acquisition module 72 is illustrated below.
Acquisition unit 82, for the Changes in weather index according to target date, acquiring the time to be predicted corresponding predetermined is gone through
Sample data in the history period, wherein sample data is divided into normal weather sample data and mutation according to Changes in weather index
Weather sample data.
Fig. 9 is the optimization of the acquisition module 72 of electric heating distribution transformer load prediction device according to an embodiment of the present invention
Structural schematic diagram, as shown in figure 9, the acquisition module 72 is in addition to containing all mechanisms in Fig. 8, further includes: determination unit 92.It is right below
The acquisition module 72 is illustrated.
Determination unit 92 is connected to above-mentioned acquisition unit 82, in the Changes in weather index according to target date, acquisition
Before sample data in time to be predicted corresponding order history period, in the following manner, Changes in weather index is determined
E:Wherein, P1, P2For the load accumulated value in morning target date, at predetermined number time point in the afternoon.
Figure 10 is that electric heating distribution transformer load prediction device according to an embodiment of the present invention obtains the knot of module 74
Structure schematic diagram, as shown in Figure 10, it includes at least one of: first processing units 102, the second processing unit that this, which obtains module 74,
104.Module 74 is obtained to this below to be illustrated.
First processing units 102, for obtaining normal weather in the case where sample data is normal weather sample data
The corresponding mean daily temperature of sample data and date type;The corresponding mean daily temperature of normal weather sample data based on acquisition
And date type is predicted to obtain using combination forecasting to the normal weather sample data in the order history period
The load prediction data of the electric heating distribution transformer of time to be predicted, wherein the corresponding normal day of normal weather sample data
Gas bag includes at least one of: fine day, and at the cloudy day, the rainy day, snowy day, date type includes: working day or day off;
The second processing unit 104, for obtaining mutation weather in the case where sample data is mutation weather sample data
The corresponding Daily minimum temperature of sample data and Changes in weather index;Mutation weather sample data corresponding day based on acquisition is minimum
Temperature and Changes in weather index carry out the mutation weather sample data in the order history period using combination forecasting
Prediction obtains the load prediction data of the electric heating distribution transformer of time to be predicted.
Figure 11 is the optimization structural schematic diagram of electric heating distribution transformer load prediction device according to an embodiment of the present invention,
As shown in figure 11, the device is in addition to containing all structures in Fig. 7, further includes: processing module 112.The electric heating distribution is become below
Depressor load prediction device is illustrated.
Processing module 112, be connected to it is above-mentioned obtain module 74, for use combination forecasting, when to order history
Between sample data in section predicted after obtaining the load prediction data of the electric heating distribution transformer of time to be predicted, it is right
Obtained load prediction data carries out Error processing, obtains error result, wherein error result includes at least one of:
Square error RMSE, mean absolute percentage error MAPE, absolute percent error AE, maximum relative error emax。
According to another aspect of an embodiment of the present invention, a kind of storage medium is additionally provided, which includes storage
Program, wherein equipment where control storage medium executes the electric heating distribution transformer of above-mentioned any one in program operation
Load forecasting method.
According to another aspect of an embodiment of the present invention, a kind of processor is additionally provided, which is used to run program,
In, program executes the electric heating distribution transformer load forecasting method of above-mentioned any one when running.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
In the above embodiment of the invention, it all emphasizes particularly on different fields to the description of each embodiment, does not have in some embodiment
The part of detailed description, reference can be made to the related descriptions of other embodiments.
In several embodiments provided herein, it should be understood that disclosed technology contents can pass through others
Mode is realized.Wherein, the apparatus embodiments described above are merely exemplary, such as the division of the unit, Ke Yiwei
A kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or components can combine or
Person is desirably integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Between coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or communication link of unit or module
It connects, can be electrical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
On unit.It can some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can for personal computer, server or network equipment etc.) execute each embodiment the method for the present invention whole or
Part steps.And storage medium above-mentioned includes: that USB flash disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic or disk etc. be various to can store program code
Medium.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (10)
1. a kind of electric heating distribution transformer load forecasting method characterized by comprising
Acquire the sample data in the time to be predicted corresponding order history period;
Using combination forecasting, when being predicted to obtain described to be predicted to the sample data in the order history period
Between the electric heating distribution transformer load prediction data, wherein the combination forecasting is integrated empirical modal point
Solve the combination of EEMD prediction model and backpropagation BP neural network prediction model, wherein the EEMD prediction model and described
BP neural network prediction model is obtained using multiple groups training data by machine learning training, in the multiple groups training data
Every group of training data include: historical data in historical time and after the historical time section predicted time the electricity
The load prediction data of heating distribution transformer.
2. the method according to claim 1, wherein acquiring the time to be predicted corresponding order history
The sample data in period includes:
According to the Changes in weather index of target date, acquire in the time to be predicted corresponding order history period
The sample data, wherein the sample data is divided into normal weather sample data and mutation according to the Changes in weather index
Weather sample data.
3. according to the method described in claim 2, it is characterized in that, referring to according to the Changes in weather of the target date
Number, before acquiring the sample data in time to be predicted corresponding order history period, further includes:
In the following manner, the Changes in weather index E is determined:
Wherein, P1, P2For the load accumulated value in morning target date, at predetermined number time point in the afternoon.
4. the method according to claim 1, wherein using the combination forecasting, to the order history
The sample data in period is predicted to obtain the described of the electric heating distribution transformer of the time to be predicted
Load prediction data includes:
In the case where the sample data is normal weather sample data, the normal weather sample data corresponding day is obtained
Mean temperature and date type;The corresponding mean daily temperature of the normal weather sample data and date type based on acquisition,
Using the combination forecasting, the normal weather sample data in the order history period is predicted to obtain
The load prediction data of the electric heating distribution transformer of the time to be predicted, wherein the normal weather sample
The corresponding normal weather of data includes at least one of: fine day, and at the cloudy day, the rainy day, snowy day, the date type includes: work
Day or day off;
And/or
In the case where the sample data is mutation weather sample data, the mutation weather sample data corresponding day is obtained
Minimum temperature and Changes in weather index;The corresponding Daily minimum temperature of the mutation weather sample data and weather based on acquisition become
Change index, using the combination forecasting, the mutation weather sample data in the order history period is carried out
Prediction obtains the load prediction data of the electric heating distribution transformer of the time to be predicted.
5. method according to claim 1 to 4, which is characterized in that the combination forecasting is being used, it is right
The sample data in the order history period is predicted to obtain the electric heating distribution of the time to be predicted
After the load prediction data of transformer, further includes:
Error processing is carried out to the obtained load prediction data, obtains error result, wherein the error result include with
It is at least one lower: root-mean-square error RMSE, mean absolute percentage error MAPE, absolute percent error AE, maximum relative error
emax。
6. a kind of electric heating distribution transformer load prediction device characterized by comprising
Acquisition module, for acquiring the sample data in the time to be predicted corresponding order history period;
Module is obtained, for using combination forecasting, the sample data in the order history period is measured in advance
To the load prediction data of the electric heating distribution transformer of the time to be predicted, wherein the combination forecasting is
The combination of integrated empirical mode decomposition EEMD prediction model and backpropagation BP neural network prediction model, wherein the EEMD
Prediction model and the BP neural network prediction model are obtained using multiple groups training data by machine learning training, described
Every group of training data in multiple groups training data include: historical data in historical time and after the historical time section it is pre-
Survey the load prediction data of the electric heating distribution transformer of time.
7. device according to claim 6, which is characterized in that the acquisition module includes:
It is corresponding described predetermined to acquire the time to be predicted for the Changes in weather index according to target date for acquisition unit
The sample data in historical time section, wherein the sample data is divided into normal weather according to the Changes in weather index
Sample data and mutation weather sample data.
8. device according to claim 7, which is characterized in that the acquisition module further include:
Determination unit, for it is corresponding to acquire the time to be predicted in the Changes in weather index according to the target date
The order history period in the sample data before, in the following manner, determine the Changes in weather index E:
Wherein, P1, P2For the load accumulated value in morning target date, at predetermined number time point in the afternoon.
9. device according to claim 6, which is characterized in that the module that obtains includes:
First processing units, for obtaining the normal day in the case where the sample data is normal weather sample data
The corresponding mean daily temperature of gas sample notebook data and date type;The normal weather sample data corresponding day based on acquisition is flat
Equal temperature and date type, using the combination forecasting, to the normal weather sample in the order history period
Notebook data is predicted to obtain the load prediction data of the electric heating distribution transformer of the time to be predicted,
In, the corresponding normal weather of the normal weather sample data includes at least one of: fine day, cloudy day, rainy day, snowy day, institute
Stating date type includes: working day or day off;
And/or
The second processing unit, for obtaining the mutation day in the case where the sample data is mutation weather sample data
The corresponding Daily minimum temperature of gas sample notebook data and Changes in weather index;The mutation weather sample data based on acquisition is corresponding
Daily minimum temperature and Changes in weather index, using the combination forecasting, to described prominent in the order history period
Change of weather gas sample notebook data is predicted to obtain the load prediction of the electric heating distribution transformer of the time to be predicted
Data.
10. device according to any one of claims 6 to 9, which is characterized in that described device further include:
Processing module, for using the combination forecasting, to the sample data in the order history period
It is predicted after obtaining the load prediction data of the electric heating distribution transformer of the time to be predicted, to obtaining
The load prediction data carry out Error processing, obtain error result, wherein the error result include it is following at least it
One: root-mean-square error RMSE, mean absolute percentage error MAPE, absolute percent error AE, maximum relative error emax。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811584211.9A CN109636059A (en) | 2018-12-24 | 2018-12-24 | Electric heating distribution transformer load forecasting method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811584211.9A CN109636059A (en) | 2018-12-24 | 2018-12-24 | Electric heating distribution transformer load forecasting method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109636059A true CN109636059A (en) | 2019-04-16 |
Family
ID=66077093
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811584211.9A Pending CN109636059A (en) | 2018-12-24 | 2018-12-24 | Electric heating distribution transformer load forecasting method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109636059A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111142049A (en) * | 2020-01-16 | 2020-05-12 | 合肥工业大学 | Intelligent transformer fault diagnosis method based on edge cloud cooperation mechanism |
CN111581883A (en) * | 2020-05-09 | 2020-08-25 | 国网上海市电力公司 | Method for calculating and predicting load on automation device |
CN112488404A (en) * | 2020-12-07 | 2021-03-12 | 广西电网有限责任公司电力科学研究院 | Multithreading efficient prediction method and system for large-scale power load of power distribution network |
CN112564085A (en) * | 2020-10-22 | 2021-03-26 | 国网山东省电力公司济宁供电公司 | Method and system for predicting maximum power load of electric heating distribution transformer |
CN115343959A (en) * | 2022-09-23 | 2022-11-15 | 国网北京市电力公司 | Electric heating load self-adaptive control method, device, equipment and medium |
CN116701481A (en) * | 2023-04-21 | 2023-09-05 | 国网宁夏电力有限公司 | Electric heating electricity consumption fluctuation analysis system and method |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109063902A (en) * | 2018-07-17 | 2018-12-21 | 广东工业大学 | A kind of short-term load forecasting method, device, equipment and storage medium |
-
2018
- 2018-12-24 CN CN201811584211.9A patent/CN109636059A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109063902A (en) * | 2018-07-17 | 2018-12-21 | 广东工业大学 | A kind of short-term load forecasting method, device, equipment and storage medium |
Non-Patent Citations (2)
Title |
---|
李香龙等: "基于EEMD-BP神经网络的含电采暖的配电变压器短期负荷预测", 《电测与仪表》 * |
梁智等: "基于VMD与PSO优化深度信念网络的短期负荷预测", 《电网技术》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111142049A (en) * | 2020-01-16 | 2020-05-12 | 合肥工业大学 | Intelligent transformer fault diagnosis method based on edge cloud cooperation mechanism |
CN111142049B (en) * | 2020-01-16 | 2020-12-29 | 合肥工业大学 | Intelligent transformer fault diagnosis method based on edge cloud cooperation mechanism |
CN111581883A (en) * | 2020-05-09 | 2020-08-25 | 国网上海市电力公司 | Method for calculating and predicting load on automation device |
CN112564085A (en) * | 2020-10-22 | 2021-03-26 | 国网山东省电力公司济宁供电公司 | Method and system for predicting maximum power load of electric heating distribution transformer |
CN112564085B (en) * | 2020-10-22 | 2023-04-18 | 国网山东省电力公司济宁供电公司 | Method and system for predicting maximum power load of electric heating distribution transformer |
CN112488404A (en) * | 2020-12-07 | 2021-03-12 | 广西电网有限责任公司电力科学研究院 | Multithreading efficient prediction method and system for large-scale power load of power distribution network |
CN115343959A (en) * | 2022-09-23 | 2022-11-15 | 国网北京市电力公司 | Electric heating load self-adaptive control method, device, equipment and medium |
CN115343959B (en) * | 2022-09-23 | 2024-04-16 | 国网北京市电力公司 | Self-adaptive control method, device, equipment and medium for electric heating load |
CN116701481A (en) * | 2023-04-21 | 2023-09-05 | 国网宁夏电力有限公司 | Electric heating electricity consumption fluctuation analysis system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109636059A (en) | Electric heating distribution transformer load forecasting method and device | |
Heydari et al. | A novel composite neural network based method for wind and solar power forecasting in microgrids | |
Wang et al. | Deterministic and probabilistic forecasting of photovoltaic power based on deep convolutional neural network | |
Li et al. | Prediction for tourism flow based on LSTM neural network | |
Ke et al. | Short-term electrical load forecasting method based on stacked auto-encoding and GRU neural network | |
Huang et al. | One‐day‐ahead hourly forecasting for photovoltaic power generation using an intelligent method with weather‐based forecasting models | |
CN107679557A (en) | Driving model training method, driver's recognition methods, device, equipment and medium | |
Du et al. | Power load forecasting using BiLSTM-attention | |
CN106251001A (en) | A kind of based on the photovoltaic power Forecasting Methodology improving fuzzy clustering algorithm | |
Dou et al. | Hybrid model for renewable energy and loads prediction based on data mining and variational mode decomposition | |
Senthil Kumar | Improved prediction of wind speed using machine learning. | |
CN114970362B (en) | Power grid load scheduling prediction method and system under multi-energy structure | |
CN109816177A (en) | A kind of Load aggregation quotient short-term load forecasting method, device and equipment | |
CN109508826A (en) | The schedulable capacity prediction methods of electric car cluster of decision tree are promoted based on gradient | |
Eseye et al. | Short-term wind power forecasting using a double-stage hierarchical hybrid GA-ANN approach | |
Samet et al. | Evaluation of neural network-based methodologies for wind speed forecasting | |
CN112329997A (en) | Power demand load prediction method and system, electronic device, and storage medium | |
CN116644970A (en) | Photovoltaic power prediction method based on VMD decomposition and lamination deep learning | |
CN115603446A (en) | Power distribution station area operation monitoring system based on convolution neural network and cloud edge synergistic effect | |
Zanjani et al. | Short-term Load Forecasting using the Combined Method of Wavelet Transform and Neural Networks Tuned by the Gray Wolf Optimization Algorithm | |
Malmir et al. | Optimization of data mining with evolutionary algorithms for cloud computing application | |
Zhai et al. | Combining PSO-SVR and Random Forest Based Feature Selection for Day-ahead Peak Load Forecasting. | |
Kumar | Improved Prediction of Wind Speed using Machine Learning. | |
CN116632834A (en) | Short-term power load prediction method based on SSA-BiGRU-Attention | |
KEREM et al. | Multi-step forward forecasting of electrical power generation in lignite-fired thermal power plant |
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 |
Application publication date: 20190416 |
|
RJ01 | Rejection of invention patent application after publication |