CN102930354B - A kind of community power consumption prediction method and device - Google Patents
A kind of community power consumption prediction method and device Download PDFInfo
- Publication number
- CN102930354B CN102930354B CN201210439494.4A CN201210439494A CN102930354B CN 102930354 B CN102930354 B CN 102930354B CN 201210439494 A CN201210439494 A CN 201210439494A CN 102930354 B CN102930354 B CN 102930354B
- Authority
- CN
- China
- Prior art keywords
- power consumption
- history
- data
- electricity
- community
- 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 44
- 230000005611 electricity Effects 0.000 claims abstract description 140
- 238000004458 analytical method Methods 0.000 claims description 30
- 238000003062 neural network model Methods 0.000 claims description 21
- 238000010276 construction Methods 0.000 claims description 18
- 238000010586 diagram Methods 0.000 description 13
- 210000002569 neuron Anatomy 0.000 description 12
- 210000004027 cell Anatomy 0.000 description 9
- 230000002354 daily effect Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 7
- 210000005036 nerve Anatomy 0.000 description 6
- 230000003203 everyday effect Effects 0.000 description 4
- 230000007935 neutral effect Effects 0.000 description 4
- 230000006399 behavior Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000001932 seasonal effect Effects 0.000 description 2
- 238000004378 air conditioning Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 210000000225 synapse Anatomy 0.000 description 1
Abstract
The invention discloses a kind of community power consumption prediction method and device.Wherein method includes: obtain the history power consumption data of described community；Obtain the historical temperature data that described power consumption is corresponding in real time；According to described historical temperature data and described history power consumption data, build electricity demand forecasting model；Obtain the temperature data in following certain time；Described community power consumption in future is predicted according to the temperature data in described following certain time and described forecast model.The present invention is by utilizing historical temperature data and history power consumption data construct electricity demand forecasting model, it is achieved that according to temperature prediction community power consumption.
Description
Technical field
The present invention relates to, with electrical domain, be specifically related to a kind of community power consumption prediction method and device.
Background technology
Along with the development of society and improving constantly of living standards of the people, increasing to the demand of electric energy.By
Property while electrical energy production and consumption, the production of the electric energy that needs to make rational planning for, it just can be made to meet consumption.
The production of main electric energy of making rational planning for by the way of following electricity consumption is predicted in prior art.
At present, power consumption prediction is mainly by considering that the impact of power consumption is come by electrovalence policy and electricity consumption period
Carry out.But in real life, the electricity consumption of electricity consumption behavior especially residential quarter can be subject to a great extent
Temperature, season etc. factor impact.Such as, when temperature drift or on the low side time, community can because use air-conditioning and
Power consumption is made to increase.The power consumption in summer also will be typically higher than the power consumption in spring.
Therefore, a kind of power consumption prediction method considering the factor impacts such as temperature it is badly in need of at present.
Summary of the invention
In view of this, it is an object of the invention to provide a kind of power consumption prediction method based on temperature impact.
Embodiments providing a kind of community power consumption prediction method, the method specifically includes:
Obtain the history power consumption data of described community；
Obtain the historical temperature data that described power consumption is corresponding in real time；
According to described historical temperature data and described history power consumption data, build electricity demand forecasting model；
Obtain the temperature data in following certain time；
Described community future is predicted according to the temperature data in described following certain time and described forecast model
Power consumption.
Preferably, described method also includes:
Store described community power consumption in future；
Obtain the described community actual power consumption within described following certain time；
According to described actual power consumption, described following power consumption is carried out error analysis, obtain error analysis knot
Really；
Continue to predict described community power consumption according to described error analysis result.
Preferably, described method also includes:
Obtain the history season information that described power consumption is corresponding in real time；
Described according to described historical temperature data and described history power consumption data, build electricity demand forecasting mould
Type includes:
According to described historical temperature data, described history season information and described history power consumption data construct
Electricity demand forecasting model；
Obtain the season information in following certain time；
Described predict described community according to the temperature data in described following certain time and described forecast model
Following power consumption includes:
According to the temperature data in described following certain time, season information in following certain time and described
Forecast model predicts described community power consumption in future.
Preferably, described community power consumption in future includes: described community daily power consumption in future and/or described little
District's peak value in future power consumption.
Preferably, described method also includes:
Obtain the history electricity price data the most corresponding with described history power consumption data and history electricity consumption period；
Described according to described historical temperature data and described history power consumption data, build electricity demand forecasting mould
Type includes:
According to described historical temperature data, described history electricity price data, described history electricity consumption period and described
History power consumption data, build electricity demand forecasting model；
Obtain the electricity price data in following certain time and determine the following electricity consumption period；
Described predict described community according to the temperature data in described following certain time and described forecast model
Following power consumption includes:
According to the temperature data in described following certain time, the electricity price data in following certain time, future
Electricity consumption period and described forecast model predict described community power consumption in future
Preferably, described according to described historical temperature data and described history power consumption data, build electricity consumption
Amount forecast model includes:
With described historical temperature data for input parameter, with described history power consumption data as output parameter, structure
Build neural network model.
The embodiment of the invention also discloses a kind of community power consumption prediction device, described device includes:
History power consumption unit, for obtaining the history power consumption data of described community；
Historical temperature unit, for obtaining the historical temperature data that described power consumption is corresponding in real time；
Model construction unit, is connected with described history power consumption unit and historical temperature unit, for according to institute
State historical temperature data and described history power consumption data, build electricity demand forecasting model；
Temperature unit, the temperature data in obtaining following certain time；
Predicting unit, is connected with described model construction unit and described temperature unit, for according to described future
Temperature data and described forecast model in certain time predict described community power consumption in future.
Preferably, described device also includes:
Memory element, is connected with described predicting unit, is used for storing described community power consumption in future；
Actual power consumption unit, for obtaining the actual electricity consumption within described following certain time of the described community
Amount；
Error analysis unit, is connected with described memory element and described actual power consumption unit, for according to institute
State actual power consumption and described following power consumption is carried out error analysis, obtain error analysis result；
Described predicting unit, is connected with described error analysis unit, specifically for tying according to described error analysis
Fruit continues to predict described community power consumption in future.
Preferably, described device also includes:
History season information unit, for obtaining the history season information that described power consumption is corresponding in real time；
Described model construction unit, is connected, specifically for going through described in basis with described history season information unit
History temperature data, described history season information and described history power consumption data construct electricity demand forecasting mould
Type；
Season information unit, the season information in obtaining following certain time；
Described predicting unit, is connected with described season information unit, specifically for according to a described following timing
Interior temperature data, the season information in following certain time and described forecast model predict described community not
Carry out power consumption.
Preferably, described community power consumption in future includes: described community daily power consumption in future and/or described little
District's peak value in future power consumption.
Preferably, described device also includes:
History electricity price data cell, for obtaining the history electricity price the most corresponding with described history power consumption data
Data；
Segment unit during history electricity consumption, for obtaining the history electricity price the most corresponding with described history power consumption data
Data；
Described model construction unit, is connected with segment unit when described history electricity price data cell and history electricity consumption,
Specifically for according to described historical temperature data, described history electricity price data, the described history electricity consumption period and
Described history power consumption data, build electricity demand forecasting model；
Electricity price data cell, the electricity price data in obtaining following certain time；
Segment unit during electricity consumption, is used for determining the following electricity consumption period；
Described predicting unit, is connected with segment unit when described electricity price data cell and described electricity consumption, specifically for
According to the temperature data in described following certain time, the electricity price data in following certain time, following electricity consumption
Period and described forecast model predict described community power consumption in future.
Preferably, described model construction unit, specifically for described historical temperature data for input parameter,
With described history power consumption data as output parameter, build neural network model.
Compared with the existing technology, beneficial effects of the present invention:
The present invention builds power consumption by utilizing cell history power consumption data and corresponding historical temperature data
Forecast model, and the temperature data obtaining following certain time is predicted, in this process, it is contemplated that temperature
The impact of Du Dui community electricity consumption, it is achieved that based on temperature prediction community power consumption.
Further, in a preferred embodiment of the invention, season information and electricity price, electricity consumption have been considered
The period impact on electricity consumption, hinge structure, to the consideration of electricity consumption influence factor more comprehensively, thus improves
The accuracy of electricity demand forecasting.
Accompanying drawing explanation
Fig. 1 is the embodiment of the present invention 1 method flow diagram；
Fig. 2 is artificial neuron's model schematic in the embodiment of the present invention；
Fig. 3 is single hidden layer BP network architecture schematic diagram in the embodiment of the present invention；
Fig. 4 is the embodiment of the present invention 2 method flow diagram；
Fig. 5 is the embodiment of the present invention 3 method flow diagram；
Fig. 6 is for the embodiment of the present invention 4 method flow diagram；
Fig. 7 is the following 7 days daily power consumption chart schematic diagrams in embodiment of the present invention small area；
Fig. 8 is the following peak value power consumption bar diagram in 7 days in embodiment of the present invention community；
Fig. 9 is the embodiment of the present invention 5 structure drawing of device.
Detailed description of the invention
In order to make those skilled in the art be more fully understood that the scheme of the embodiment of the present invention, below in conjunction with the accompanying drawings
With embodiment, the embodiment of the present invention is described in further detail.
The embodiment of the present invention 1 provides a kind of community power consumption prediction method, sees Fig. 1, and the method specifically includes:
S11, obtain the history power consumption data of described community.
Concrete, the history of this community by arranging peripheral interface, can be directly read from the ammeter of residential quarter
Power consumption data, and record the time that these power consumption data are corresponding, the history power consumption data that then will obtain
It is stored in data base with the corresponding time.
S12, obtain the historical temperature data that described power consumption is corresponding in real time.
With obtain history power consumption data be similar to, in the present invention, can by arrange peripheral interface as with meteorology
Office arranges interface, obtains the historical temperature data corresponding with history power consumption data.Concrete, can basis
The power consumption data time before recorded obtains corresponding historical temperature data.
S13, according to described historical temperature data and described history power consumption data, build electricity demand forecasting
Model.
Prior art has the multiple model being predicted according to historical data, such as neural network model.This portion
Point content will be described below.
S14, the temperature data obtained in following certain time.
It is similar to step S12, can be by the interface with weather bureau, it is thus achieved that following certain time such as the temperature in 7 days
Degrees of data.
S15, predict described community according to the temperature data in described following certain time and described forecast model
Following power consumption.
In above-mentioned steps S13, refer to neural network model, in the present invention with artificial nerve network model and
As a example by BP error backward propagation method model, the structure of forecast model is described in detail.
Artificial neural network (Artificial Neural Networks, ANN) model is that a kind of application is similar
Structure in cerebral nerve synapse connection carries out the mathematical model of information processing, is to be led to by some simple elements
Cross the networking of large-scale parallel coupling configuration.Neuron is the basic processing unit of neutral net, and it is one years old
As be the nonlinear device of multiple input single output, typical artificial neuron meta-model is as shown in Figure 2.
The input/output relation of neuron can be described as:
Wherein xj, j=1,2 ..., n is the input signal of neuron, θiFor threshold value, wjiRepresent from neuron j to
The connection weights of neuron i, f be activation primitive (also known as transmission function), it must continuously differentiable, load
Neuron activation functions conventional in prediction is S type function.
Artificial nerve network model has the strongest self study and complicated nonlinear function approximation ability, the suitableeest
Together in load forecast problem, it it is one of practical Forecasting Methodology of getting the nod in the world.It highlights excellent
Point is that a large amount of unstructuredness, non-precision rule are had adaptation function, has imformation memory, independently learns
The feature that habit, knowledge reasoning and optimization calculate, particularly its self study and adaptation function are that traditional algorithm is hoped
Dirt not and.
When building artificial nerve network model, it is thus necessary to determine that input and output parameter.In the present invention,
Need, with historical temperature data for input parameter, to build artificial neuron with history power consumption data for output parameter
Network model.When follow-up being predicted, need the temperature data of input following certain time, the most manually
The parameter of neural network model output is the power consumption data of following certain time.
BP (BackPropagation) neutral net is also referred to as error backward propagation method, and it is by non-thread
Property converter unit composition feedforward network.BP neutral net is by input layer, output layer and hidden layer (hidden layer
Can be one or more layers) constitute, it is not connected to between node layer, the output of every node layer only affects next
The input of node layer, sees accompanying drawing 3, for single hidden layer BP network architecture schematic diagram.
The basic thought of BP algorithm is: the study of whole network is inverse by the forward-propagating of input signal and error
To propagating two process compositions.Forward-propagating process refers to that input signal is inputted by input layer, through the power of network
After the transmission function effect of value, threshold value and neuron, export from output layer.If output valve and expected value it
Between error more than ormal weight, then be modified, proceed to error-duration model and broadcast the stage, i.e. error passes through hidden layer
Successively return to input layer, and by error by " sharing " to each layer neuron by " gradient decline " principle,
Thus obtain the error signal of each layer neuron, as the foundation of amendment weights.Above two processes are the most
Secondary carry out.The process that weights are constantly revised, namely the training process of network.This circulation is performed until
Till the output error of network is reduced to permissible value or arrives the frequency of training set.
In the present invention, the process building BP neural network model is similar with building artificial nerve network model, this
Place is no longer discussed in detail.
For improving the accuracy of prediction further, in the embodiment of the present invention 2, the result doped can be entered
Row error analysis also utilizes error analysis modified result forecast model.Detailed process is as shown in Figure 4:
S21, storage above-described embodiment 1 predict the following power consumption obtained.
S22, obtain the described community actual power consumption within described following certain time.
S23, according to described actual power consumption, the following power consumption that obtains of prediction is carried out error analysis, and raw
Become error analysis result.
S24, according to above-mentioned error analysis result continue predict described community following power consumption.Concrete,
Need the temperature data of certain time, forecast model and described error analysis prediction of result according to follow-up acquisition
Described community power consumption in future.
Mentioning in background technology, different seasons, the power consumption of community also can be different.Reality in the present invention
Execute in example 3, it is also contemplated that the impact of seasonal factor.Seeing Fig. 5, it specifically predicts that process is as follows:
S31, obtain the history power consumption data of described community.
S32, obtain the historical temperature data that described power consumption is corresponding in real time.
S33, obtain the history season information that described power consumption is corresponding in real time.
S34, according to described historical temperature data, described history season information and described history power consumption number
According to building electricity demand forecasting model.
S35, the temperature data obtained in following certain time.
S36, the season information obtained in following certain time.
S37, according to the temperature data in described following certain time, season information in following certain time
Described community power consumption in future is predicted with described forecast model.
It should be noted that in this embodiment, it was predicted that the preferred above-mentioned neural network model of model.This time step
Rapid S34 is specially with historical temperature data, history season information for input parameter, with history power consumption data
Neural network model is built for output parameter.In corresponding step S37, need to be with the temperature in following certain time
Degrees of data, season information in following certain time do input parameter, and its output parameter is the future of community
Power consumption.
The prediction of power consumption is used for the production of follow-up electric energy of making rational planning for.For ensureing the consumption of electric energy, electric energy
Production not only need to ensure the electricity consumption of community every day, in addition it is also necessary to ensure the community use when electricity consumption summit
Electricity.Therefore, in a preferred embodiment of the invention, need to dope the daily power consumption of described community and little
The peak value power consumption in district.Wherein peak value power consumption is also needed to record the time of its correspondence.
In real life, community electricity consumption behavior is also affected by electricity price and electricity consumption period.Such as evening 8
O'clock typically more than power consumption after morning by 10 o'clock, during electricity price height, power consumption also can significantly reduce.For entering one
Step improves the accuracy of electricity demand forecasting, in the embodiment of the present invention 4, and can be in conjunction with electricity price data and electricity consumption
Community power consumption is predicted by the period.Seeing Fig. 6, this embodiment specifically includes:
S41, obtain the history power consumption data of described community.
S42, obtain described power consumption historical temperature data corresponding in real time, history electricity price data and history and use
The electricity period.
S43, according to described historical temperature data, described history electricity price data, the described history electricity consumption period with
And described history power consumption data, build electricity demand forecasting model.
S44, the temperature data in acquisition following certain time, the electricity price data in following certain time are the most really
The fixed following electricity consumption period.
S45, according to the temperature data in described following certain time, electricity price data in following certain time,
Following electricity consumption period and described forecast model predict described community power consumption in future.
When this electricity demand forecasting model is neural network model, can be with historical temperature data, history electricity price number
It is input parameter according to, history electricity consumption period, builds neutral net mould with history power consumption data for output parameter
Type.
It should be noted that in a preferred embodiment of the invention, above-described embodiment 4 may also be combined with season information
Impact, community power consumption in future is predicted.
Check contrast for the convenience of the user, after predicting community power consumption in future, drawing software can be called and carry out
Draw, the following power consumption of community is patterned displaying.Concrete, available chart represents community future
Peak value power consumption in certain time and peak value power consumption moment, represent that community is certain in future with bar diagram
Daily power consumption in time, wherein, bar diagram highly represents the daily power consumption in different sky.As it is shown in fig. 7,
For the peak value power consumption of certain community every day of following 7 days graphically illustrated and peak value power consumption time.Fig. 8
The daily power consumption of certain community following 7 day every day for representing with bar diagram.
Corresponding said method embodiment, the embodiment of the present invention 5 additionally provides a kind of community power consumption prediction device,
Seeing Fig. 9, this device specifically includes:
History power consumption unit 11, for obtaining the history power consumption data of described community.
Concrete, history power consumption unit 11 can pass through peripheral interface, directly reads from the ammeter of residential quarter
The history power consumption data of this community, and record the time that these power consumption data are corresponding, then by going through of obtaining
History power consumption data and corresponding time are stored in data base.
Historical temperature unit 12, for obtaining the historical temperature data that described power consumption is corresponding in real time.
Being similar to history power consumption unit 11, in the present invention, historical temperature unit 12 can be by arranging periphery
Interface, as arranged interface with weather bureau, obtains the historical temperature data corresponding with history power consumption data.Tool
Body, corresponding historical temperature data can be obtained according to the power consumption data time recorded before.
Model construction unit 13, is connected, for basis with described history power consumption unit and historical temperature unit
Described historical temperature data and described history power consumption data, build electricity demand forecasting model.
Prior art has the multiple model being predicted according to historical data, such as neural network model.This portion
Point content will be described below.
Temperature unit 14, the temperature data in obtaining following certain time.
Being similar to historical temperature unit 12, temperature unit 14 can be by the interface with weather bureau, it is thus achieved that following one
Temperature data in fixing time such as 7 days futures.
Predicting unit 15, is connected with described model construction unit and described temperature unit, for according to described not
Carry out the temperature data in certain time and described forecast model predicts described community power consumption in future.
In the specific embodiment of the invention, model construction unit 13 can be used according to historical temperature data and history
Electric quantity data, builds neural network model such as artificial nerve network model, BP neural network model.Relevant god
The description of embodiment of the method part is can be found in through the introduction of network model.
When neural network model to be built, model construction unit 13, specifically for historical temperature data being
Input parameter, with history power consumption data as output parameter, builds neural network model.
For improving the accuracy of prediction further, in embodiments of the present invention, it is also possible to the result of prediction is entered
Row error analysis also utilizes error analysis modified result forecast model.To this, relative above-described embodiment, described
Device also includes:
Memory element, is connected with described predicting unit, is used for storing described community power consumption in future.
Actual power consumption unit, for obtaining the actual electricity consumption within described following certain time of the described community
Amount.
Error analysis unit, is connected with described memory element and described actual power consumption unit, for according to institute
State actual power consumption and described following power consumption is carried out error analysis, obtain error analysis result.
Described predicting unit, is connected with described error analysis unit, specifically for tying according to described error analysis
Fruit continues to predict described community power consumption in future.Concrete, it was predicted that it is certain that unit needs according to follow-up acquisition
Community power consumption in future described in the temperature data of time, forecast model and described error analysis prediction of result.
As mentioned in the background art, in different seasons, the power consumption of community also can be different.The present invention's
In another embodiment, it is also contemplated that the impact of seasonal factor.Corresponding, said apparatus also includes:
History season information unit, for obtaining the history season information that described power consumption is corresponding in real time.
Described model construction unit, is connected, specifically for going through described in basis with described history season information unit
History temperature data, described history season information and described history power consumption data construct electricity demand forecasting mould
Type.
Season information unit, the season information in obtaining following certain time.
Described predicting unit, is connected with described season information unit, specifically for according to a described following timing
Interior temperature data, the season information in following certain time and described forecast model predict described community not
Carry out power consumption.
It should be noted that in this embodiment, it was predicted that the preferred above-mentioned neural network model of model.Now mould
Type construction unit is specifically for historical temperature data, history season information for input parameter, with history electricity consumption
Amount data are that output parameter builds neural network model.Corresponding predicting unit, specifically for future certain
Temperature data in time, the season information in following certain time do input parameter, and its output parameter is
The following power consumption of community.
The prediction of power consumption is used for the production of follow-up electric energy of making rational planning for.For ensureing the consumption of electric energy, electric energy
Production not only need to ensure the electricity consumption of community every day, in addition it is also necessary to ensure the community use when electricity consumption summit
Electricity.Therefore, in a preferred embodiment of the invention, it was predicted that unit needs to dope the day electricity consumption of described community
Amount and the peak value power consumption of community.Wherein peak value power consumption is also needed to record the time of its correspondence.
In real life, community electricity consumption behavior is also affected by electricity price and electricity consumption period.Such as evening
8 o'clock to 10 o'clock more than power consumption after morning, and during electricity price height, power consumption also can significantly reduce.For entering
The accuracy of one step raising electricity demand forecasting, in embodiments of the present invention, can be in conjunction with electricity price data and electricity consumption
Community power consumption is predicted by the period.Now, relative said apparatus, described device also includes:
History electricity price data cell, for obtaining the history electricity price the most corresponding with described history power consumption data
Data.
Segment unit during history electricity consumption, for obtaining the history electricity price the most corresponding with described history power consumption data
Data.
Described model construction unit, is connected with segment unit when described history electricity price data cell and history electricity consumption,
Specifically for according to described historical temperature data, described history electricity price data, the described history electricity consumption period and
Described history power consumption data, build electricity demand forecasting model.
Electricity price data cell, the electricity price data in obtaining following certain time.
Segment unit during electricity consumption, is used for determining the following electricity consumption period.
Described predicting unit, is connected with segment unit when described electricity price data cell and described electricity consumption, specifically for
According to the temperature data in described following certain time, the electricity price data in following certain time, following electricity consumption
Period and described forecast model predict described community power consumption in future.
Wherein, model construction unit, it is also possible to historical temperature data, history electricity price data, history electricity consumption
Period is input parameter, builds neural network model with history power consumption data for output parameter.
It should be noted that in the present invention, above-described embodiment may also be combined with the impact of season information, to little
District's power consumption in future is predicted.
Check that contrast, described device also include graphical representation unit for the convenience of the user, specifically in prediction
After the power consumption in future of community, call drawing software and draw, the following power consumption of community is carried out figure
Change and show.
Concrete, graphical representation unit available chart represents the peak value power consumption in the certain time in future of community
And the peak value power consumption moment, represent community daily power consumption within following certain time with bar diagram, wherein,
Bar diagram highly represents the daily power consumption in different sky.
It should be noted that said apparatus embodiment is corresponding with embodiment of the method, therefore to device section
Dividing and no longer describe in detail, relevant portion sees embodiment of the method.
Being described in detail the embodiment of the present invention above, detailed description of the invention used herein is to this
Bright being set forth, the explanation of above example is only intended to help to understand the method and apparatus of the present invention；With
Time, for one of ordinary skill in the art, according to the thought of the present invention, in detailed description of the invention and application
All will change in scope, in sum, this specification content should not be construed as limitation of the present invention.
Claims (10)
1. a community power consumption prediction method, it is characterised in that described method includes:
Obtain the history power consumption data of described community；
Obtain the historical temperature data that described power consumption is corresponding in real time；
According to described historical temperature data and described history power consumption data, build electricity demand forecasting model；
Obtain the temperature data in following certain time；
Described community future is predicted according to the temperature data in described following certain time and described forecast model
Power consumption；
Wherein, described electricity demand forecasting model is neural network model；
Described method also includes:
Obtain the history electricity price data the most corresponding with described history power consumption data and history electricity consumption period；
Described according to described historical temperature data and described history power consumption data, build electricity demand forecasting
Model includes:
According to described historical temperature data, described history electricity price data, described history electricity consumption period and institute
State history power consumption data, build electricity demand forecasting model；
Obtain the electricity price data in following certain time and determine the following electricity consumption period；
Described predict described community according to the temperature data in described following certain time and described forecast model
Following power consumption includes:
According to the temperature data in described following certain time, electricity price data in following certain time, not
Carry out the electricity consumption period and described forecast model predicts described community power consumption in future.
Method the most according to claim 1, it is characterised in that described method also includes:
Store described community power consumption in future；
Obtain the described community actual power consumption within described following certain time；
According to described actual power consumption, described following power consumption is carried out error analysis, obtain error analysis knot
Really；
Continue to predict described community power consumption according to described error analysis result.
Method the most according to claim 1, it is characterised in that described method also includes:
Obtain the history season information that described power consumption is corresponding in real time；
Described according to described historical temperature data and described history power consumption data, build electricity demand forecasting
Model includes:
According to described historical temperature data, described history season information and described history power consumption data structure
Build electricity demand forecasting model；
Obtain the season information in following certain time；
Described predict described community according to the temperature data in described following certain time and described forecast model
Following power consumption includes:
According to the temperature data in described following certain time, season information in following certain time and institute
State forecast model and predict described community power consumption in future.
Method the most according to claim 1, it is characterised in that described community power consumption in future includes:
Described community daily power consumption in future and/or described community peak value in future power consumption.
5. according to the method described in any one of Claims 1-4, it is characterised in that go through described in described basis
History temperature data and described history power consumption data, build electricity demand forecasting model and include:
With described historical temperature data for input parameter, with described history power consumption data as output parameter,
Build neural network model.
6. a community power consumption prediction device, it is characterised in that described device includes:
History power consumption unit, for obtaining the history power consumption data of described community；
Historical temperature unit, for obtaining the historical temperature data that described power consumption is corresponding in real time；
Model construction unit, is connected, for basis with described history power consumption unit and historical temperature unit
Described historical temperature data and described history power consumption data, build electricity demand forecasting model；
Temperature unit, the temperature data in obtaining following certain time；
Predicting unit, is connected with described model construction unit and described temperature unit, for according to described not
Carry out the temperature data in certain time and described forecast model predicts described community power consumption in future；
History electricity price data cell, for obtaining the history electricity the most corresponding with described history power consumption data
Valence mumber evidence；
Segment unit during history electricity consumption, for obtaining the history electricity the most corresponding with described history power consumption data
Valence mumber evidence；
Described model construction unit, is connected with segment unit when described history electricity price data cell and history electricity consumption,
Specifically for according to described historical temperature data, described history electricity price data, the described history electricity consumption period with
And described history power consumption data, build electricity demand forecasting model；
Electricity price data cell, the electricity price data in obtaining following certain time；
Segment unit during electricity consumption, is used for determining the following electricity consumption period；
Described predicting unit, is connected with segment unit when described electricity price data cell and described electricity consumption, specifically uses
According to the temperature data in described following certain time, the electricity price data in following certain time, future
Electricity consumption period and described forecast model predict described community power consumption in future；
Wherein, described electricity demand forecasting model is neural network model.
Device the most according to claim 6, it is characterised in that described device also includes:
Memory element, is connected with described predicting unit, is used for storing described community power consumption in future；
Actual power consumption unit, for obtaining the actual electricity consumption within described following certain time of the described community
Amount；
Error analysis unit, is connected, for basis with described memory element and described actual power consumption unit
Described actual power consumption carries out error analysis to described following power consumption, obtains error analysis result；
Described predicting unit, is connected with described error analysis unit, specifically for according to described error analysis
Result continues to predict described community power consumption in future.
Device the most according to claim 6, it is characterised in that described device also includes:
History season information unit, for obtaining the history season information that described power consumption is corresponding in real time；
Described model construction unit, is connected with described history season information unit, specifically for according to described
Historical temperature data, described history season information and described history power consumption data construct electricity demand forecasting
Model；
Season information unit, the season information in obtaining following certain time；
Described predicting unit, is connected with described season information unit, specifically for certain according to described future
Temperature data in time, the season information in following certain time and the prediction of described forecast model are described little
District's power consumption in future.
Device the most according to claim 8, it is characterised in that described community power consumption in future includes:
Described community daily power consumption in future and/or described community peak value in future power consumption.
10. according to the device described in any one of claim 6 to 9, it is characterised in that described model construction list
Unit, specifically for described historical temperature data for input parameter, be defeated with described history power consumption data
Go out parameter, build neural network model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210439494.4A CN102930354B (en) | 2012-11-06 | 2012-11-06 | A kind of community power consumption prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210439494.4A CN102930354B (en) | 2012-11-06 | 2012-11-06 | A kind of community power consumption prediction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102930354A CN102930354A (en) | 2013-02-13 |
| CN102930354B true CN102930354B (en) | 2016-08-10 |
Family
ID=47645148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210439494.4A Active CN102930354B (en) | 2012-11-06 | 2012-11-06 | A kind of community power consumption prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102930354B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473613B (en) * | 2013-09-09 | 2017-01-11 | 武汉理工大学 | Landscape structure-surface temperature-electricity consumption coupling model and application thereof |
| TWI489725B (en) * | 2013-11-07 | 2015-06-21 | Inst Information Industry | Apparatus and method for creating a power consumption model and computer program product thereof |
| CN103854068B (en) * | 2013-12-06 | 2017-02-08 | 国家电网公司 | Method for forecasting residential quarter short-term loads |
| CN104751233A (en) * | 2013-12-27 | 2015-07-01 | 台达电子工业股份有限公司 | Contract capacity optimization system and optimization method |
| CN105703358B (en) * | 2014-07-11 | 2019-09-03 | 英科德技术股份有限公司 | Device, server, system and method for energy measurement |
| CN104317910A (en) * | 2014-10-27 | 2015-01-28 | 国家电网公司 | Data processing method and device |
| CN104346659A (en) * | 2014-11-26 | 2015-02-11 | 北京邮电大学 | Short-term power generation prediction method applied to high-concentration-ratio photovoltaic power generation system |
| CN106934497B (en) * | 2017-03-08 | 2021-04-16 | 青岛卓迅电子科技有限公司 | Intelligent community power consumption real-time prediction method and device based on deep learning |
| CN107330548A (en) * | 2017-06-16 | 2017-11-07 | 广东电网有限责任公司惠州供电局 | Residential electricity consumption Forecasting Methodology based on time series models |
| CN110210672A (en) * | 2019-06-04 | 2019-09-06 | 北京百度网讯科技有限公司 | The method and apparatus of electricity demand forecasting, electronic equipment, computer-readable medium |
| CN111523807B (en) * | 2020-04-24 | 2023-04-28 | 广西电网有限责任公司崇左供电局 | Electric energy substitution potential analysis method based on time sequence and neural network |
| CN112163709B (en) * | 2020-10-09 | 2024-04-09 | 珠海格力电器股份有限公司 | Method and device for electricity utilization promotion, storage medium, and electronic device |
| CN114456862B (en) * | 2020-11-04 | 2022-10-21 | 中国石油化工股份有限公司 | Optimization method for natural gas cryogenic denitrification pretreatment process parameters |
| CN112907064B (en) * | 2021-02-08 | 2024-04-02 | 国网安徽省电力有限公司蚌埠供电公司 | Electric quantity prediction method and device based on adaptive window, storage medium and terminal |
| CN112907062B (en) * | 2021-02-08 | 2024-04-02 | 国网安徽省电力有限公司蚌埠供电公司 | Power grid electric quantity prediction method, device, medium and terminal integrating temperature characteristics |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521674A (en) * | 2011-12-07 | 2012-06-27 | 福建省电力有限公司 | Computer generating method for short-term load forecasting in summer in power consumer electricity-utilization information acquisition system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| BRPI0606352A2 (en) * | 2005-01-18 | 2009-06-16 | Mc Energy Inc | method for calculating future energy consumption, method for calculating a future cost basis for energy consumption and method for budgeting future energy consumption for a property |
| CN102682349A (en) * | 2012-05-14 | 2012-09-19 | 云南电力试验研究院（集团）有限公司电力研究院 | Electricity consumption intelligent prediction system and method |
-
2012
- 2012-11-06 CN CN201210439494.4A patent/CN102930354B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521674A (en) * | 2011-12-07 | 2012-06-27 | 福建省电力有限公司 | Computer generating method for short-term load forecasting in summer in power consumer electricity-utilization information acquisition system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102930354A (en) | 2013-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102930354B (en) | A kind of community power consumption prediction method and device | |
| Kankal et al. | Modeling and forecasting of Turkey’s energy consumption using socio-economic and demographic variables | |
| De Marchi et al. | Combining participative and institutional approaches with multicriteria evaluation. An empirical study for water issues in Troina, Sicily | |
| CN102183621B (en) | Aquaculture dissolved oxygen concentration online forecasting method and system | |
| Xia et al. | Short, medium and long term load forecasting model and virtual load forecaster based on radial basis function neural networks | |
| Iskin et al. | Exploring renewable energy pricing with analytic network process—Comparing a developed and a developing economy | |
| CN105844371A (en) | Electricity customer short-term load demand forecasting method and device | |
| CN105335816A (en) | Electric power communication operation trend and business risk analyzing method based on deep learning | |
| CN107908638A (en) | The operation of power networks efficiency rating method and system matched somebody with somebody are excavated based on big data | |
| Pukšec et al. | Assessing the impact of energy saving measures on the future energy demand and related GHG (greenhouse gas) emission reduction of Croatia | |
| CN106951990A (en) | Electric load intelligent Forecasting and device | |
| CN107730041A (en) | Short-Term Load Forecasting Method based on improved genetic wavelet neural network | |
| CN104037943A (en) | Method and system for monitoring voltage and capable of improving power grid voltage quality | |
| CN104573857A (en) | Power grid load rate prediction method based on intelligent algorithm optimization and combination | |
| CN104834975A (en) | Power network load factor prediction method based on intelligent algorithm optimization combination | |
| Qiu et al. | Hierarchical multi-agent reinforcement learning for repair crews dispatch control towards multi-energy microgrid resilience | |
| Yilmaz | Real measure of a transmission line data with load fore-cast model for the future | |
| Hoekstra et al. | Agent-based model for the adoption and impact of electric vehicles in real neighbourhoods | |
| Pandey et al. | Applications of artificial intelligence in power system operation, control and planning: a review | |
| Şeker | Long term electricity load forecasting based on regional load model using optimization techniques: A case study | |
| Hao | Deep reinforcement learning for the optimization of building energy control and management | |
| Medykovskyy et al. | Development of a regional energy efficiency control system on the basis of intelligent components | |
| Wu et al. | A deep learning-based strategy to the energy management-advice for time-of-use rate of household electricity consumption | |
| Ramezani et al. | Short-term electric load forecasting using neural networks | |
| CN104036366A (en) | Method for evaluating power grid development supporting degree through electric power telecommunication network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |