CN106682999A - Electric power user baseline load calculating method and apparatus thereof - Google Patents
Electric power user baseline load calculating method and apparatus thereof Download PDFInfo
- Publication number
- CN106682999A CN106682999A CN201611033053.9A CN201611033053A CN106682999A CN 106682999 A CN106682999 A CN 106682999A CN 201611033053 A CN201611033053 A CN 201611033053A CN 106682999 A CN106682999 A CN 106682999A
- Authority
- CN
- China
- Prior art keywords
- load
- load curve
- baseline
- curve
- feature
- 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
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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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"
Abstract
The invention discloses an electric power user baseline load calculating method and an apparatus thereof. The method is characterized by according to a history day load curve set and a reference load curve set, forming several characteristic load curves; determining a key factor influencing a user baseline load and establishing a decision tree associated with the characteristic load curves and the key factor; and using the decision tree to carry out prediction calculating on the user baseline load, and feeding back a calculation result to each service system. A lot of data loading time is saved, accuracy of a baseline load calculating result is effectively increased, and an accurate data basis is provided for reasonably making and implementing electric power demand side management and demand response measures. Simultaneously, the apparatus is used for executing the method; an automation degree of electric power user baseline load prediction calculating is increased; baseline load prediction calculating efficiency is improved; and a problem that each service system is easy to frequently generate faults during a data interaction process is avoided.
Description
Technical field
The present invention relates to Power System and its Automation technical field, more particularly to a kind of power consumer baseline carry calculation
Method and device.
Background technology
Load characteristic analysis be intelligent grid research basis, according to load characteristic carry out user's baseline load it is pre- at last
The important decision foundation of power network expansion planning, management and running, power construction and load management, for ensure power grid security, economy,
High-quality operation is significant;Demand response is the abbreviation of electricity needs response, is referred to when wholesale power market price is raised
Or system reliability it is compromised when, the inductivity that supplier of electricity sends that power consumer receives reduce the direct compensation of load notify or
After person's power price rising signals, change its intrinsic custom power mode, the electricity consumption for reaching reduction or elapsing certain period is born
Lotus and respond supply of electric power, so as to ensure the stabilization of power grids, and suppress the acts and efforts for expediency that electricity price rises.
Traditional electric load characteristic analysis method mainly has two classes:One is by analysis of Influential Factors, i.e., in remaining variables
Keep it is constant on the premise of extract dependent variable one by one, qualitative or quantitative description its influence degree to independent variable;Two is by row
Industry classification analysises, i.e., first to every profession and trade or refine to being analyzed with electrical feature for certain type of user, qualitative or quantitative
To impact of all types of users to regional power grid electrical feature.
But in specific practice, a great problem that the baseline load prediction based on demand response is always present.Mesh
Front because load species is various, feature difference is big, and load variations randomness is strong, accurately calculates the baseline load difficulty of demand response
Greatly, the particularly single baseline load prediction accuracy for considering a certain factor is very low.And current baseline load prediction side
Method and system lack the automation means to baseline load prediction, cause baseline load predictive efficiency low, and each operation system exists
Easily frequently error during interaction data.
The content of the invention
The goal of the invention of the present invention is to provide a kind of power consumer baseline Load Calculation Method and device, existing to solve
In technology, because load species is various, feature difference is big, and load variations randomness is strong, accurately calculates the baseline negative of demand response
Lotus difficulty is big, and the particularly single baseline load prediction accuracy for considering a certain factor is very low, and, baseline load prediction side
Method and system lack the automation means to baseline load prediction, cause baseline load predictive efficiency low, and each operation system exists
The technical problem easily frequently malfunctioned during interaction data.
A kind of first aspect according to embodiments of the present invention, there is provided power consumer baseline Load Calculation Method, including:
Historical load data is obtained, history daily load curve collection is obtained, included in the history daily load curve collection
History daily load curve is corresponding with the data structure of the reference load curve included in the reference load curve set for prestoring;
According to the history daily load curve collection and the reference load curve set, several feature load curves are formed, will
The feature load curve is converted to and represented by CIM/OWL Ontologies, and several described feature load curves form feature load
Curve set;
It is determined that affecting the key factor of user's baseline load;
Set up the decision tree for associating the feature load curve and the key factor;
Using the decision tree, calculating is predicted to user's baseline load, result of calculation is fed back to into each business
System.
Further, it is described according to history daily load curve collection and the reference load curve set, form several feature loads
Curve is concretely comprised the following steps:
Judge whether the history daily load curve for being contrasted meets default similar standard with the reference load curve;
If meeting default similar standard, the history daily load curve is formed into several with the reference load curve
Feature load curve.
Further, whether the history daily load curve for judging to be contrasted meets pre- with the reference load curve
If similar standard is specially:
Calculating the history daily load curve concentrates whole history daily load curves every with the reference load curve set
Euclidean distance of the bar with reference to load curve;
Whether the Euclidean distance is judged less than or equal to default threshold value,
If the Euclidean distance is less than or equal to default threshold value, the history daily load curve is with described with reference to negative
Lotus curve meets default similar standard;
If the Euclidean distance is more than default threshold value, the history daily load curve and the reference load curve
It is unsatisfactory for default similar standard.
Further, if described meet default similar standard, by the history daily load curve and the reference load
Curve forms several feature load curves and is specially:
Calculate the load meansigma methodss of history daily load curve All Time point corresponding with the reference load curve, institute
State meansigma methodss and form feature load curve with corresponding All Time point;Array meets the history day of default similar standard
Load curve forms several feature load curves with the reference load curve.
Further, the step of determination affects the key factor of user's baseline load specifically includes:
Obtaining affects the factor of user's baseline load;
Extract the peak value and valley of the historical load data and calculate the average of the historical load data, form three
Data sequence;
With reference to the data sequence, using gray relative analysis method, calculate between the historical load data and the factor
Grey relational grade;
According to the grey relational grade, it is determined that affecting the key factor of user's baseline load.
Further, the foundation associates the feature load curve and the decision tree step of the key factor is specifically wrapped
Include:
According to the feature load curve, calculate such another characteristic load curve and refer to the GINI of the key factor
Number;
According to the GINI indexes, the decision tree of the association feature load curve model and the key factor is set up;
It is described calculating such another characteristic load curve be with the formula of the GINI indexes of the key factor:
Wherein, D be such another characteristic load curve set, m be key factor number, i be key factor sequence number, Pi tables
Show the probability that any feature load curve is affected by factor i in D, Pi is equal to the feature load curve affected by key factor i in D
Total number of the bar number divided by feature load curve in D.
Further, the utilization decision tree, specifically includes the step of be predicted calculating to user's baseline load:
By the decision tree, obtain feature load curve corresponding with baseline load to be calculated and belong to category load
The historical load data and key factor of curve;
Forecasting Methodology is selected to be predicted, the Forecasting Methodology includes linear regression, similartrend prediction and god
Jing neural network forecasts, and smoothing weights calculating is carried out, obtain baseline load result of calculation.
Further, methods described also includes:While said method is performed, also by the historical load data, described
History daily load curve, the feature load curve, the key factor, the decision tree and baseline carry calculation knot
Fruit stores.
Further, the utilization decision tree, before calculating is predicted to user's baseline load, also includes:
Start service interface, circular wait needs the instruction for carrying out baseline carry calculation and/or carrying out data transmission, described
Instruction is sent by each operation system;
The instruction is received, and confirmation is sent to each operation system by the service interface and received the instruction
Message.
Another aspect according to embodiments of the present invention, there is provided a kind of power consumer baseline load computing device, the dress
Put the power consumer baseline Load Calculation Method that the first aspect for performing embodiments of the invention is provided, it is characterised in that
Described device includes:
The memorizer of the reference load curve set that is stored with and the control process device being connected with the memorizer;
The control process device include electricity consumption feature modeling unit, key factor identification unit, decision tree set up unit and
Baseline load prediction and calculation unit;Wherein,
The electricity consumption feature modeling unit, for obtaining historical load data, obtains history daily load curve collection;It is described to go through
History daily load curve included in history daily load curve collection and the ginseng included in the reference load curve set for prestoring
Examine the data structure correspondence of load curve;It is additionally operable to, according to the history daily load curve collection and the reference load curve set,
Several feature load curves are formed, the feature load curve is converted to and is represented by CIM/OWL Ontologies, several described spies
Levy load curve and form feature load curve collection;
The key factor determining unit, for determining the key factor for affecting user's baseline load;
The decision tree sets up unit, for setting up the decision-making for associating the feature load curve and the key factor
Tree;
The baseline load prediction and calculation unit, for using the decision tree, carrying out to user's baseline load pre-
Survey and calculate, baseline load result of calculation is fed back to into each operation system.
From above technical scheme, power consumer baseline Load Calculation Method provided in an embodiment of the present invention and device,
Methods described obtains history daily load curve collection by obtaining historical load data, and the history daily load curve is concentrated and wrapped
The data structure of the history daily load curve for containing and the reference load curve included in the reference load curve set for prestoring
Correspondence;According to the history daily load curve collection and the reference load curve set, several feature load curves are formed, will be described
Feature load curve is converted to and represented by CIM/OWL Ontologies, and several described feature load curves form feature load curve
Collection;And the key factor for affecting user's baseline load is determined, associate the feature load curve and the key so as to set up
The decision tree of factor;Using the decision tree, the feature belonging to the quick baseline load for determining that power consumer is to be calculated is born
The classification of lotus curve, and the historical load data of the power consumer, save mass data and are loaded into the time, effectively improve baseline negative
The accuracy of lotus result of calculation, accurate data are provided for rational with enforcement demand Side Management and demand response measure
Foundation, meanwhile, described device is used to perform methods described, improves the automatization calculated for the load prediction of power consumer baseline
Degree, improves baseline load prediction computational efficiency, it is to avoid what each operation system easily frequently malfunctioned during interaction data asks
Topic.
Description of the drawings
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to institute in embodiment
The accompanying drawing that needs are used is briefly described, it should be apparent that, drawings in the following description are only some enforcements of the present invention
Example, for those of ordinary skill in the art, on the premise of not paying creative work, can be being obtained according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is to be preferable to carry out a kind of power consumer baseline Load Calculation Method flow chart for exemplifying according to one;
Fig. 2 is to be preferable to carry out a kind of power consumer baseline load computing device structure block diagram for exemplifying according to one.
Wherein, 1- memorizeies, 2- control process devices, 21- electricity consumption feature modeling units, 22- key factor identification units,
23- decision trees set up unit, 24- baseline load prediction and calculation units.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Whole description, it is clear that described embodiment is only a part of embodiment of the invention, rather than the embodiment of whole.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
First aspect according to embodiments of the present invention, there is provided a kind of power consumer baseline Load Calculation Method, refering to figure
1, it is the flow chart of the power consumer baseline Load Calculation Method, from figure 1 it appears that methods described includes:
Step S1, acquisition historical load data, obtain history daily load curve collection, and the history daily load curve concentrates institute
Comprising the data of history daily load curve and the reference load curve included in the reference load curve set for prestoring tie
Structure correspondence;
Wherein, the historical load data of a certain power consumer refers to the user before current point in time in limited period
Power load data, the terminal of the limited period before the current point in time includes but is not limited to current point in time, described to work as
The unit of front time point can be hour, or one day, or specific current time;For example, it is assumed that when current
Between point be on January 10th, 2016, then the load data of the power consumer can be from January 9 to the previous moon in 2016 to
Power load data in the period on December 9th, 2015;The length in the period can be arranged voluntarily as needed;In addition,
The history daily load curve is to be described to form according to the coordinate of power load and its corresponding time point composition, can be more
Intuitively show power load Changing Pattern in certain period of the user before current point in time.
It should be noted that history daily load curve included in the history daily load curve collection with prestore
The data structure correspondence of the reference load curve included in reference load curve set, refers to the time point that two curves are included
Unanimously.
Step S2, according to the history daily load curve collection and the reference load curve set, form several feature loads
Curve, the feature load curve is converted to by CIM (Common Information Model, Common Information Model)/OWL
(the Internet Ontology Language, Ontology of Web Language) Ontology represents, several described feature load curve shapes
Into feature load curve collection;
For same power consumer, time point that the history daily load curve is included according to it is different and each not phase
Together, for different power consumers, the history daily load curve is included in addition due to the consumption habit difference between user
Time point is different and different therefore different classes of history daily load curve history of forming daily load curve collection;And it is described
Reference load curve set is pre-stored within memorizer, and each of which bar all has respective spy with reference to load curve
Levy.On this basis, the comparison history daily load curve collection and the reference load curve set described in step S2, according to comparing knot
Fruit shape is further included into several feature load curve refinement steps:
Whether the history daily load curve that judgement is computed meets default similar standard with the reference load curve;
If meeting default similar standard, the history daily load curve is formed into several with the reference load curve
Feature load curve.
Further, whether the history daily load curve for judging to be computed meets pre- with the reference load curve
If similar standard is specially:
Calculating the history daily load curve concentrates whole history daily load curves every with the reference load curve set
Euclidean distance of the bar with reference to load curve;Specifically, selected characteristic time point;The reference load curve set is traveled through, according to institute
Characteristic time point is stated, the history daily load curve is calculated one by one and is concentrated whole history daily load curves bent with the reference load
Line is concentrated per bar with reference to the Euclidean distance of load curve;Wherein, the characteristic time point, can sets itself as needed, example
It such as can be 8 points to late 10 points of evening prime time at night, reason is that during this period of time, most of resident sees
The custom of TV, therefore power load has certain rule to follow;For example when can also be certain of weekend or legal festivals and holidays
Between point etc..
Whether the Euclidean distance is judged less than or equal to default threshold value,
If the Euclidean distance is less than or equal to default threshold value, the history daily load curve is with described with reference to negative
Lotus curve meets default similar standard;
If the Euclidean distance is more than default threshold value, the history daily load curve and the reference load curve
It is unsatisfactory for default similar standard.
Further, if described meet default similar standard, by the history daily load curve and the reference load
Curve forms several feature load curves and is specially:
Calculate the load meansigma methodss of history daily load curve All Time point corresponding with the reference load curve, institute
State meansigma methodss and form feature load curve with corresponding All Time point;The similar history daily load curve of array and institute
State reference load curve and form several feature load curves.
Step S3, the key factor for determining impact user's baseline load;
The baseline load of power consumer can be affected by some specific factors, for example, weather conditions;If the day on the same day
Gas is thunderstorm, and ordinarily resident user can reduce the use of electrical appliance in family, to electrical appliance can be avoided to be subject to thunderbolt youngster to damage;
If weather is the cloudy day, that most of power consumer can all increase the use of illuminating lamp on daytime.For another example, time factor;
In 8 points to ten one points of the morning, the power load of most of resident all can be negative to ten one points of electricity consumption of evening less than at 8 points in evening
Lotus.For another example, date factor;The power load at legal festivals and holidays or weekend is higher than workaday power load;Meanwhile, equally
Reason, power consumer baseline load also suffers from the impact of the factors such as temperature, humidity and specific event.Based on this, invent
People first includes the factor that may all affect baseline load, recognizes the key factor for affecting user's baseline load again afterwards;
The refinement step that step S3 is further included is:
Obtaining affects the factor of user's baseline load;
Extract the peak value and valley of the historical load data and calculate the average of the historical load data, form three
Data sequence;
With reference to the data sequence, using gray relative analysis method, calculate between the historical load data and the factor
Grey relational grade;
According to the grey relational grade, it is determined that affecting the key factor of user's baseline load.
By determining the key factor for affecting baseline load, will be non-key in the factor that may all affect baseline load
Factor is rejected, the non-key factor the baseline load is not affected or the impact that produces it is negligible because
Element, key factor when baseline load prediction is carried out, only can be listed in the scope of basis for forecasting, save the same of calculating time
When, it is to avoid the interference of non-key factor, further improves for the prediction computational efficiency of baseline load.
Step S4, the decision tree for setting up the association feature load curve and the key factor;
The foundation associates the feature load curve and the decision tree step of the key factor is specifically included:
According to the feature load curve, calculate such another characteristic load curve and refer to the GINI of the key factor
Number;
According to the GINI indexes, the decision tree of the association feature load curve model and the key factor is set up;
It is described calculating such another characteristic load curve be with the formula of the GINI indexes of the key factor:
Wherein, D be such another characteristic load curve set, m be key factor number, i be key factor sequence number, Pi tables
Show the probability that any feature load curve is affected by factor i in D, Pi is equal to the feature load curve affected by key factor i in D
Total number of the bar number divided by feature load curve in D.
The feature load curve using the decision tree, belonging to the quick baseline load for determining that power consumer is to be calculated
Classification, and the historical load data of the power consumer saves mass data and is loaded into the time, effectively improves baseline carry calculation
As a result accuracy.
Step S5, using the decision tree, calculating is predicted to user's baseline load, result of calculation is fed back to
Each operation system.
The utilization decision tree described in step S5, calculating institute is predicted to user's baseline load further
Including refinement step be:
By the decision tree, obtain feature load curve corresponding with baseline load to be calculated and belong to category load
The historical load data and key factor of curve;
Forecasting Methodology is selected to be predicted, the Forecasting Methodology includes linear regression, similartrend prediction and god
Jing neural network forecasts, and smoothing weights calculating is carried out, obtain baseline load result of calculation.
Further, methods described also includes:While methods described is performed, by the historical load data, described go through
History daily load curve, the feature load curve, the key factor, the decision tree and the baseline load result of calculation
Store.
Further, methods described also includes, using the decision tree, to user's baseline load calculating is predicted
Before, starting service interface, circular wait needs the instruction for carrying out baseline carry calculation and/or carrying out data transmission, the instruction
Sent by each operation system;
The instruction is received, and confirmation is sent to each operation system by the service interface and received the instruction
Message.
Wherein, above-mentioned each operation system refers to the grid company operation system related to baseline load, such as marketing system
Or scheduling system, the application of baseline Load Calculation Method and device is determined by grid company.Above-mentioned circular wait is referred to
For each operation system, inquire about whether receive the instruction one by one, if receiving the instruction, correspondingly perform, if do not had
The instruction is received, is then inquired about for next operation system, all operation systems are queried after finishing, then again
Inquire about one by one.
In addition, feeding back to result of calculation before each operation system described in step S5, methods described also includes, will count
Calculate result and be converted into standard interface interactive information, the standard interface interactive information refers to and information institute is exchanged with other operation systems
The information exchange that needs meet is required.
From above technical scheme, power consumer baseline Load Calculation Method provided in an embodiment of the present invention, the side
Method obtains history daily load curve collection, going through included in the history daily load curve collection by obtaining historical load data
History daily load curve is corresponding with the data structure of the reference load curve included in the reference load curve set for prestoring;Root
According to the history daily load curve collection and the reference load curve set, several feature load curves are formed, the feature is born
Lotus Curve transform is to be represented by CIM/OWL Ontologies, and several described feature load curves form feature load curve collection;And really
Making affects the key factor of user's baseline load, so as to set up associate the feature load curve and the key factor certainly
Plan tree;Using the decision tree, the feature load curve belonging to the quick baseline load for determining that power consumer is to be calculated
Classification, and the historical load data of the power consumer, save mass data and are loaded into the time, effectively improve baseline carry calculation knot
The accuracy of fruit, accurate data foundation is provided for rational with enforcement demand Side Management and demand response measure.
Another aspect according to embodiments of the present invention, there is provided a kind of power consumer baseline load computing device, the dress
Put for performing baseline Load Calculation Method provided in an embodiment of the present invention, be the power consumer baseline load gauge refering to Fig. 2
The structural representation of device is calculated, from figure 2 it can be seen that described device includes:
The memorizer 1 of the reference load curve set that is stored with and the control process device 2 being connected with the memorizer 1;
It is single that the control process device 2 includes that electricity consumption feature modeling unit 21, key factor identification unit 22, decision tree is set up
Unit 23 and baseline load prediction and calculation unit 24;Wherein,
The electricity consumption feature modeling unit 21, for obtaining historical load data, obtains history daily load curve collection;It is described
History daily load curve included in history daily load curve collection with included in the reference load curve set for prestoring
The data structure correspondence of reference load curve;It is additionally operable to, according to the history daily load curve collection and the reference load curve
Collection, forms several feature load curves, the feature load curve is converted to and is represented by CIM/OWL Ontologies, several institutes
State feature load curve and form feature load curve collection;
The key factor identification unit 22, for determining the key factor for affecting user's baseline load;
The decision tree sets up unit 23, for setting up the decision-making for associating the feature load curve and the key factor
Tree;
The baseline load prediction and calculation unit 24, for using the decision tree, carrying out to user's baseline load
Prediction is calculated, and baseline load result of calculation is fed back to into each operation system;
It should be noted that the memorizer 1 be used for store the historical load data, the history daily load curve,
The feature load curve, the key factor, the decision tree and the result of calculation.
Preferably, the baseline load prediction and calculation unit 24 is being used to utilize the decision tree, to user's baseline
Load is predicted before calculating, is additionally operable to:
Start service interface, circular wait needs the instruction for carrying out baseline carry calculation and/or carrying out data transmission, described
Instruction is sent by each operation system;
The instruction is received, and confirmation is sent to each operation system by the service interface and received the instruction
Message;
The baseline load result of calculation is converted into feed back to each operation system after standard interface interactive information.
From above technical scheme, power consumer baseline load computing device provided in an embodiment of the present invention, for holding
Row power consumer baseline Load Calculation Method provided in an embodiment of the present invention, methods described is obtained by obtaining historical load data
To history daily load curve collection, the history daily load curve included in the history daily load curve collection and the ginseng for prestoring
Examine the data structure correspondence of the reference load curve included in load curve collection;According to the history daily load curve collection and institute
Reference load curve set is stated, several feature load curves are formed, the feature load curve is converted to by CIM/OWL bodies pair
As representing, several described feature load curves form feature load curve collection;And determine affect user's baseline load key
Factor, so as to set up the decision tree for associating the feature load curve and the key factor;It is quick true using the decision tree
Determine the classification of the feature load curve belonging to power consumer baseline load to be calculated, and the history of the power consumer is born
Lotus data, save mass data and are loaded into the time, effectively improve the accuracy of baseline load result of calculation, are rational and enforcement
Demand Side Management and demand response measure provide accurate data foundation, meanwhile, described device is used to perform methods described,
The automaticity that the load prediction of power consumer baseline is calculated is improve, baseline load prediction computational efficiency is improved, it is to avoid
The problem that each operation system easily frequently malfunctions during interaction data.
Those skilled in the art will readily occur to its of the present invention after considering description and putting into practice invention disclosed herein
Its embodiment.The application is intended to any modification of the present invention, purposes or adaptations, these modifications, purposes or
Person's adaptations follow the general principle of the present invention and including the undocumented common knowledge in the art of the present invention
Or conventional techniques.Description and embodiments are considered only as exemplary, and true scope and spirit of the invention are by following
Claim is pointed out.
It should be appreciated that the precision architecture for being described above and being shown in the drawings is the invention is not limited in, and
And can without departing from the scope carry out various modifications and changes.The scope of the present invention is only limited by appended claim.
Claims (10)
1. a kind of power consumer baseline Load Calculation Method, it is characterised in that methods described includes:
Historical load data is obtained, history daily load curve collection, the history included in the history daily load curve collection is obtained
Daily load curve is corresponding with the data structure of the reference load curve included in the reference load curve set for prestoring;
According to the history daily load curve collection and the reference load curve set, several feature load curves are formed, will be described
Feature load curve is converted to and represented by CIM/OWL Ontologies, and several described feature load curves form feature load curve
Collection;
It is determined that affecting the key factor of user's baseline load;
Set up the decision tree for associating the feature load curve and the key factor;
Using the decision tree, calculating is predicted to user's baseline load, result of calculation is fed back to into each operation system.
2. power consumer baseline Load Calculation Method according to claim 1, it is characterised in that described negative according to history day
Lotus curve set and the reference load curve set, form concretely comprising the following steps for several feature load curves:
Judge whether the history daily load curve for being contrasted meets default similar standard with the reference load curve;
If meeting default similar standard, the history daily load curve is formed into several features with the reference load curve
Load curve.
3. power consumer baseline Load Calculation Method according to claim 2, it is characterised in that what the judgement was contrasted
The history daily load curve is specially with whether the reference load curve meets default similar standard:
Calculate the history daily load curve and concentrate whole history daily load curves with every Radix Glehniae in the reference load curve set
Examine the Euclidean distance of load curve;
Judge the Euclidean distance whether less than or equal to default threshold value;
If the Euclidean distance is less than or equal to default threshold value, the history daily load curve is bent with the reference load
Line meets default similar standard;
If the Euclidean distance is more than default threshold value, the history daily load curve is discontented with the reference load curve
The default similar standard of foot.
4. power consumer baseline Load Calculation Method according to claim 2, it is characterised in that if described meet default
Similar standard, then form the history daily load curve and the reference load curve several feature load curves and be specially:
The load meansigma methodss of history daily load curve All Time point corresponding with the reference load curve are calculated, it is described flat
Average forms feature load curve with corresponding All Time point;Array meets the history daily load of default similar standard
Curve forms several feature load curves with the reference load curve.
5. power consumer baseline Load Calculation Method according to claim 1, it is characterised in that the determination affects user
The step of key factor of baseline load, specifically includes:
Obtaining affects the factor of user's baseline load;
Extract the peak value and valley of the historical load data and calculate the average of the historical load data, form three data
Sequence;
With reference to the data sequence, using gray relative analysis method, the ash between the historical load data and the factor is calculated
The color degree of association;
According to the grey relational grade, it is determined that affecting the key factor of user's baseline load.
6. power consumer baseline Load Calculation Method according to claim 1, it is characterised in that the foundation association is described
The decision tree step of feature load curve and the key factor is specifically included:
According to the feature load curve, the GINI indexes of such another characteristic load curve and the key factor are calculated;
According to the GINI indexes, the decision tree of the association feature load curve model and the key factor is set up;
It is described calculating such another characteristic load curve be with the formula of the GINI indexes of the key factor:
Wherein, D be such another characteristic load curve set, m be key factor number, i be key factor sequence number, PiIn representing D
The probability that any feature load curve is affected by factor i, PiEqual to the feature load curve bar number affected by key factor i in D
Divided by the total number of feature load curve in D.
7. power consumer baseline Load Calculation Method according to claim 6, it is characterised in that the utilization decision tree,
The step of being predicted calculating to user's baseline load specifically includes:
By the decision tree, obtain feature load curve corresponding with baseline load to be calculated and belong to category load curve
Historical load data and key factor;
Forecasting Methodology is selected to be predicted, the Forecasting Methodology includes linear regression, similartrend prediction and nerve net
Network is predicted, and carries out smoothing weights calculating, obtains baseline load result of calculation.
8. power consumer baseline Load Calculation Method according to claim 1, it is characterised in that methods described also includes:
While perform claim requires the method described in 1, by the historical load data, the history daily load curve, the spy
Levy load curve, the key factor, the decision tree and the baseline load result of calculation to store.
9. power consumer baseline Load Calculation Method according to claim 1, it is characterised in that the utilization decision tree,
Before calculating is predicted to user's baseline load, also include:
Start service interface, circular wait needs the instruction for carrying out baseline carry calculation and/or carrying out data transmission, the instruction
Sent by each operation system;
The instruction is received, and confirmation is sent to each operation system by the service interface and received disappearing for the instruction
Breath.
10. a kind of power consumer baseline load computing device, described device is used for the electricity that perform claim requires described in any one of 1-9
Power user's baseline Load Calculation Method, it is characterised in that described device includes:The memorizer of the reference load curve set that is stored with and
The control process device being connected with the memorizer;
The control process device includes that electricity consumption feature modeling unit, key factor identification unit, decision tree set up unit and baseline
Load prediction computing unit;Wherein,
The electricity consumption feature modeling unit, for obtaining historical load data, obtains history daily load curve collection;
In history daily load curve included in the history daily load curve collection and the reference load curve set for prestoring
Comprising reference load curve data structure correspondence;It is additionally operable to, according to the history daily load curve collection and the reference
Load curve collection, forms several feature load curves, and the feature load curve is converted to by CIM/OWL Ontology tables
Show, several described feature load curves form feature load curve collection;
The key factor determining unit, for determining the key factor for affecting user's baseline load;
The decision tree sets up unit, for setting up the decision tree for associating the feature load curve and the key factor;
The baseline load prediction and calculation unit, for using the decision tree, to user's baseline load meter being predicted
Calculate, baseline load result of calculation is fed back to into each operation system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611033053.9A CN106682999A (en) | 2016-11-18 | 2016-11-18 | Electric power user baseline load calculating method and apparatus thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611033053.9A CN106682999A (en) | 2016-11-18 | 2016-11-18 | Electric power user baseline load calculating method and apparatus thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106682999A true CN106682999A (en) | 2017-05-17 |
Family
ID=58866061
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611033053.9A Pending CN106682999A (en) | 2016-11-18 | 2016-11-18 | Electric power user baseline load calculating method and apparatus thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106682999A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111583063A (en) * | 2020-05-11 | 2020-08-25 | 国网四川省电力公司电力科学研究院 | Business start and end time detection method based on standard template and storage medium |
CN114662043A (en) * | 2022-05-25 | 2022-06-24 | 广东电网有限责任公司佛山供电局 | Real-time evaluation method for user load response condition and related device thereof |
CN115630753A (en) * | 2022-12-19 | 2023-01-20 | 西南交通大学 | Load baseline prediction method for electrolytic hydrogen production based on new energy multi-space-time scene |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100082172A1 (en) * | 2008-09-25 | 2010-04-01 | Korea Electric Power Corporation | Load forecasting analysis system for calculating customer baseline load |
KR20130082960A (en) * | 2011-12-26 | 2013-07-22 | 주식회사 케이티 | Method and system of producing realtime customer baseline load |
CN104239983A (en) * | 2014-10-13 | 2014-12-24 | 东南大学 | Big data perspective based demand response cutting load measurement method |
CN104881706A (en) * | 2014-12-31 | 2015-09-02 | 天津弘源慧能科技有限公司 | Electrical power system short-term load forecasting method based on big data technology |
KR20160022540A (en) * | 2014-08-20 | 2016-03-02 | 서강대학교산학협력단 | Method for estimating customer baseline load using data maning and apparatus thereof |
-
2016
- 2016-11-18 CN CN201611033053.9A patent/CN106682999A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100082172A1 (en) * | 2008-09-25 | 2010-04-01 | Korea Electric Power Corporation | Load forecasting analysis system for calculating customer baseline load |
KR20130082960A (en) * | 2011-12-26 | 2013-07-22 | 주식회사 케이티 | Method and system of producing realtime customer baseline load |
KR20160022540A (en) * | 2014-08-20 | 2016-03-02 | 서강대학교산학협력단 | Method for estimating customer baseline load using data maning and apparatus thereof |
CN104239983A (en) * | 2014-10-13 | 2014-12-24 | 东南大学 | Big data perspective based demand response cutting load measurement method |
CN104881706A (en) * | 2014-12-31 | 2015-09-02 | 天津弘源慧能科技有限公司 | Electrical power system short-term load forecasting method based on big data technology |
Non-Patent Citations (1)
Title |
---|
蔡洋主编: "《电力系统运行管理》", 31 August 1985, 北京:水利电力出版社 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111583063A (en) * | 2020-05-11 | 2020-08-25 | 国网四川省电力公司电力科学研究院 | Business start and end time detection method based on standard template and storage medium |
CN111583063B (en) * | 2020-05-11 | 2022-07-01 | 国网四川省电力公司电力科学研究院 | Business start and end time detection method based on standard template and storage medium |
CN114662043A (en) * | 2022-05-25 | 2022-06-24 | 广东电网有限责任公司佛山供电局 | Real-time evaluation method for user load response condition and related device thereof |
CN115630753A (en) * | 2022-12-19 | 2023-01-20 | 西南交通大学 | Load baseline prediction method for electrolytic hydrogen production based on new energy multi-space-time scene |
CN115630753B (en) * | 2022-12-19 | 2023-03-03 | 西南交通大学 | Load baseline prediction method for electrolytic hydrogen production based on new energy multi-space-time scene |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108647891A (en) | Data exception classification, Reasons method and device | |
Hilorme et al. | Smart grid concept as a perspective for the development of Ukrainian energy platform | |
CN109376971B (en) | Load curve prediction method and system for power consumers | |
CN110991700A (en) | Weather and electricity utilization correlation prediction method and device based on deep learning improvement | |
CN109583629A (en) | Improved similar historical day short-term load forecasting method and device based on deviation self-correcting | |
CN101888087A (en) | Method for realizing distributed super-short-term area load forecasting in distribution network terminal | |
CN104820876A (en) | Short-term load forecasting method and system | |
CN111680841B (en) | Short-term load prediction method, system and terminal equipment based on principal component analysis | |
CN106067137A (en) | A kind of credit rating evaluation methodology based on intelligent grid monitoring system | |
CN107506845A (en) | A kind of electricity sales amount Forecasting Methodology and its system based on multi-model fusion | |
CN111191811A (en) | Cluster load prediction method and device and storage medium | |
CN107133825A (en) | The Forecasting Methodology and device of a kind of businessman's volume of the flow of passengers | |
CN106682999A (en) | Electric power user baseline load calculating method and apparatus thereof | |
CN106910017A (en) | It is a kind of that analysis method and system are experienced based on the electric energy that user behavior data is excavated | |
Dong et al. | A multifactorial short-term load forecasting model combined with periodic and non-periodic features-a case study of qingdao, China | |
CN113191828B (en) | User electricity price value grade label construction method, device, equipment and medium | |
CN111626543A (en) | Method and device for processing power related data | |
CN109635996A (en) | Room rate tendency look-forward model construction method, device and storage medium | |
Lin et al. | An inventory system with investment to reduce yield variability and set-up cost | |
CN113887809A (en) | Power distribution network supply and demand balance method, system, medium and computing equipment under double-carbon target | |
CN104252647B (en) | Electro-load forecast method based on anti-distance weighting interpolation method | |
CN116865254A (en) | Power load index prediction method, system, equipment and medium | |
CN106651005B (en) | Baseline load prediction method and device | |
Lei et al. | Forecasting short-term residential electricity consumption using a deep fusion model | |
CN111080037A (en) | Short-term power load prediction method and device based on deep neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170517 |