CN105260803B - A kind of system power consumption prediction technique - Google Patents
A kind of system power consumption prediction technique Download PDFInfo
- Publication number
- CN105260803B CN105260803B CN201510752495.8A CN201510752495A CN105260803B CN 105260803 B CN105260803 B CN 105260803B CN 201510752495 A CN201510752495 A CN 201510752495A CN 105260803 B CN105260803 B CN 105260803B
- Authority
- CN
- China
- Prior art keywords
- user
- power consumption
- analysis
- influence factor
- factor
- 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 60
- 230000005611 electricity Effects 0.000 claims abstract description 71
- 238000004458 analytical method Methods 0.000 claims abstract description 67
- 238000004088 simulation Methods 0.000 claims description 17
- 238000013179 statistical model Methods 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 3
- 238000001816 cooling Methods 0.000 description 18
- 230000000694 effects Effects 0.000 description 10
- 238000010438 heat treatment Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Abstract
Accurately to carry out electricity demand forecasting, the present invention provides a kind of system power consumption prediction techniques, comprising: S1 data acquisition acquires the power information of the users such as resident, industry, business by intelligent electric meter, influence factor information is obtained by external system, while generating the user's history table；S2 user power utilization law-analysing is fitted analysis to the user power utilization information collected in conjunction with history lists, and analysis obtains user power utilization rule；S3 analysis of Influential Factors, is arranged fluctuation of load range according to user type, carries out analysis of Influential Factors to the time point beyond fluctuation range, obtains each influence factor to the influence value of the user power consumption；The prediction of S4 system power consumption, in conjunction with user power utilization law-analysing and analysis of Influential Factors, the short-term electricity consumption following to system is predicted.Prediction technique of the invention is a completely new developing direction of power-system short-term electricity demand forecasting from now on, has extensive use and significance in prediction field.
Description
Technical field
The invention belongs to electric system analysis of electric power consumption fields, are related to a kind of system power consumption prediction technique.
Background technique
The research of the short-term electricity demand forecasting of distribution network system has very long history, for a long time, short-term electricity demand forecasting
The system that object is generally confined within the whole network, domestic and foreign scholars have made the research work of a large amount of theory and method to this, have proposed
A variety of prediction techniques to differ from one another, such as time series method, expert system approach, artificial neural network method.
Short-term electricity consumption have it is obvious periodically, as the similitude of daily power consumption curve, same week type day it is similar
Property, similitude of festivals or holidays etc., furthermore influence of the short-term electricity consumption vulnerable to various environmental factors, such as weather, recreational and sports activities,
This makes the variation of electricity consumption the random process of non-stationary occur.Therefore presently, there are prediction technique, as time series method, specially
Family's systems approach, artificial neural network method etc., practical application effect is unsatisfactory, it is difficult to establish system power consumption and numerous influences
Relational model between factor.However, prediction result be formulate power grid hair, Transaction algorithm, Power System Planning, plan, electricity consumption,
Therefore can the element task of the departments such as scheduling propose that relatively reliable, accurate solution is always the class of people's research
Topic.
Summary of the invention
The purpose of the present invention is to provide a kind of new system power consumption prediction technique, this method can make up existing prediction
The deficiency of method.
A kind of system power consumption prediction technique, comprising the following steps:
S1 data acquisition
The power information that resident, industry, the users such as business are acquired by intelligent electric meter, by external system obtain influence because
Prime information, while generating the user's history table.
S2 user power utilization law-analysing
Analysis is fitted to the user power utilization information collected in conjunction with user's history table, analysis obtains user power utilization rule
Rule.
S3 analysis of Influential Factors
Fluctuation of load range is set according to user type, analysis of Influential Factors is carried out to the time point beyond fluctuation range,
Each influence factor is obtained to the influence value of the user power consumption.
The prediction of S4 system power consumption
In conjunction with user power utilization law-analysing and analysis of Influential Factors, the short-term electricity consumption following to system is predicted.
The user power utilization information that step S1 is obtained is real-time power information；The mode for generating user's history table is that timing will obtain
The information taken forms user power utilization information user's history lists.
Influence factor described in step S1 specifically includes that weather conditions, festivals or holidays factor, social event factor etc..
Preferably, the user's history table includes: User ID, user power consumption, Weather information, festivals or holidays type, correlation
Social event information.
Preferably, the user can be formed with every Weather information for counting a user power consumption every other hour and obtaining at that time
When historical statistics point；The when historical statistics point of the user is merged when daily morning 0~4 to be formed the user when daily power consumption,
Day user's history lists of Weather information, festivals or holidays type, related social event information.
Fitting analyzing method described in step S2 is linear analysis.
Preferably, linear analysis method described in step S2 are as follows: according to user's history table obtained by step S1 according to user type
It is y with user's daily power consumption to user power utilization information with electricity consumption type, is x with the date, carries out linear fit by following formula
Analysis, analysis obtain user power utilization law curve.
Y (x)=(xx-n+…+xx-2+xx-1+x)/(n+1)
N value is the value number in sampling process, and value range is divided according to user type and electricity consumption type, value
Quantity can be chosen according to system operations ability.In general, 5 working day points, weekends should not be less than for working day electricity consumption value
Electricity consumption value should not be the average value no more than 3 identical festivals or holidays, continuous electricity consumption less than 6 weekend points, holiday electricity consumption value
Value should not be less than 15 data points.
Analysis of Influential Factors method described in step S3 are as follows:
1) judge whether user is in active state, judgment method is to judge that user's daily power consumption is according to user type
It is no to be lower than threshold value.
2) as user is in active state, then judge whether the time point user power consumption exceeds fluctuation range.
3) weather conditions, festivals or holidays factor, the unicity of social event factor are carried out to the time point beyond fluctuation range
Analysis；Wherein weather conditions are that difference calculates, by calculating the gas epidemic disaster at the time point and the previous day, wind-force difference and being
It is no that there are the special variations such as rainfall, snowfall.
4) for example single influence factor, which then counts the influence factor, influences average value, maximum value and the variance of user power consumption
Value；Wherein weather conditions need to count corresponding difference；Unifactor model is formed according to statistical result.
5) for example multinomial influence factor then combines single factors obtained by step 3) to judge each influence factor using following methods
Influence coefficient:
A. all possible combination contained by multinomial influence factor is counted.
B. it is fitted under each combination that each influence factor average value is tired and the matching degree with true value.
C. it chooses closest to combination fitting and counts each influence factor in multinomial influence factor to user power consumption
Influence coefficient；Wherein weather conditions coefficient is in conjunction with difference；Form the statistical model under the multinomial influence factor.
D. each influence coefficient of same model is averaged.
Analysis obtains each influence factor to the influence value of the user power consumption.
Further, the weather conditions are main are as follows: temperature, humidity, rainfall/snow and wind-force；Festivals or holidays factor is main are as follows:
Common weekend, legal additional holiday, special event holiday；Social event factor is mainly associated event according to user type and draws
Point.
Preferably, the progress of Hadoop distributed computing framework can be used in analysis of Influential Factors described in step S3.
System power consumption prediction technique described in step S4 are as follows:
The following influence factor information each in a short time is obtained first, then according to the following influence factor selection existing in a short time
Corresponding simulation model, in conjunction with user power utilization law curve and corresponding influence factor simulation model, simulation and forecast is following short-term
The electricity consumption of the interior user；The use of the system realm can be obtained in the electricity demand forecasting for counting all users in each system realm
Power quantity predicting.
Preferably, system power consumption prediction technique of the present invention can be used for the short-term forecast of system power consumption.
Compared with existing prediction technique, prediction technique of the present invention is had the advantages that
1. analysis method of the present invention is based on user power utilization information, since user is determined by industry attribute, activity tool
There is apparent regularity, influence factor is relatively single, and the relationship of electricity consumption and influence factor is simpler and regular, and electricity consumption is special
Property be easier to hold, therefore analysis method of the present invention can more accurately grasp user power utilization rule.
2. the present invention has fully considered the major influence factors for influencing user power consumption, and calculates each influence factor
Influence value and when multinomial factor collective effect each factor contribution proportion, therefore, prediction technique of the present invention with it is existing pre-
Survey method is compared to more accurate.
3. reducing the operation pressure of system invention introduces the big data technology of Hadoop frame, improving system
Performance.
4. prediction technique of the present invention has stronger learning ability, as the increase of statistical model can more accurately have
The carry out electricity demand forecasting of effect, the process for increasing new prediction model is relatively simple, and increases new model and also do not need to carry out again
Analysis.
5. prediction technique of the invention is a completely new developing direction of electric system electricity demand forecasting from now on, will more added with
The stability and economy of operation of power networks are improved conducive to grid company, in addition, the dimensions such as user's electricity and industry attribute, GIS are believed
The combination of breath, it can be achieved that industry and plot electricity consumption trend study, there is extensive use and significance in prediction field.
Detailed description of the invention
Fig. 1 electricity demand forecasting method structure chart；
Fig. 2 analysis of Influential Factors method structure chart.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing, to the present invention into
One step is described in detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to limit this
Invention.
With the extensive use of the large area installation and power information acquisition system of intelligent electric meter, a large number of users is obtained
Power information to study user power consumption changing rule, since user is determined by industry attribute, production activity
With itself apparent regularity, influence factor is relatively single, and the relationship of electricity consumption and influence factor is simpler, uses electrical characteristics
It is easier to hold, therefore analysis of electric power consumption point closer to power demand is more conducive to grasp electricity consumption variational regularity.
Based on the above analysis, applicants have invented a kind of short-term electricity demand forecastings of the system based on mass users electricity consumption data
New method, this method is large number of due to being related to power grid user, and data volume is big, computationally intensive, traditional Computational frame without
The competent so a large amount of data of method calculate work, therefore introduce the big data technology based on Hadoop frame.
According to one embodiment of present invention, the specific implementation flow of system power consumption prediction technique of the present invention is as schemed
Shown in 1:
S1 data acquisition
The step is mainly completed the extraction to the user power utilizations information gathering data such as resident, industry, business, storage and is turned
It changes.The user power utilization information that the present invention is obtained by intelligent electric meter, the information are real-time power information；It is every to count one every other hour
Secondary user power consumption simultaneously obtains the when historical statistics point that Weather information at that time forms the user；By the use when daily morning 0~4
The when historical statistics point at family merges to form the user when daily power consumption, Weather information, festivals or holidays type, related social event information
Day user's history lists.Wherein, Weather information, festivals or holidays type, related social event information are obtained by external system.
By taking 0301 Room user of Chaoyang District, Beijing City Jiahe garden No. 1 building of cell as an example, which is resident, and User ID is
A-BJCYJHY-01-0301 will form the when historical statistics point of the user per system every other hour, as shown in the table:
Historical statistics point when 1 A-BJCYJHY-01-0301 user of table
The when historical statistics point of the user is merged when daily morning 0~4 to form the user when daily power consumption, weather are believed
Day user's history lists of breath, festivals or holidays type, related social event information, as shown in the table:
2 A-BJCYJHY-01-0301 user of table day user's history lists
S2 user power utilization law-analysing
According to electricity consumption user type, preliminary classification is carried out to user power utilization type, such as: resident's electricity consumption is divided into: work
Day electricity consumption, weekend electricity consumption, holiday electricity consumption；Industrial user classifies without electricity consumption；Commercial user is divided into: working day electricity consumption and holiday
Electricity consumption.
According to user's history table obtained by step S1 according to user type and electricity consumption type to user power utilization information with user day
Electricity consumption is y, is x with the date, carries out linear fit analysis by following formula, analysis obtains user power utilization law curve.
Y (x)=(xx-n+…+xx-2+xx-1+x)/(n+1)
N value is the value number in sampling process, and value range is divided according to user type and electricity consumption type, value
Quantity can be chosen according to system operations ability.In general, 5 working day points, weekends should not be less than for working day electricity consumption value
Electricity consumption value should not be the average value no more than 3 identical festivals or holidays, continuous electricity consumption less than 6 weekend points, holiday electricity consumption value
Value should not be less than 15 data points.
S3 analysis of Influential Factors
Influence factor includes: date factor, weather conditions, festivals or holidays factor, social event factor, wherein weather conditions master
It wants are as follows: gas epidemic disaster, rainfall/snow, wind-force etc.；Festivals or holidays factor is main are as follows: weekend, legal additional holiday, special event holiday
Deng；Social event factor is mainly associated event according to user type and divides.Analysis of Influential Factors is distributed using Hadoop
Computational frame carries out, and fluctuation of load range is arranged according to user type first, then carries out to the time point beyond fluctuation range
Analysis of Influential Factors.
As shown in Fig. 2, analysis of Influential Factors method can be with are as follows:
1) judge whether user is in active state, judgment method is to judge that user's daily power consumption is according to user type
It is no to be lower than threshold value.
2) as user is in active state, judge whether the time point user power consumption exceeds fluctuation range.
3) influence factors such as weather, festivals or holidays, social event are carried out to the time point beyond fluctuation range and carries out unicity
Analysis；Wherein weather conditions are that difference calculates, by calculating the gas epidemic disaster at the time point and the previous day, wind-force difference and being
It is no that there are the special variations such as rainfall, snowfall.
4) for example single influence factor, which then counts the influence factor, influences average value, maximum value and the variance of user power consumption
Value；Wherein weather conditions need to count corresponding difference；Unifactor model is formed according to statistical result.
5) for example multinomial influence factor then combines single factors obtained by step 3) to judge each influence factor using following methods
Influence coefficient:
A. all possible combination contained by multinomial influence factor is counted.
B. it is fitted under each combination that each influence factor average value is tired and the matching degree with true value.
C. it chooses closest to combination fitting and counts each influence factor in multinomial influence factor to user power consumption
Influence coefficient；Wherein weather conditions coefficient is in conjunction with difference；Form the statistical model under the multinomial influence factor.
D. each influence coefficient of same model is averaged.
Analysis obtains each influence factor to the influence value of the user power consumption.
By taking 0301 Room user of Chaoyang District, Beijing City Jiahe garden No. 1 building of cell as an example, which is resident, and User ID is
The electricity consumption fluctuation range of the resident is arranged according to Chaoyang District, Beijing City ordinarily resident's electricity consumption by A-BJCYJHY-01-0301
It is ± 2 degree, activity threshold is 2 degree, and ordinarily resident is associated with social event and specifically includes that grave news, important competitive sports, periphery
Business advertising campaign etc. in range.
On May 1st, 2015 is International Labour Day legal festivals and holidays -, counts user spring--3 days legal festivals and holidays average value
Electricity consumption is 25 degree, 28 degree, 24 degree, and the electricity consumption on practical user's spring-International Labour Day legal festivals and holidays-is 24 degree, 28 degree, 24
Degree judges using the legal festivals and holidays -3 days average value as the legal festivals and holidays -3 days simulation models.
On May 9th, 2015 is weekend, which is 25 degree, but its practical electricity consumption is 28 degree,
Then carry out analysis of Influential Factors to the day: analysis show that this day, there are 3 influence factors, were respectively as follows: cooling 3 DEG C, shower first；
Obtain that cooling, shower is on the independent influence statistical value of the user later :+0.7 degree of cooling/DEG C, shower+1.2 spend, Fitting Analysis obtains
It is out weekend electricity consumption average value and the sum of cooling, shower influence value closest to the simulation model of true value under 3 influence factors.
Calculate this analysis in each influence factor influence coefficient are as follows: cooling+0.7 degree/DEG C, shower+1.2 spend, with other same model
The influence coefficient of same affect factor be averaged and obtain user spring-weekend-cooling, shower simulation model are as follows: weekend
The degree of electricity consumption average value+0.7021/DEG C/cooling+1.2112 degree/shower.
On May 13rd, 2015 is common working day, which is 15 degree, but its practical electricity consumption
Be 24 degree, then carry out analysis of Influential Factors to the day: analysis show that this day there are 3 influence factors, is respectively as follows: unskilled labor first
Make day, yin~fine, 9.5 DEG C of heating；Judge yin~it is fine be not counted in analysis of Influential Factors for common weather events, later obtain liter
Temperature is on the independent influence statistical value of the user: heat up for+1 degree/DEG C, Fitting Analysis obtain logical working day, heat up influence factor
The simulation model of the lower user power consumption closest to true value is commonly work the sum of daily level mean value and heating influence value.Meter
Calculation obtains in this analysis, the influence coefficient of heating be+0.9474 degree/DEG C and other influenced after coefficient is averaged with class model
Obtain spring-common working day-Elevated Temperature Conditions lower die analog model is common work+1.0211/ DEG C of daily power consumption/heating.
With Beijing economic and technological development zone, for the electronic medical instruments Co., Ltd of Beijing Feitian, Customs Assigned Number B-
BJJK0011, producing class business electrical amount its fluctuation range is arranged according to Beijing economic and technological development zone is ± 5%.
On May 1st, 2015, linear analysis show that the daily power consumption is 3230 degree, but its practical electricity consumption is 3782 degree, is increased
Width is 17.09%, exceeds fluctuation range, therefore carries out analysis of Influential Factors to it: first analysis obtain there are 3 influence because
Element: 2 DEG C rainfall, cooling, International Labour Day public holiday-obtain the independent influence of rainfall, cooling, public holiday on the user later
Statistical value: rainfall is+227 degree, cooling be+26.3 degree/DEG C, the public holiday, Fitting Analysis was obtained there is no independent influence factor
The insufficient practical amplification of the sum of rainfall, cooling influence value, judges that for special event, legal vacation is calculated in International Labour Day public holiday-
The influence value on day-International Labour Day is+62 degree/day, forms public holiday in spring-- rainfall, the simulation model for the factor that cools down is to use level
+ 227 degree of the mean value/degree of rainfall+26.3/DEG C/cooling+62 degree/day/legal festivals and holidays.
By taking Chaoyang District, Beijing City Carrefour hypermarket Wangjing shop as an example, Customs Assigned Number C-BJCYJLF003, according to Beijing
It is ± 10% that its fluctuation range, which is arranged, in the common trade company's electricity consumption in Chaoyang District.
On May 1st, 2015, linear analysis show that the daily power consumption is 1112 degree, but its practical electricity consumption is 1521 degree, is increased
Width is 36.78%, exceeds fluctuation range, therefore carries out analysis of Influential Factors to it: first analysis obtain there are 3 influence because
Element: 2 DEG C rainfall, cooling, International Labour Day public holiday-obtain the independent influence of rainfall, cooling, public holiday on the user later
Statistical value: rainfall be -101 degree, cooling be+18.9 degree/DEG C, the public holiday+211 spend, Fitting Analysis obtain cooling influence value and
The sum of public holiday closest to actual value, judges that rainfall influences coefficient in this analysis for 0, similar pattern is not present in judgement
Type, then with cooling for+18.9 degree/DEG C, the public holiday+211 spends simulation model as public holiday in spring-- rainfall, cooling.
The prediction of S4 system power consumption
Prediction technique of the present invention are as follows:
The following influence factor information each in a short time is obtained first, then according to the following influence factor selection existing in a short time
Corresponding simulation model, in conjunction with user power utilization law curve and corresponding influence factor simulation model, simulation and forecast is following short-term
The electricity consumption of the interior user；The use of the system realm can be obtained in the electricity demand forecasting for counting all users in each system realm
Power quantity predicting.
By taking 0301 Room user of Chaoyang District, Beijing City Jiahe garden No. 1 building of cell as an example, which is resident, and User ID is
A-BJCYJHY-01-0301, the electricity consumption of following 3 days users from needing to predict on August 10th, 2015.Deterministic process is as follows:
Following 3 days from August 10th, 1.2015 are common working day, transfer 27~31 July and August 3~7 altogether
10 workaday electricity consumptions form predicted mean vote.
2. from analysis on August 10th, 2015 following 3 days without legal section vacation, without related social event, Weather information are as follows: 8
There is within 13rd thunder shower the moon, the following average daily temperature temperature difference on the 3rd is respectively less than 3 DEG C, is unsatisfactory for threshold value, therefore choose simulation in following 3 days
Model are as follows: on 11 summer-of August on common working day, 12 summer-of August common working day, August 13rd are commonly to work summer-
Day-thunder shower model.
The electricity consumption that simulation is calculated following 3 days is respectively as follows: 21 degree, 21 degree, 23 degree, practical electricity consumption is 21 degree,
22 degree, 22 degree.
Compared with existing prediction technique, prediction technique of the present invention is had the advantages that
1. analysis method of the present invention is based on user power utilization information, since user is determined by industry attribute, activity tool
There is apparent regularity, influence factor is relatively single, and the relationship of electricity consumption and influence factor is simpler and regular, and electricity consumption is special
Property be easier to hold, therefore analysis method of the present invention can more accurately grasp user power utilization rule.
2. the present invention has fully considered the major influence factors for influencing user power consumption, and calculates each influence factor
Influence value and when multinomial factor collective effect each factor contribution proportion, therefore, prediction technique of the present invention with it is existing pre-
Survey method is compared to more accurate.
3. reducing the operation pressure of system invention introduces the big data technology of Hadoop frame, improving system
Performance.
4. prediction technique of the present invention has stronger learning ability, as the increase of statistical model can more accurately have
The carry out electricity demand forecasting of effect, the process for increasing new prediction model is relatively simple, and increases new model and also do not need to carry out again
Analysis.
5. prediction technique of the invention is a completely new developing direction of electric system electricity demand forecasting from now on, will more added with
The stability and economy that operation of power networks is improved conducive to grid company, in addition, the dimensions such as user power consumption and industry attribute, GIS
The combination of information, it can be achieved that industry and plot electricity consumption trend study, there is extensive use and significance in prediction field.
It should be noted that and understand, in the case where not departing from spirit and scope required by the claims in the present invention, energy
It is enough that various modifications and improvements are made to the present invention of foregoing detailed description.It is therefore desirable to protection technical solution range not by
The limitation of given any specific exemplary teachings.
Claims (6)
1. a kind of system power consumption prediction technique, which comprises the following steps:
S1 data acquisition
The power information that resident, industry, commercial user are acquired by intelligent electric meter obtains influence factor letter by external system
Breath, while generating the user's history table；
S2 user power utilization law-analysing
Analysis is fitted to the user power utilization information collected in conjunction with user's history table, analysis obtains user power utilization rule；
S3 analysis of Influential Factors
Fluctuation of load range is set according to user type, analysis of Influential Factors is carried out to the time point beyond fluctuation range, is obtained
Influence value of each influence factor to the user power consumption；
The prediction of S4 system power consumption
In conjunction with user power utilization law-analysing and analysis of Influential Factors, the short-term electricity consumption following to system is predicted；
The user power utilization information that step S1 is obtained is real-time power information；The mode timing for generating user's history table will acquire
Information forms the user's history table of the user power utilization information；The influence factor includes: weather conditions, festivals or holidays factor, society
Event factor；
The user's history table includes: User ID, user power consumption, Weather information, festivals or holidays type, related social event letter
Breath；
The generating mode of the user's history table is: per user power consumption of statistics every other hour and the weather of acquisition at that time is believed
Breath forms the when historical statistics point of the user；Merge the when historical statistics point of the user to form the user when daily morning 0~4
When daily power consumption, Weather information, festivals or holidays type, related social event information day history lists；
Analysis of Influential Factors described in step S3 are as follows:
1) judge whether user is in active state, judgment method is to judge whether user's daily power consumption is low according to user type
In threshold value；
2) as user is in active state, judge whether the time point user power consumption exceeds fluctuation range；
3) weather conditions, festivals or holidays factor, the unicity of social event factor is carried out to the time point beyond fluctuation range to analyze；
Wherein weather conditions are that difference calculates, by calculating the gas epidemic disaster at the time point and the previous day, wind-force difference and whether there is
Rainfall, the special variation of snowfall；
4) for example single influence factor, which then counts the influence factor, influences average value, maximum value and the variance yields of user power consumption；
Wherein weather conditions need to count corresponding difference；Unifactor model is formed according to statistical result；
5) for example multinomial influence factor then combines single factors obtained by step 3) to judge the shadow of each influence factor using following methods
Ring coefficient:
A. all possible combination contained by multinomial influence factor is counted；
B. it is fitted under each combination that each influence factor average value is tired and the matching degree with true value；
C. it chooses closest to combination fitting and counts each influence factor in multinomial influence factor to the influence of user power consumption
Coefficient；Wherein weather conditions coefficient is in conjunction with difference；Form the statistical model under the multinomial influence factor；
D. each influence coefficient of same model is averaged.
2. system power consumption prediction technique according to claim 1, which is characterized in that the weather conditions are as follows: temperature, wet
Degree, rainfall/snow and wind-force；Festivals or holidays factor are as follows: common weekend, legal additional holiday, special event holiday；Social event factor
Event is associated according to user type to divide.
3. system power consumption prediction technique according to claim 1, which is characterized in that Fitting Analysis described in step S2 is linear
Fitting Analysis.
4. system power consumption prediction technique according to claim 3, which is characterized in that the analysis of linear fit described in step are as follows:
It is with user's daily power consumption to user power utilization information according to user type and electricity consumption type according to user's history table obtained by step S1
Y is x with the date, carries out linear fit analysis by following formula, analysis obtains user power utilization law curve；
Y (x)=(xx-n+…+xx-2+xx-1+x)/(n+1)
N value is the value number in sampling process, and value range is divided according to user type and electricity consumption type, value quantity
It can be chosen according to system operations ability.
5. system power consumption prediction technique according to claim 1, which is characterized in that analysis of Influential Factors described in step S3 can
It is carried out using Hadoop distributed computing framework.
6. system power consumption prediction technique according to claim 1, which is characterized in that the prediction of system power consumption described in step S4
Method are as follows: the following influence factor information each in a short time is obtained first, then according to the following influence factor selection existing in a short time
Corresponding simulation model, in conjunction with user power utilization law curve and corresponding influence factor simulation model, simulation and forecast is following short-term
The electricity consumption of the interior user；The use of the system realm can be obtained in the electricity demand forecasting for counting all users in each system realm
Power quantity predicting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510752495.8A CN105260803B (en) | 2015-11-06 | 2015-11-06 | A kind of system power consumption prediction technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510752495.8A CN105260803B (en) | 2015-11-06 | 2015-11-06 | A kind of system power consumption prediction technique |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105260803A CN105260803A (en) | 2016-01-20 |
| CN105260803B true CN105260803B (en) | 2019-04-19 |
Family
ID=55100482
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510752495.8A Active CN105260803B (en) | 2015-11-06 | 2015-11-06 | A kind of system power consumption prediction technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105260803B (en) |
Families Citing this family (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105809573A (en) * | 2016-03-02 | 2016-07-27 | 深圳供电局有限公司 | Big data analysis based load nature authentication method |
| CN106447067A (en) * | 2016-06-30 | 2017-02-22 | 国网江苏省电力公司南通供电公司 | Court daily power consumption prediction method based on adaptive nonlinear regression model |
| CN106408183A (en) * | 2016-09-09 | 2017-02-15 | 国家电网公司 | Power supply index regulating method |
| CN106529747A (en) * | 2017-01-04 | 2017-03-22 | 成都四方伟业软件股份有限公司 | Power load predicting method and system based on large data |
| CN106992525B (en) * | 2017-05-26 | 2020-09-04 | 国网山东省电力公司泰安供电公司 | Power load management method and device |
| CN107292455A (en) * | 2017-07-31 | 2017-10-24 | 华自科技股份有限公司 | Electricity demand forecasting method, device, readable storage medium storing program for executing and computer equipment |
| CN107808348A (en) * | 2017-09-28 | 2018-03-16 | 昆明能讯科技有限责任公司 | A kind of sale of electricity company provides the method that purchase of electricity declares prediction for client |
| CN108493946A (en) * | 2018-04-12 | 2018-09-04 | 辽宁石油化工大学 | Electric energy control method, device and equipment based on user power utilization analysis |
| CN108680805A (en) * | 2018-05-08 | 2018-10-19 | 广东昊阳电力建设有限公司 | Online power analysis forecasting system and electric system |
| CN109460849A (en) * | 2018-09-10 | 2019-03-12 | 杭州中恒云能源互联网技术有限公司 | A kind of user power utilization Deviation Analysis Method based on intelligent algorithm |
| CN109993424B (en) * | 2019-03-26 | 2023-06-23 | 广东艾胜物联网科技有限公司 | Non-interference type load decomposition method based on width learning algorithm |
| CN110136024B (en) * | 2019-03-28 | 2023-10-10 | 上海芯联芯智能科技有限公司 | Method and device for acquiring electricity utilization characteristics, electricity utilization habits and electricity consumption predicted values of users |
| 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 |
| CN110288215A (en) * | 2019-06-17 | 2019-09-27 | 国网甘肃省电力公司兰州供电公司 | A kind of power customer service Situation Awareness evaluation method based on entropy assessment |
| CN110472991B (en) * | 2019-06-27 | 2022-10-28 | 拉扎斯网络科技（上海）有限公司 | Data processing method, device, server and storage medium |
| CN110414724A (en) * | 2019-07-10 | 2019-11-05 | 东软集团股份有限公司 | For predicting method, apparatus, readable storage medium storing program for executing and the electronic equipment of power consumption |
| CN110689250B (en) * | 2019-09-20 | 2023-04-14 | 深圳供电局有限公司 | Method and system for processing user electricity consumption data and computer readable medium |
| CN111026791B (en) * | 2019-12-26 | 2021-08-10 | 广东电网有限责任公司电力科学研究院 | User type judgment method based on resident fine-grained electricity consumption data |
| CN111680938B (en) * | 2020-08-13 | 2020-11-06 | 国网浙江省电力有限公司 | Power flow type big data based rework and production monitoring method and system and readable medium |
| CN112070302A (en) * | 2020-09-08 | 2020-12-11 | 北京中宸泓昌科技有限公司 | Early warning method and system for power utilization safety of power grid |
| CN112434892A (en) * | 2020-12-21 | 2021-03-02 | 深圳供电局有限公司 | Differential power demand prediction method and system based on multi-factor combination learning |
| CN113239029A (en) * | 2021-05-18 | 2021-08-10 | 国网江苏省电力有限公司镇江供电分公司 | Completion method for missing daily freezing data of electric energy meter |
| CN113449919B (en) * | 2021-06-29 | 2022-12-09 | 国网山东省电力公司菏泽供电公司 | Power consumption prediction method and system based on feature and trend perception |
| CN113705999A (en) * | 2021-08-19 | 2021-11-26 | 国网河南省电力公司周口供电公司 | Electric power big data automatic reasoning platform and electric power distribution system |
| CN114444397A (en) * | 2022-01-28 | 2022-05-06 | 国家电网有限公司 | Short-term power consumption prediction algorithm based on user portrait data model |
| CN114529072A (en) * | 2022-02-11 | 2022-05-24 | 杭州致成电子科技有限公司 | Regional electric quantity prediction method based on time series |
| CN115423301B (en) * | 2022-09-01 | 2023-04-25 | 杭州达中科技有限公司 | Intelligent electric power energy management and control method, device and system based on Internet of things |
| CN115796404A (en) * | 2022-11-10 | 2023-03-14 | 国网江苏省电力有限公司营销服务中心 | Power utilization prediction method and system for agent electricity purchasing users |
| CN115829117B (en) * | 2022-11-29 | 2023-09-19 | 湖北华中电力科技开发有限责任公司 | Method and system for predicting electricity consumption in any area in short time based on terminal electricity consumption |
| CN115660225A (en) * | 2022-12-13 | 2023-01-31 | 浙江万胜智能科技股份有限公司 | Electricity load prediction management method and system based on ammeter communication module |
| CN116011685A (en) * | 2023-03-22 | 2023-04-25 | 国网陕西省电力有限公司西咸新区供电公司 | Electric quantity prediction method, electric quantity prediction device, computer equipment and storage medium |
| CN116526479B (en) * | 2023-07-03 | 2023-10-13 | 国网北京市电力公司 | Method, device, equipment and medium for predicting power supply quantity |
| CN116937569A (en) * | 2023-07-26 | 2023-10-24 | 广东永光新能源设计咨询有限公司 | Intelligent energy storage method and device for photovoltaic power generation and electronic equipment |
| CN116709062B (en) * | 2023-08-07 | 2023-10-20 | 安徽融兆智能有限公司 | Electricity consumption information acquisition equipment with detection function |
| CN117689070A (en) * | 2023-12-13 | 2024-03-12 | 北京朗杰科技有限公司 | Ammeter management system based on internet of things equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102723785A (en) * | 2012-07-03 | 2012-10-10 | 广东电网公司电力科学研究院 | Electric power load prediction method and electric power load prediction system based on collection of larger user intelligent terminal |
| US8600572B2 (en) * | 2010-05-27 | 2013-12-03 | International Business Machines Corporation | Smarter-grid: method to forecast electric energy production and utilization subject to uncertain environmental variables |
| CN104992239A (en) * | 2015-06-24 | 2015-10-21 | 国网山东省电力公司电力科学研究院 | Correlation coefficient-based industry electricity consumption law forecasting method |
-
2015
- 2015-11-06 CN CN201510752495.8A patent/CN105260803B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8600572B2 (en) * | 2010-05-27 | 2013-12-03 | International Business Machines Corporation | Smarter-grid: method to forecast electric energy production and utilization subject to uncertain environmental variables |
| CN102723785A (en) * | 2012-07-03 | 2012-10-10 | 广东电网公司电力科学研究院 | Electric power load prediction method and electric power load prediction system based on collection of larger user intelligent terminal |
| CN104992239A (en) * | 2015-06-24 | 2015-10-21 | 国网山东省电力公司电力科学研究院 | Correlation coefficient-based industry electricity consumption law forecasting method |
Non-Patent Citations (1)
| Title |
|---|
| 基于大数据技术的短期负荷分析与预测;王德金;《华东电力》;20141030;第42卷(第10期);参见正文第2007-2010页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105260803A (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105260803B (en) | A kind of system power consumption prediction technique | |
| WO2018082523A1 (en) | Load cycle mode identification method | |
| CN108593990B (en) | Electricity stealing detection method based on electricity consumption behavior mode of electric energy user and application | |
| CN103413253B (en) | A kind of classification of the annual peak load based on economy, meteorologic factor Forecasting Methodology | |
| CN110097297A (en) | A kind of various dimensions stealing situation Intellisense method, system, equipment and medium | |
| Chen et al. | Day-ahead prediction of hourly electric demand in non-stationary operated commercial buildings: A clustering-based hybrid approach | |
| CN102999791A (en) | Power load forecasting method based on customer segmentation in power industry | |
| CN111896680B (en) | Greenhouse gas emission analysis method and system based on satellite remote sensing data | |
| CN103226736B (en) | Based on the long-medium term power load forecasting method of cluster analysis and grey target theory | |
| CN109508835A (en) | A kind of wisdom Power Network Short-Term Electric Load Forecasting method of integrated environment feedback | |
| CN108959424A (en) | A kind of operating method of the city electricity consumption map for power system load monitoring | |
| CN109858728A (en) | Load forecasting method based on branch trade Analysis of Electrical Characteristics | |
| CN110782153A (en) | Modeling method and system for comprehensive energy efficiency assessment system of enterprise park | |
| Chen et al. | An energy planning oriented method for analyzing spatial-temporal characteristics of electric loads for heating/cooling in district buildings with a case study of one university campus | |
| CN105868073A (en) | Data center energy saving strategy implementation method based on time series analysis | |
| Zhang et al. | Effect of physical, environmental, and social factors on prediction of building energy consumption for public buildings based on real-world big data | |
| CN108256724B (en) | Power distribution network open capacity planning method based on dynamic industry coefficient | |
| CN115660225A (en) | Electricity load prediction management method and system based on ammeter communication module | |
| CN104715156A (en) | Building energy-saving potential dynamic assessment method | |
| CN104134102B (en) | Long-term electricity needs distribution forecasting method in power network based on LEAP models | |
| CN105913366A (en) | Industrial electric power big data-based regional industry business climate index building method | |
| Ahmad et al. | Statistical distributions of load profiling data | |
| CN104751253B (en) | Distribution power flow Forecasting Methodology based on B- spline Basis bottom developed curve cluster | |
| CN104134096A (en) | Intelligent load prediction method based on province and prefecture integration | |
| Gerossier et al. | A novel method for decomposing electricity feeder load into elementary profiles from customer information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |