CN107563544A - A kind of gray model power predicating method based on day character vector optimization - Google Patents
A kind of gray model power predicating method based on day character vector optimization Download PDFInfo
- Publication number
- CN107563544A CN107563544A CN201710702936.2A CN201710702936A CN107563544A CN 107563544 A CN107563544 A CN 107563544A CN 201710702936 A CN201710702936 A CN 201710702936A CN 107563544 A CN107563544 A CN 107563544A
- Authority
- CN
- China
- Prior art keywords
- mrow
- day
- month
- power consumption
- 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.)
- Pending
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of gray model power predicating method based on day character vector optimization, method fills vacancies in the proper order model as major prognostic model using dimensions such as grey, and the daily power consumption data of nearest ten days are predicted using day character vector method, daily power consumption data are combined to obtain to the higher moon electricity demand forecasting result of the degree of accuracy of this month and next month with of that month prediction data.The power predicating method of the present invention, simple and easy, positive effect can be as the main method of branch trade power quantity predicting and the auxiliary revision method of power system power quantity predicting.
Description
Technical field
The present invention relates to a kind of gray model power predicating method based on day character vector optimization, belong to power marketing intelligence
Can applied technical field.
Background technology
Electric Power Forecasting Based on Artificial, which is that power grid enterprises are scientific and reasonable, arranges every production and operating activities, ensures electricity net safety stable fortune
Capable important foundation work, and government department formulate the important evidence of electric power relevant policies.Current economic structural adjustment is increasingly
In-depth, Power Market is complicated and changeable, and new changing trend is presented to the demand of quantity of electricity in socio-economic development, particularly
Power system reform deepens constantly, and market access degree expands day by day, and the increase of electricity market main body, it is pre- how to effectively improve electricity
Surveying particularly middle or short term power quantity predicting accuracy turns into the common issue that each main market players faces.The middle or short term of month power consumption data
Prediction has a variety of methods, and it is a kind of more effective method that wherein the dimension such as grey, which fills vacancies in the proper order prediction,.But use a certain prediction side
Method can increase the uncertainty of prediction, and the combined prediction of a variety of methods can reduce asking for the prediction imperfect tape of single method
Topic.
The content of the invention
The technical problems to be solved by the invention are the defects of overcoming prior art, there is provided one kind is excellent based on day character vector
The gray model power predicating method of change, with day character vector method and waits dimension to fill vacancies in the proper order Grey Modelss, realizes moon power consumption
Prediction.
In order to solve the above technical problems, the present invention provides a kind of gray model power quantity predicting based on day character vector optimization
Scheme.It is then pre- using the history moon power consumption of 12 months first with following ten days daily power consumption of day character vector method prediction
Sequencing row, the moon power consumption in this month is predicted according to Grey Modelss, utilizes the daily power consumption knot in ten days futures predicted
Fruit adjusts the concept structure gray model that the dimension such as of that month electricity demand forecasting result, foundation is filled vacancies in the proper order, and then predicts the moon of next month
Power consumption.
The technology used in the present invention means are as follows：
A kind of gray model power predicating method based on day character vector optimization, comprises the following steps：
1) following ten days daily power consumption data since being predicted day are calculated according to day character vector method；
2) of that month moon power consumption is predicted by gray model algorithm；
3) based on etc. dimension fill vacancies in the proper order with gray model algorithm prediction next month moon power consumption.
The computational methods of foregoing step 1) daily power consumption data are：
Meteorological factor 1-1) is calculated, calculates each history max. daily temperature, minimum temperature, mean temperature and humidity four respectively
Sub- factor and the matching factor of prediction day, calculation formula are as follows：
Wherein, SjFor the matching factor of j-th of sub- factor, j=1,2,3,4, represent respectively maximum temperature, minimum temperature,
Mean temperature and humidity, minjDay and the least absolute value of history day difference, max are predicted for j-th of sub- factorjFor j-th it is sub- because
Element prediction day and the maximum value of history day difference, valuejThe absolute of day and history day difference is predicted for j-th of sub- factor
Value, ρ is regulation coefficient；
Then weighted average is done to the matching factor of every sub- factor and obtains meteorological factor T：
Time factor D 1-2) is calculated, calculation formula is as follows：
D=β1 Mod (t, 7)β2 int(t/7) (3)
Wherein, β1For day attenuation coefficient, β2For all attenuation coefficients, t is the date differences of prediction day and history day, and mod is to take
Cofunction, int are bracket function；
The week factor W 1-3) is calculated, calculation formula is as follows：
W=1- | f (X1)-f(X0)| (4)
Wherein, f (X1) be history day week type, f (X0) for prediction day week type, X1Value be Monday, week
Two, Wednesday, Thursday, Friday, Saturday, Sunday, f (X1) value be：F (Monday)=0.1, f (Tuesday)=f (Wednesday)=f (weeks
Four)=0.2, f (Friday)=0.3, f (Saturday)=0.7, f (Sunday)=1；
Analogue forecasting method 1-4) is calculated, is integrated by the way of meteorological factor, time factor, week fac-tor
Similarity factor, take analogue forecasting method maximum first 5 are used as similar day set, to the daily power consumption of similar day with similarity factor
Daily power consumption after being weighted averagely as weight as prediction day.
Foregoing history day is to predict the date in 35 days a few days ago.
Foregoing ρ values are 0.5.
Foregoing β1Value is 0.96, β2Value is 0.94.
Foregoing step 2) prediction it is of that month the moon power consumption process it is as follows：
Smoothing factor 2-1) is calculated, is calculated monthly with the average moon power consumption divided by average year power consumption of the first three years monthly
Smoothing factor；
2-2) utilize the history moon power consumption data structure forecasting sequence x of 12 months(0), with moon power consumption divided by smooth system
Several modes are smoothed to forecasting sequence；
2-3) ought power consumption month in and month out by the prediction of grey GM (1,1) model formation；
2-4) prediction result be multiplied by this month smoothing factor reduced after ought electricity demand forecasting result month in and month out.
Foregoing step 3) prediction next month the moon power consumption process it is as follows：
3-1) the daily power consumption result in ten days futures predicted using step 1) is added up, with the electricity after cumulative and
Replacement step 2-2) in of that month electricity demand forecasting result phase same date electricity, form new of that month electricity demand forecasting result；
3-2) according to etc. the concept filled vacancies in the proper order of dimension, by after adjustment ought be before electricity demand forecasting result adds smoothing processing month in and month out
The forecasting sequence that power consumption forms month in and month out of history 12 simultaneously removes first data in former forecasting sequence, obtains new pre- sequencing
Row；
3-3) new forecasting sequence is smoothed；
The moon power consumption of next month 3-4) is predicted by grey GM (1,1) model formation；
3-5) prediction result be multiplied by this month smoothing factor reduced after next month moon electricity demand forecasting result.
Foregoing grey GM (1,1) model formation is：
Wherein, k=1,2,3 ... 11 represent k-th month, x(0)For forecasting sequence,Represent kth+1 month
Month electricity demand forecasting result, x(0)(1) moon power consumption of 1st month is represented, a is development coefficient, and u is grey actuating quantity, and a, u are equal
For the differential equationCoefficient.
What the present invention was reached has the beneficial effect that：
The power predicating method of the present invention, simple and easy, positive effect can be as the main side of branch trade power quantity predicting
The auxiliary revision method of method and power system power quantity predicting；This method improves the power quantity predicting degree of accuracy simultaneously, is carried for dispatching of power netwoks
Supported for more accurate data, traffic department is shifted to an earlier date the method for operation for rationally, economically arranging future, so as to improve
The performance driving economy of power network, higher profit is brought for Utilities Electric Co., more solid guarantee is provided for power grid operation.
Brief description of the drawings
Fig. 1 is the moon electricity demand forecasting overall framework figure of the present invention；
Fig. 2 is the moon electricity demand forecasting flow chart of the present invention；
Fig. 3 is the daily power consumption prediction flow chart of the present invention；
Fig. 4 is power quantity predicting curve in embodiment.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings.Following examples are only used for clearly illustrating the present invention
Technical scheme, and can not be limited the scope of the invention with this.
Electric Power Forecasting Based on Artificial, which is that power grid enterprises are scientific and reasonable, arranges every production and operating activities, ensures electricity net safety stable fortune
Capable important foundation work, and government department formulate the important evidence of electric power relevant policies.Current economic structural adjustment is increasingly
In-depth, Power Market is complicated and changeable, and new changing trend is presented to the demand of quantity of electricity in socio-economic development, particularly
Power system reform deepens constantly, and market access degree expands day by day, and the increase of electricity market main body, it is pre- how to effectively improve electricity
Surveying particularly middle or short term power quantity predicting accuracy turns into the common issue that each main market players faces.The present invention is by using the whole society
The Study on Forecasting Method of electricity and trade classification power consumption, make every effort to provide new thinking and side for quantity of electricity analysis prediction work
Method.
The gray model power predicating method based on day character vector optimization of the present invention includes daily power consumption prediction, the moon is used
Power quantity predicting and prediction data amendment three parts content, pre- flow gauge are as shown in Figure 2.
Data Analysis Services procedure level relation during prediction including basic data layer as shown in figure 1, be used to collect
Electric quantity data (the industry moon electricity, industry day electricity), meteorological data (maximum temperature, minimum temperature, mean temperature, humidity), section
Holiday data；Data analysis layer is used to analyze branch trade electricity and meteorological, festivals or holidays relation；Construction of A Model layer is used to build
Vertical day character vector model and gray model；As a result presentation layer is used to be shown industry-specific moon power quantity predicting result.
Daily power consumption prediction detailed process be：Analyze since predict day the meteorology in ten days futures, the time, week etc. because
The similitude of element, following ten days daily power consumption data are calculated according to day character vector method.In definition prediction 35 days a few days ago
Date is referred to as history day, predicts that the meteorological data of day passes through weather forecast data acquisition.Pre- flow gauge as shown in figure 3, including with
Lower step：
1-1) calculates meteorological factor, calculates each history max. daily temperature, minimum temperature, mean temperature and humidity four respectively
Individual sub- factor and the matching factor of prediction day, weighted average is done to the matching factor of every sub- factor and obtains meteorological factor
Value.The matching factor calculation formula of every sub- factor is as follows：
Wherein, SjFor j-th of sub- factor matching factor (j=1,2,3,4, represent respectively maximum temperature, minimum temperature,
Mean temperature and humidity)；minjDay and the least absolute value of history day difference are predicted for j-th of sub- factor；maxjFor j-th of son
Factor predicts day and the maximum value of history day difference；valuejThe exhausted of day and history day difference is predicted for j-th of sub- factor
To value；ρ is regulation coefficient, typically takes 0.5.Obtain all SjAfter value, do weighted average and obtain meteorological factor T：
A meteorological factor is daily obtained for each history.
1-2) calculates time factor, according to the principle of " near big and far smaller ", the value of time factor is calculated using formula, is calculated
Formula is as follows：
D=β1 Mod (t, 7)β2 int(t/7) (3)
Wherein, β1For day attenuation coefficient, 0.96 is taken in of the invention；β2For all attenuation coefficients, 0.94 is taken in of the invention；T is pre-
Survey the date differences of day and history day；Mod is remainder function；Int is bracket function.
1-3) calculates the week factor, calculates the week factor according to history day and the difference of prediction week day type, calculates public
Formula is as follows：
W=1- | f (X1)-f(X0)| (4)
Wherein, f (X1) be history day week type, f (X0) for prediction day week type, X1Value be Monday, week
Two, Wednesday, Thursday, Friday, Saturday, on Sunday, mapping ruler is set to according to actual conditions：F (Monday)=0.1, f (Tuesday)=
F (Wednesday)=f (Thursday)=0.2, f (Friday)=0.3, f (Saturday)=0.7, f (Sunday)=1.
1-4) calculates analogue forecasting method, is obtained by the way of meteorological factor, time factor, week fac-tor comprehensive
Similarity factor is closed, take analogue forecasting method maximum first 5 are used as similar day set, to the daily power consumption of similar day with similar system
Count the daily power consumption as prediction day after being weighted averagely as weight.
The detailed process of month electricity demand forecasting is：The moon power consumption in this month is calculated by gray model algorithm, with reference to not
The daily power consumption data come ten days draw it is new ought power consumption data month in and month out, according to etc. dimension fill vacancies in the proper order the concept of gray model, use
New moon power consumption data calculate the moon power consumption data of next month, finally draw a moon electricity demand forecasting result.Including following step
Suddenly：
2-1) calculates smoothing factor, is calculated monthly with the average moon power consumption divided by average year power consumption of the first three years monthly
Smoothing factor.
2-2) predictions ought power consumption month in and month out, utilize the history moon power consumption data structure forecasting sequence x of 12 months(0), use
The mode of month power consumption divided by smoothing factor handles forecasting sequence.Pass through x(0)Grey GM (1,1) model formation prediction ought be month in and month out
Power consumption, prediction result be multiplied by this month smoothing factor reduced after cross ought electricity demand forecasting result month in and month out.Gray model
Formula is as follows：
Wherein, k=1,2,3 ... 11 represent k-th month；x(0)For forecasting sequence；A is development coefficient；U acts on for grey
Amount；；A, u is the differential equationCoefficient.
2-3) predicts next power consumption month in and month out.The daily power consumption result in the ten days futures predicted using step 1) is carried out
Cumulative, with the electricity and replacement step 2-2 after cumulative) in of that month electricity demand forecasting result phase same date electricity, formed new
Of that month electricity demand forecasting result, according to etc. the concept filled vacancies in the proper order of dimension, the of that month electricity demand forecasting result after adjustment is added smooth
The history 12 of the before processing forecasting sequence that power consumption forms month in and month out simultaneously removes first data in former forecasting sequence, then,
New forecasting sequence is handled (moon power consumption divided by smoothing factor) with smoothing factor, and it is pre- to resettle grey GM (1,1) model
The moon power consumption of next month is surveyed, the moon power consumption of next month to predicting is smoothed to obtain final prediction knot again
Fruit.
By taking certain province as an example, industry-specific moon electricity demand forecasting in 2016 is carried out, prediction steps are as follows：
A, the moon power consumption of this month and next month are predicted for timing node with No. 1 of every month.No. 1 to No. 10 is calculated respectively
Day prediction electricity.The predicted value of the maximum temperature for predicting day, minimum temperature, mean temperature and humidity is obtained first, in history day
The middle meteorological factor coefficient that history day is calculated according to formula (1).The time factor of history day is calculated by formula (3), then is passed through
Formula (4) calculates the week factor of history day.By the meteorological factor of the history day calculated, time factor and week factor phase
Multiply, calculate the similarity factor of history day, the history day of 5 before similarity factor is taken as similar day set, to the day of similar day
Power consumption is used as the daily power consumption of prediction day after being weighted using similarity factor as weight averagely.
B, the prediction daily power consumption of No. 1 to No. 10 is obtained by previous step, then daily power consumption summed, obtain future 10
The sum of its daily power consumption.
C, it is calculating sample to take certain provincial and autonomous regional branches industry 2013, the moon power consumption of 2014,2,015 3, is calculated respectively industry-specific
The average moon power consumption and average year power consumption of every month, then obtained with average moon power consumption divided by average year power consumption each
Month industry-specific smoothing factor.
D, using predicted month as timing node, the moon power consumption for taking first 12 months is forecasting sequence, such as predicts in January, 2016
The moon power consumption then forecasting sequence be in January, 2015 to December moon power consumption.The moon power consumption number in forecasting sequence is used first
According to divided by industry-specific smoothing factor every month, then according to gray model formula (5) predict this month moon power consumption, finally
The smoothing factor that this month the sector is multiplied by with prediction result obtains final of that month moon electricity demand forecasting result, will predict early stage
10 days futures daily power consumption sum replace the electricity that correspond to date in of that month electricity demand forecasting result and, form new of that month use
Power quantity predicting result.
E, last data in forecasting sequence are replaced with of that month electricity demand forecasting result, rejects the in forecasting sequence
One data, thus form a new forecasting sequence being made up of 12 data.Forecasting sequence is carried out in the same fashion
Smoothing processing, the moon power consumption result of next month is predicted by formula (5), then is obtained after being smoothed to prediction result
Final secondary electricity demand forecasting value month in and month out.
Fig. 4 is certain whole society of province moon electricity demand forecasting curve in 2016 and actual moon power consumption curve, it can be seen that prediction
As a result it is closer to actual result.
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, some improvement and deformation can also be made, these are improved and deformation
Also it should be regarded as protection scope of the present invention.
Claims (8)
1. a kind of gray model power predicating method based on day character vector optimization, it is characterised in that comprise the following steps：
1) following ten days daily power consumption data since being predicted day are calculated according to day character vector method；
2) of that month moon power consumption is predicted by gray model algorithm；
3) based on etc. dimension fill vacancies in the proper order with gray model algorithm prediction next month moon power consumption.
2. a kind of gray model power predicating method based on day character vector optimization according to claim 1, its feature
It is, the computational methods of step 1) the daily power consumption data are：
1-1) calculate meteorological factor, calculate respectively each history max. daily temperature, minimum temperature, mean temperature and humidity four it is sub- because
Element and the matching factor of prediction day, calculation formula are as follows：
<mrow>
<msub>
<mi>s</mi>
<mi>j</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>min</mi>
<mi>j</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&rho;max</mi>
<mi>j</mi>
</msub>
</mrow>
<mrow>
<msub>
<mi>value</mi>
<mi>j</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&rho;max</mi>
<mi>j</mi>
</msub>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, SjFor the matching factor of j-th of sub- factor, j=1,2,3,4, maximum temperature, minimum temperature, average temperature are represented respectively
Degree and humidity, minjDay and the least absolute value of history day difference, max are predicted for j-th of sub- factorjFor j-th of sub- factor prediction
Day and the maximum value of history day difference, valuejDay and the absolute value of history day difference are predicted for j-th of sub- factor, ρ is
Regulation coefficient；
Then weighted average is done to the matching factor of every sub- factor and obtains meteorological factor T：
<mrow>
<mi>T</mi>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mn>4</mn>
</mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>S</mi>
<mn>1</mn>
</msub>
<mo>+</mo>
<msub>
<mi>S</mi>
<mn>2</mn>
</msub>
<mo>+</mo>
<msub>
<mi>S</mi>
<mn>3</mn>
</msub>
<mo>+</mo>
<msub>
<mi>S</mi>
<mn>4</mn>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
Time factor D 1-2) is calculated, calculation formula is as follows：
<mrow>
<mi>D</mi>
<mo>=</mo>
<msup>
<msub>
<mi>&beta;</mi>
<mn>1</mn>
</msub>
<mrow>
<mi>mod</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>,</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<msup>
<msub>
<mi>&beta;</mi>
<mn>2</mn>
</msub>
<mrow>
<mi>int</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>/</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, β1For day attenuation coefficient, β2For all attenuation coefficients, t is the date differences of prediction day and history day, and mod is remainder letter
Number, int is bracket function；
The week factor W 1-3) is calculated, calculation formula is as follows：
W=1- | f (X1)-f(X0)| (4)
Wherein, f (X1) be history day week type, f (X0) for prediction day week type, X1Value be Monday, Tuesday,
Wednesday, Thursday, Friday, Saturday, Sunday, f (X1) value be：F (Monday)=0.1, f (Tuesday)=f (Wednesday)=f (Thursday)
=0.2, f (Friday)=0.3, f (Saturday)=0.7, f (Sunday)=1；1-4) calculate analogue forecasting method, using meteorological factor,
Time factor, the mode of week fac-tor obtain analogue forecasting method, take maximum preceding 5 conducts of analogue forecasting method similar
Day set, the daily power consumption of similar day is weighted using similarity factor as weight it is average after be used as the day electricity consumption for predicting day
Amount.
3. a kind of gray model power predicating method based on day character vector optimization according to claim 2, its feature
It is, the history day is to predict the date in 35 days a few days ago.
4. a kind of gray model power predicating method based on day character vector optimization according to claim 2, its feature
It is, the ρ values are 0.5.
5. a kind of gray model power predicating method based on day character vector optimization according to claim 2, its feature
It is, the β1Value is 0.96, β2Value is 0.94.
6. a kind of gray model power predicating method based on day character vector optimization according to claim 1, its feature
Be, the step 2) prediction it is of that month the moon power consumption process it is as follows：
Smoothing factor 2-1) is calculated, monthly smooth is calculated with the first three years average moon power consumption divided by average year power consumption monthly
Coefficient；
2-2) utilize the history moon power consumption data structure forecasting sequence x of 12 months(0), with moon power consumption divided by smoothing factor
Mode is smoothed to forecasting sequence；
2-3) ought power consumption month in and month out by the prediction of grey GM (1,1) model formation；
2-4) prediction result be multiplied by this month smoothing factor reduced after ought electricity demand forecasting result month in and month out.
7. a kind of gray model power predicating method based on day character vector optimization according to claim 6, its feature
Be, the step 3) prediction next month the moon power consumption process it is as follows：
3-1) the daily power consumption result in ten days futures predicted using step 1) is added up, with the electricity after cumulative and replacement
Step 2-2) in of that month electricity demand forecasting result phase same date electricity, form new of that month electricity demand forecasting result；
3-2) according to etc. the concept filled vacancies in the proper order of dimension, by after adjustment ought electricity demand forecasting result is added before smoothing processing month in and month out history
12 forecasting sequences that power consumption forms month in and month out simultaneously remove first data in former forecasting sequence, obtain new forecasting sequence；
3-3) new forecasting sequence is smoothed；
The moon power consumption of next month 3-4) is predicted by grey GM (1,1) model formation；
3-5) prediction result be multiplied by this month smoothing factor reduced after next month moon electricity demand forecasting result.
8. a kind of gray model power predicating method based on day character vector optimization according to claim 6 or 7, it is special
Sign is that grey GM (1, the 1) model formation is：
<mrow>
<msup>
<mover>
<mi>x</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>a</mi>
</mrow>
</msup>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msup>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
<mo>-</mo>
<mfrac>
<mi>u</mi>
<mi>a</mi>
</mfrac>
<mo>)</mo>
</mrow>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>a</mi>
<mi>k</mi>
</mrow>
</msup>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, k=1,2,3 ... 11 represent k-th month, x(0)For forecasting sequence,Represent that the kth moon of+1 month is used
Power quantity predicting result, x(0)(1) moon power consumption of 1st month is represented, a is development coefficient, and u is grey actuating quantity, and a, u are micro-
Divide equationCoefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710702936.2A CN107563544A (en) | 2017-08-16 | 2017-08-16 | A kind of gray model power predicating method based on day character vector optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710702936.2A CN107563544A (en) | 2017-08-16 | 2017-08-16 | A kind of gray model power predicating method based on day character vector optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107563544A true CN107563544A (en) | 2018-01-09 |
Family
ID=60974490
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710702936.2A Pending CN107563544A (en) | 2017-08-16 | 2017-08-16 | A kind of gray model power predicating method based on day character vector optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107563544A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113191574A (en) * | 2021-05-28 | 2021-07-30 | 上海申瑞继保电气有限公司 | Daily electricity prediction method for single product production line |
| CN116937752A (en) * | 2023-09-14 | 2023-10-24 | 广州德姆达光电科技有限公司 | Charging and discharging control method for outdoor mobile energy storage power supply |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825040A (en) * | 2015-12-29 | 2016-08-03 | 海南电力技术研究院 | Short-term power load prediction method |
| CN106447534A (en) * | 2016-09-22 | 2017-02-22 | 国网上海市电力公司 | Method for determining stability of power mode based on gray relational analysis |
-
2017
- 2017-08-16 CN CN201710702936.2A patent/CN107563544A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825040A (en) * | 2015-12-29 | 2016-08-03 | 海南电力技术研究院 | Short-term power load prediction method |
| CN106447534A (en) * | 2016-09-22 | 2017-02-22 | 国网上海市电力公司 | Method for determining stability of power mode based on gray relational analysis |
Non-Patent Citations (3)
| Title |
|---|
| DAO_1990: "预测之相似日算法", 《HTTPS://BLOG.CSDN.NET/DAO_1990/ARTICLE/DETAILS/50269205》 * |
| 李春涛等: "基于改进灰色模型的短期负荷预测", 《电气开关》 * |
| 段锋等: "两种改进的电力负荷灰色预测模型的应用", 《科学技术与工程》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113191574A (en) * | 2021-05-28 | 2021-07-30 | 上海申瑞继保电气有限公司 | Daily electricity prediction method for single product production line |
| CN116937752A (en) * | 2023-09-14 | 2023-10-24 | 广州德姆达光电科技有限公司 | Charging and discharging control method for outdoor mobile energy storage power supply |
| CN116937752B (en) * | 2023-09-14 | 2023-12-26 | 广州德姆达光电科技有限公司 | Charging and discharging control method for outdoor mobile energy storage power supply |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Imani et al. | Electrical load forecasting using customers clustering and smart meters in Internet of Things | |
| Mandal et al. | A neural network based several-hour-ahead electric load forecasting using similar days approach | |
| Khan et al. | Genetic algorithm based optimized feature engineering and hybrid machine learning for effective energy consumption prediction | |
| US7403928B2 (en) | Identify data sources for neural network | |
| Panagopoulos | Assessing the impacts of socio-economic and hydrological factors on urban water demand: A multivariate statistical approach | |
| CN103002164A (en) | Telephone traffic forecasting method of electric power call center | |
| JP2016181060A (en) | Electric power demand prediction system | |
| CN113869713A (en) | Optimal scheduling method and system for mobile charging vehicle | |
| Liu et al. | Daily tourism demand forecasting: the impact of complex seasonal patterns and holiday effects | |
| CN107563544A (en) | A kind of gray model power predicating method based on day character vector optimization | |
| CN115481918A (en) | Active sensing and predictive analysis system for unit state based on source network load storage | |
| Mellios et al. | A multivariate analysis of the daily water demand of Skiathos Island, Greece, implementing the artificial neuro-fuzzy inference system (ANFIS) | |
| CN107895211A (en) | A kind of long-medium term power load forecasting method and system based on big data | |
| Mey et al. | Prediction of energy consumption for variable customer portfolios including aleatoric uncertainty estimation | |
| Zheng et al. | Short-term energy consumption prediction of electric vehicle charging station using attentional feature engineering and multi-sequence stacked Gated Recurrent Unit | |
| CN115952918B (en) | Ordered power usage pattern generation method and system for novel power load management | |
| CN110659775A (en) | LSTM-based improved electric power short-time load prediction algorithm | |
| Souzanchi-K et al. | A multi adaptive neuro fuzzy inference system for short term load forecasting by using previous day features | |
| CN106779241B (en) | Short-term passenger flow prediction method for rail transit | |
| Prakash et al. | Ensemble Method for Short-Term Load Forecasting Using LSTM, SVR, and FFNN Taking into Account Seasonal Dependency. | |
| JPH11126102A (en) | Predictive method for demand of gas | |
| JP2000003353A (en) | Water distribution prediction method | |
| Salleh et al. | Experiment on electricity consumption prediction using long short-term memory architecture on residential electrical consumer | |
| CN110430328A (en) | Call center's telephone traffic prediction method and system based on LightGBM model | |
| Steens et al. | A forecast based load management approach for commercial buildings--comparing LSTM and standardized load profile techniques |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180109 |
|
| RJ01 | Rejection of invention patent application after publication |