CN107273997B - Method and system for predicting daily electricity consumption of platform area based on frequency coefficient linear regression model - Google Patents
Method and system for predicting daily electricity consumption of platform area based on frequency coefficient linear regression model Download PDFInfo
- Publication number
- CN107273997B CN107273997B CN201710280886.3A CN201710280886A CN107273997B CN 107273997 B CN107273997 B CN 107273997B CN 201710280886 A CN201710280886 A CN 201710280886A CN 107273997 B CN107273997 B CN 107273997B
- Authority
- CN
- China
- Prior art keywords
- day
- week
- standard
- prediction
- linear regression
- 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
- 230000005611 electricity Effects 0.000 title claims abstract description 74
- 238000012417 linear regression Methods 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000006467 substitution reaction Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000005612 types of electricity Effects 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Abstract
The invention relates to a method for predicting the daily electricity consumption of a platform area based on a day of the week coefficient linear regression model, which comprises the following steps of establishing a day of the week proportion coefficient group K of the platform area according to a standard week; obtaining the corrected electricity consumption of the history day according to the week-day proportion coefficient group K; establishing an n-ary linear regression model for the influence factors; obtaining historical meteorological data close to a prediction day through inquiring a meteorological database; substituting the historical meteorological data obtained in the step S2 close to the prediction day into a formula; fitting the linear regression equation by adopting a least square method formula, and calculating each constant in the linear regression equation; substituting the constant and the influence factor of the prediction day into the n-ary linear regression model to obtain the correction electricity consumption of the prediction day, finally, the power consumption of the correction day of the prediction day is substituted into the coefficient group K of the week proportion to obtain the predicted consumption of the prediction day.
Description
Technical Field
The invention relates to a method and a system for predicting daily electricity consumption of a platform area based on a day of the week coefficient linear regression model, and belongs to the technical field of power grid electricity consumption.
Background
The power consumption prediction is a key basis for making a comprehensive production plan and making an operation plan by a power grid company, a reasonable and accurate prediction conclusion can bring forward effect to the operation decision of the company, otherwise, the deviation of the operation strategy of the company can be caused, and therefore the power consumption prediction is very important for future season or year. As can be seen from an overview of domestic and foreign market prediction technologies, the existing electricity consumption prediction technologies can be classified into three types, but the key problems of electricity consumption prediction cannot be solved.
The first type of electricity consumption prediction technology is to extrapolate trend according to actual occurrence values of historical electricity consumption, and the information contained in the prediction conclusion is a development mode of the last period of the economic environment lineage based on the prediction period, for example, chinese patent 101976301. However, if the electricity consumption situation in the prediction period is changed greatly or the direction of the electricity consumption situation turns around, the method cannot be predicted, so that the prediction conclusion of the method is often larger than the deviation which actually occurs in the current economic instability period.
The second type of electricity consumption prediction technology is to judge the growth amplitude of the predicted year based on the experience of the predicted personnel, the predicted personnel can develop prediction according to the current economic situation and the own prediction experience, the judgment of the economic situation is limited to a qualitative analysis level and cannot be quantified on a specific prediction model, the predicted experience growth is more dependent on the individual comprehensive judgment capability of the predicted personnel, and the reliability of the predicted result cannot be effectively ensured.
A third type of electricity consumption prediction technique is to extrapolate the actual occurrence of the historical electricity consumption using different algorithms, the problem of electricity consumption prediction is solved algorithmically. However, the existing algorithm is complex, and the accuracy of the prediction of the power consumption is not enough.
Disclosure of Invention
The invention aims to solve the technical problems that the prediction algorithm is complex, experience cannot be quantified on a specific prediction model, and the prediction result has large deviation and low accuracy.
In order to solve the problems, the method and the system for predicting the daily electricity consumption of the district based on the frequency coefficient linear regression model are provided, and the daily electricity consumption of the district is predicted by establishing the frequency coefficient modified linear regression model, so that the method has the advantages of simplicity and accuracy in prediction.
The invention solves the technical problems as follows:
a prediction method of the daily electricity consumption of a platform area based on a day of the week coefficient linear regression model comprises the following steps:
s1, searching for a standard week from near to far in the history day;
s2, acquiring a coefficient of proportionality K of the week and the day according to the 7-day daily electricity consumption of the standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
s3, using the actual daily electricity consumption D of the area history i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D' i ＝D i /K(i) i＝1,2,...,7 (2)；
s4, correcting value D of actual power consumption according to historical day i ' number n of influence factors of daily electricity, and establishing an n-element linear regression model related to the influence factors, namely:
D'＝a 1 *T+a 2 *Tmax+a 3 *Tmin+a 4 *RH+a 5 *V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
s5, obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through inquiring a meteorological database, wherein i-1 represents the day before the predicted day, and so on;
s6, substituting the historical meteorological data obtained in the step S5 close to the prediction day into a formula (3), namely:
s7, fitting the linear regression equation (4) by adopting a least square method formula, and calculating each constant in the linear regression equation (4), namely a 1 、a 2 、a 3 、a 4 、a 5 、b；
S8, a is carried out 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; i.e. the method comprises the following steps:
s9, correcting the cycle proportion coefficient of the corrected daily electricity consumption of the predicted day, and respectively multiplying the cycle proportion coefficient by the corresponding cycle proportion coefficient to obtain the predicted electricity consumption, namely:
as a further improvement of the present invention, further, the search conditions in S1 for the standard week are: (1) The 7 days of the week are in a comfortable zone, namely T max i ＜28&&T min i Not less than 5,i =1 to 7; (2) 7 days of the week plus 2 days before and 2 days after the week for a total of 11 daysAre non-holidays.
As a further improvement of the invention, the number of the influence factors of the n-ary linear regression model in S4 is more than or equal to 1.
As a further improvement of the present invention, further, the number of days for which queries for historical weather data near the predicted day are required is equal to the number of influencing factors minus 1.
The invention also provides a prediction system of the daily electricity consumption of the platform area based on the coefficient linear regression model of the week, which comprises,
a standard week searching unit for searching for a standard week from near to far in the history day;
the day of the week proportionality coefficient obtaining unit is used for obtaining a day of the week proportionality coefficient K according to 7 days of electricity consumption of a standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
a correction unit for the actual daily power consumption of the area for correcting the actual daily power consumption D of the area i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D' i ＝D i /K(i) i＝1,2,...,7 (2)；
a model building unit for correcting the value D of the actual electricity consumption according to the history day i ' number n of influence factors of daily electricity, and establishing an n-element linear regression model related to the influence factors, namely:
D'＝a1*T+a2*Tmax+a 3 *Tmin+a 4 *RH+a5*V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
the query unit is used for obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through a meteorological database, wherein i-1 represents the day before the predicted day, and the like;
a constant substitution unit of the model for substituting the historical meteorological data obtained by S2 close to the prediction day into formula (3), namely:
a constant solving unit of the model for fitting the linear regression equation (4) by using a least square method formula to calculate each constant, namely a, in the linear regression equation (4) 1 、a 2 、a 3 、a 4 、a 5 、b；
A correction electricity consumption solving unit for solving the electricity consumption of a 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; namely:
the prediction electricity consumption solving unit is used for carrying out coefficient correction on the correction daily electricity consumption of the prediction day, and obtaining the prediction electricity consumption according to the coefficient of the corresponding coefficient of the cycle, namely:
as a further improvement of the present invention, further, the search conditions of the standard week search unit are: (1) The 7 days of the week are in a region where the human body is more comfortable, i.e. Tmax i ＜28&&Tmin i Not less than 5,i =1 to 7; (2) The 7 days of the week plus 2 days before and 2 days after the week, for a total of 11 days, are non-holidays.
As a further improvement of the invention, the number of the influence factors of the n-ary linear regression model in the model building unit is more than or equal to 1.
As a further improvement of the present invention, further, the number of days for which queries for historical weather data near the predicted day are required is equal to the number of influencing factors minus 1.
In summary, the technical scheme of the invention mainly has the following advantages: the method is simple and has high accuracy.
Detailed Description
The invention discloses a method for predicting the daily electricity consumption of a platform area based on a day of the week coefficient linear regression model, which comprises the following steps:
s1, searching for a standard week from near to far in the history day;
s2, acquiring a coefficient of proportionality K of the week and the day according to the 7-day daily electricity consumption of the standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
s3, using the actual daily electricity consumption D of the area history i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D' i ＝D i /K(i) i＝1,2,...,7 (2)；
s4, correcting value D of actual power consumption according to historical day i ' number n of influence factors of daily electricity, and establishing an n-element linear regression model related to the influence factors, namely:
D'＝a 1 *T+a 2 *Tmax+a 3 *Tmin+a 4 *RH+a 5 *V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
s5, obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through inquiring a meteorological database, wherein i-1 represents the day before the predicted day, and so on;
s6, substituting the historical meteorological data obtained in the step S2 close to the prediction day into a formula (3), namely:
s7, fitting the linear regression equation (4) by adopting a least square method formula, and calculating each constant in the linear regression equation (4), namely a 1 、a 2 、a 3 、a 4 、a 5 、b；
S8, a is carried out 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; namely:
s9, correcting the cycle proportion coefficient of the corrected daily electricity consumption of the predicted day, and respectively multiplying the cycle proportion coefficient by the corresponding cycle proportion coefficient to obtain the predicted electricity consumption, namely:
the search conditions for the standard week in S1 were: (1) The 7 days of the week are in a region where the human body is more comfortable, i.e. Tmax i ＜28&&Tmin i Not less than 5,i =1 to 7; (2) The 7 days of the week plus 2 days before and 2 days after the week, for a total of 11 days, are non-holidays.
And S4, the number of the influence factors of the n-ary linear regression model is more than or equal to 1.
The number of days required for queries of historical meteorological data near the predicted day is equal to the number of influencing factors minus 1.
The invention also provides a prediction system of the daily electricity consumption of the platform area based on the coefficient linear regression model of the week, which comprises,
a standard week searching unit for searching for a standard week from near to far in the history day;
the day of the week proportionality coefficient obtaining unit is used for obtaining a day of the week proportionality coefficient K according to 7 days of electricity consumption of a standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
a correction unit for the actual daily power consumption of the area for correcting the actual daily power consumption D of the area i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D' i ＝D i /K(i) i＝1,2,...,7 (2)；
a model building unit for correcting the value D of the actual electricity consumption according to the history day i ' number n of influence factors of daily electricity, and establishing an n-element linear regression model related to the influence factors, namely:
D'＝a 1 *T+a 2 *Tmax+a 3 *Tmin+a 4 *RH+a 5 *V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
the query unit is used for obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through a meteorological database, wherein i-1 represents the day before the predicted day, and the like;
a constant substitution unit of the model for substituting the historical meteorological data obtained by S2 close to the prediction day into formula (3), namely:
a constant solving unit of the model for fitting the linear regression equation (4) by using a least square method formula to calculate each constant, namely a, in the linear regression equation (4) 1 、a 2 、a 3 、a 4 、a 5 、b；
A correction electricity consumption solving unit for solving the electricity consumption of a 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; namely:
the prediction electricity consumption solving unit is used for carrying out coefficient correction on the correction daily electricity consumption of the prediction day, and obtaining the prediction electricity consumption according to the coefficient of the corresponding coefficient of the cycle, namely:
the search conditions of the standard week in the standard week search unit are as follows: (1) The 7 days of the week are in a region where the human body is more comfortable, i.e. Tmax i ＜28&&Tmin i Not less than 5,i =1 to 7; (2) The 7 days of the week plus 2 days before and 2 days after the week, for a total of 11 days, are non-holidays.
The number of influence factors of the n-ary linear regression model in the model building unit is greater than or equal to 1.
The number of days required for queries of historical meteorological data near the predicted day is equal to the number of influencing factors minus 1.
In this embodiment, taking a certain distribution area (13 groups in the river and sea, area number: 14000000014460) in the south China as an example, the daily electricity (unit: kilowatt-hour) is shown in table 1:
TABLE 1 historical power consumption and meteorological data for a bay
From table 1, the standard week was found to be one satisfying S1 from month 3, 26, 2016 to month 4, 1, which satisfies the following two points:
(1) The holidays of Qingming festival are 2016, 4 months and 4 months, and the holidays of Qingming and 2016, 4 months and 1 month are two days different. Two days before the month of 2016, 3 and 26, there is no holiday.
(2) The highest temperature and the lowest temperature of the heat pump water heater from 26 days in 2016 to 4 days in 2016 meet T max i ＜28&&T min i ≥5。
As can be seen from the actual electricity consumption of the standard week 2016, month 3, 26 and month 1 of 2016, d1=208.78, d2=210.64, d3=197.73, d4=204.32, d5=198.86, d6= 226.66, and d7= 223.32. The data are taken into formula (1) to obtain the coefficient of proportionality K of the circumference.
K＝[1,1.0089,0.974,0.9786,0.9525,1.0856,1.0696]
As is clear from Table 1, the actual electric power amounts from month 4, month 7 and month 4, month 11 of 2016 are 211.4167,202.3586,230.5163,228.2099,210.9300, respectively, and the actual electric power amount correction values are obtained by substituting them into (2)216.04, 212.45, 212.34, 213.36, 210.93. Substituting the data of the correction value into equation (4), and performing least square formula fitting on the equation to calculate each constant in the linear regression equation (4) as follows: a, a 1 ＝18.3892,a 2 ＝-8.0752,a 3 ＝-5.4459,a 4 ＝2.3399,a 5 =9.5635, b= 2.6151, as follows:
| date of day | Week table | Average temperature | Maximum temperature | Minimum temperature | Humidity of the water | Wind speed |
| 2016/4/12 | 2 | 14.1 | 20.7 | 9.2 | 73.4 | 1.2 |
| 2016/4/13 | 3 | 9.9 | 13.7 | 6.3 | 71 | 0.7 |
| 2016/4/14 | 4 | 11.1 | 16.7 | 6.2 | 71.3 | 0.5 |
| 2016/4/15 | 5 | 16.3 | 21.4 | 9.1 | 54.1 | 0.3 |
| 2016/4/16 | 6 | 15.2 | 21.7 | 9.6 | 64.9 | 1.2 |
TABLE 2 predicted Meteorological data for the predicted day
Bringing the predicted data and constants in table 2 into equation (5) yields:
substituting the obtained data into equation (6) to obtain a predicted value:
the actual values of the power consumption of the areas from 12 days 4 to 16 days 2016 are as follows:
comparing the predicted value with the true value, the relative error is within 5%, and the predicted requirement is met.
In addition to the embodiments described above, other embodiments of the invention are possible. All technical schemes formed by equivalent substitution or equivalent transformation fall within the protection scope of the invention.
Claims (6)
1. The method for predicting the daily electricity consumption of the platform area based on the day of the week coefficient linear regression model is characterized by comprising the following steps of:
s1, searching for a standard week from near to far in the history days "
S2, acquiring a coefficient of proportionality K of the week and the day according to the 7-day daily electricity consumption of the standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
s3, using the actual daily electricity consumption D of the area history i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D' i ＝D i /K(i) i＝1,2,...,7 (2)；
s4, correcting value D of actual power consumption according to historical day i ' number n of influence factors of daily electricity, and establishing an n-element linear regression model related to the influence factors, namely:
D'＝a 1 *T+a 2 *Tmax+a 3 *Tmin+a 4 *RH+a 5 *V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
s5, obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through inquiring a meteorological database, wherein i-1 represents the day before the predicted day, and so on;
s6, substituting the historical meteorological data obtained in the step S5 close to the prediction day into a formula (3), namely:
s7, fitting the linear regression equation (4) by adopting a least square method formula, and calculating each constant in the linear regression equation (4), namely a 1 、a 2 、a 3 、a 4 、a 5 、b；
S8, a is carried out 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; namely:
s9, correcting the cycle scale factor of the corrected daily electricity consumption of the predicted day, and respectively multiplying the corrected daily electricity consumption corresponding to the cycle by the corresponding cycle scale factor to obtain the predicted electricity consumption, namely:
the searching conditions in the standard week in the S1 are as follows: (1) The 7 days of the week are in a region where the human body is more comfortable, i.e. Tmax i ＜28&&Tmin i Not less than 5,i =1 to 7; (2) 7 days of the week plus 2 days before and 2 days after the week, total 11 days were non-holidays.
2. The method for predicting the daily electricity consumption of a district based on a day of the week coefficient linear regression model according to claim 1, wherein the number of the influence factors of the n-ary linear regression model in S4 is 1 or more.
3. The method for predicting the daily electricity consumption of a district based on a day of the week coefficient linear regression model according to claim 1, wherein the number of days for which the historical meteorological data near the predicted day is required to be queried is equal to the number of influence factors minus 1.
4. The system for predicting the daily electricity consumption of the area based on the day of the week coefficient linear regression model is characterized by comprising,
a standard week searching unit for searching for a standard week from near to far in the history day;
the day of the week proportionality coefficient obtaining unit is used for obtaining a day of the week proportionality coefficient K according to 7 days of electricity consumption of a standard week, namely:
K＝[1,D2/D1,D3/D1,D4/D1,D5/D1,D6/D1,D7/D1] (1)
wherein, D1 is the power consumption of the standard Monday area, D2 is the power consumption of the standard Monday area, D3 is the power consumption of the standard Monday area, D4 is the power consumption of the standard Monday area, D5 is the power consumption of the standard Monday area, D6 is the power consumption of the standard Monday area, and D7 is the power consumption of the standard Monday area;
a correction unit for the actual daily power consumption of the area for correcting the actual daily power consumption D of the area i Correction value D of actual power consumption of historical day is obtained according to coefficient correction of proportion of the week to the day i ' i.e.:
D'i＝Di/K(i) i＝1,2,...,7 (2)；
a model building unit for correcting the value D of the actual electricity consumption according to the history day i ' number n of influencing factors of daily electricity, an n-ary linear regression model is built on the influence factors, namely:
D'＝a 1 *T+a 2 *Tmax+a 3 *Tmin+a 4 *RH+a 5 *V+b (3)
wherein the variables are respectively: average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V;
the query unit is used for obtaining historical meteorological data of average air temperature T, maximum air temperature Tmax, minimum air temperature Tmin, humidity RH and wind speed V which are close to a predicted day through a meteorological database, wherein i-1 represents the day before the predicted day, and the like;
a constant substitution unit of the model for substituting historical weather data obtained by the weather database near the prediction day into formula (3), namely:
a constant solving unit of the model for fitting the linear regression equation (4) by using a least square method formula to calculate each constant, namely a, in the linear regression equation (4) 1 、a 2 、a 3 、a 4 、a 5 、b；
A correction electricity consumption solving unit for solving the electricity consumption of a 1 、a 2 、a 3 、a 4 、a 5 The influence factors of the prediction date, b and the prediction date are substituted into the n-element linear regression model (3) to obtain the correction electricity consumption of the prediction dateWherein i represents a prediction day, i+1 represents a day after the prediction day, and so on; namely:
the prediction electricity consumption solving unit is used for carrying out cycle number proportionality coefficient correction on the correction daily electricity consumption of the prediction day, and respectively multiplying the corresponding cycle number proportionality coefficients according to the correction daily electricity consumption corresponding to the cycle number to obtain the prediction electricity consumption, namely:
the search conditions of the standard week in the standard week search unit are as follows: (1) The 7 days of the week are in a region where the human body is more comfortable, i.e. Tmax i ＜28&&Tmin i Not less than 5,i =1 to 7; (2) The 7 days of the week plus 2 days before and 2 days after the week, for a total of 11 days, are non-holidays.
5. The prediction system for daily electricity consumption of a district based on a day of the week coefficient linear regression model according to claim 4, wherein the number of influence factors of the n-ary linear regression model in the model establishing unit is 1 or more.
6. The day of the week coefficient linear regression model based prediction system of claim 4 wherein the number of days for the query of the historical weather data near the prediction day is equal to the number of influence factors minus 1.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2016105074462 | 2016-06-30 | ||
| CN201610507446.2A CN106447067A (en) | 2016-06-30 | 2016-06-30 | Court daily power consumption prediction method based on adaptive nonlinear regression model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107273997A CN107273997A (en) | 2017-10-20 |
| CN107273997B true CN107273997B (en) | 2024-02-02 |
Family
ID=58183793
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610507446.2A Pending CN106447067A (en) | 2016-06-30 | 2016-06-30 | Court daily power consumption prediction method based on adaptive nonlinear regression model |
| CN201710280886.3A Active CN107273997B (en) | 2016-06-30 | 2017-04-26 | Method and system for predicting daily electricity consumption of platform area based on frequency coefficient linear regression model |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610507446.2A Pending CN106447067A (en) | 2016-06-30 | 2016-06-30 | Court daily power consumption prediction method based on adaptive nonlinear regression model |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN106447067A (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222745A (en) * | 2019-11-20 | 2020-06-02 | 黑龙江电力调度实业有限公司 | Power utilization scheduling system and method |
| CN116701887B (en) * | 2023-08-07 | 2023-11-07 | 河北思极科技有限公司 | Power consumption prediction method and device, electronic equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10224990A (en) * | 1997-02-10 | 1998-08-21 | Fuji Electric Co Ltd | Method for correcting predicted value of electric power demand |
| CN103440556A (en) * | 2013-09-04 | 2013-12-11 | 国家电网公司 | Electricity consumption prediction method based on economic conduction |
| CN104598985A (en) * | 2014-12-12 | 2015-05-06 | 国家电网公司 | Power load forecasting method |
| CN104616079A (en) * | 2015-02-04 | 2015-05-13 | 国家电网公司 | Temperature change based power grid daily electricity consumption prediction method |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| WO2016031065A1 (en) * | 2014-08-29 | 2016-03-03 | 三菱電機株式会社 | Power consumption estimation device, appliance management system, power consumption estimation method, and program |
| CN105574607A (en) * | 2015-12-10 | 2016-05-11 | 四川省电力公司供电服务中心 | Electricity market monthly electricity utilization prediction method |
| CN105590174A (en) * | 2015-12-29 | 2016-05-18 | 南京因泰莱电器股份有限公司 | Enterprise power consumption load prediction method based on K-means clustering RBF neural network |
| CN105678407A (en) * | 2015-12-31 | 2016-06-15 | 国网上海市电力公司 | Daily electricity consumption prediction method based on artificial neural network |
-
2016
- 2016-06-30 CN CN201610507446.2A patent/CN106447067A/en active Pending
-
2017
- 2017-04-26 CN CN201710280886.3A patent/CN107273997B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10224990A (en) * | 1997-02-10 | 1998-08-21 | Fuji Electric Co Ltd | Method for correcting predicted value of electric power demand |
| CN103440556A (en) * | 2013-09-04 | 2013-12-11 | 国家电网公司 | Electricity consumption prediction method based on economic conduction |
| WO2016031065A1 (en) * | 2014-08-29 | 2016-03-03 | 三菱電機株式会社 | Power consumption estimation device, appliance management system, power consumption estimation method, and program |
| CN104598985A (en) * | 2014-12-12 | 2015-05-06 | 国家电网公司 | Power load forecasting method |
| CN104616079A (en) * | 2015-02-04 | 2015-05-13 | 国家电网公司 | Temperature change based power grid daily electricity consumption prediction method |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| CN105574607A (en) * | 2015-12-10 | 2016-05-11 | 四川省电力公司供电服务中心 | Electricity market monthly electricity utilization prediction method |
| CN105590174A (en) * | 2015-12-29 | 2016-05-18 | 南京因泰莱电器股份有限公司 | Enterprise power consumption load prediction method based on K-means clustering RBF neural network |
| CN105678407A (en) * | 2015-12-31 | 2016-06-15 | 国网上海市电力公司 | Daily electricity consumption prediction method based on artificial neural network |
Non-Patent Citations (2)
| Title |
|---|
| 林小红 ; 夏丽花 ; 黄美金 ; 吴昌叨 ; .福州市夏季电力气象等级预测模型初探.气象科技.2006,(06),全文. * |
| 苗键强 ; 童星 ; 康重庆 ; .考虑相关因素统一修正的节假日负荷预测模型.电力建设.2015,(10),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106447067A (en) | 2017-02-22 |
| CN107273997A (en) | 2017-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107273998B (en) | Air temperature correction method for predicting daily electricity quantity of transformer area | |
| CN105069536B (en) | A kind of power demand prediction technique based on temperature and economic growth | |
| CN109667727B (en) | Wind turbine generator yaw error inherent deviation identification and compensation method based on power curve analysis | |
| CN111507512B (en) | Correction method and device for photovoltaic ultra-short-term predicted power | |
| CN103617565A (en) | Method for establishing relationship between city micro weather and electric power air conditioner loads | |
| CN103617457A (en) | Method for predicting weather sensitivity load related to temperature in electric system | |
| CN107273997B (en) | Method and system for predicting daily electricity consumption of platform area based on frequency coefficient linear regression model | |
| CN109886567B (en) | Short-term load prediction method considering somatosensory temperature and radiation intensity | |
| CN105605666B (en) | The accurate distribution method of heating load and device | |
| CN110729764A (en) | Optimal scheduling method for photovoltaic power generation system | |
| CN109299814B (en) | Meteorological influence electric quantity decomposition prediction method | |
| CN102915387A (en) | Power grid ice region distribution diagram drawing method | |
| CN102509173A (en) | Markov chain based method for accurately forecasting power system loads | |
| CN111008727A (en) | Power distribution station load prediction method and device | |
| CN107169596B (en) | Human comfort-based prediction method for daily electricity consumption of area | |
| CN114266375A (en) | Regional short-term power load peak prediction method based on multi-innovation algorithm | |
| CN109214458A (en) | A kind of city load quantization method based on historical data | |
| CN112493100B (en) | Cotton moisture monitoring drip irrigation control method and system based on soil water potential | |
| CN107256435B (en) | Fixed value correction method based on predicted value of daily electricity quantity of station area | |
| CN110991938B (en) | Energy consumption calculation method for multi-product production line | |
| CN110598896A (en) | Photovoltaic power prediction method based on prediction error correction | |
| CN116542163A (en) | Historical data driven short-term heating thermal load prediction method | |
| CN115081681A (en) | Prophet algorithm-based wind power prediction method | |
| CN105528735A (en) | Abnormal data point correction method based on measured wind speed and spatial correlation | |
| CN105698412A (en) | Control method of concentrated solar water heating system using weather forecast |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |