CN116432821B - Meteorological condition-based next-day power load peak prediction method - Google Patents

Meteorological condition-based next-day power load peak prediction method Download PDF

Info

Publication number
CN116432821B
CN116432821B CN202310217831.3A CN202310217831A CN116432821B CN 116432821 B CN116432821 B CN 116432821B CN 202310217831 A CN202310217831 A CN 202310217831A CN 116432821 B CN116432821 B CN 116432821B
Authority
CN
China
Prior art keywords
daily
load
meteorological
peak value
peak
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
Application number
CN202310217831.3A
Other languages
Chinese (zh)
Other versions
CN116432821A (en
Inventor
曲晓黎
尤琦
王洁
张金满
杨琳晗
赵增保
李文晴
刘浩
周朔
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hebei Meteorological Service Center Hebei Meteorological Film And Television Center
Original Assignee
Hebei Meteorological Service Center Hebei Meteorological Film And Television Center
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hebei Meteorological Service Center Hebei Meteorological Film And Television Center filed Critical Hebei Meteorological Service Center Hebei Meteorological Film And Television Center
Priority to CN202310217831.3A priority Critical patent/CN116432821B/en
Publication of CN116432821A publication Critical patent/CN116432821A/en
Application granted granted Critical
Publication of CN116432821B publication Critical patent/CN116432821B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • 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
    • 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
    • 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
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The invention discloses a method for predicting the peak value of the next daily electrical load based on meteorological conditions, which comprises the following steps: s1, collecting peak value data of power load and meteorological observation data of a corresponding period; s2, calculating the correlation between the daily electricity load peak value and the meteorological element, performing curve fitting and determining a meteorological element response threshold value; s3, selecting weather elements with obvious correlation, and calculating the relative risk of different weather elements to daily electrical load peaks; s4, calculating the peak cumulative variation of the daily electrical load according to the meteorological elements and the peak variation rate of the daily electrical load caused by the meteorological elements; s5, determining a random component of the electricity load of the main holiday according to the peak value historical data of the daily electricity load; and S6, calculating a peak value of the daily electrical load, and carrying out early warning when the peak value of the daily electrical load reaches a certain magnitude compared with the amplitude of the peak value of the daily electrical load.

Description

Meteorological condition-based next-day power load peak prediction method
Technical Field
The invention relates to a power load prediction method, in particular to a method for predicting the peak value of the next-day power load based on meteorological conditions.
Background
The electric power is the pulse of national economy, the accurate and efficient electric load prediction is the basis for reasonably arranging the power generation plan of the power grid, and is also the premise of ensuring the safe and reliable operation of the electric power system.
The temperate continental monsoon climate is characterized in that the climate is dry and windy in spring, hot and rainy in summer and dry and cold in winter. Therefore, in the periods of spring farmland irrigation, summer air conditioning electricity utilization, winter heating and the like, the condition of power load surge often occurs, and great challenges are brought to the power grid dispatching of the power department.
On the premise that industrial electricity, namely economic load is relatively stable, electricity load change caused by meteorological conditions becomes an important consideration in power grid dispatching work. Therefore, the prediction of the power load peak value based on the meteorological conditions is beneficial to the accurate scheduling of the power departments and the arrangement of line maintenance work. However, there is currently no method for effectively assessing and predicting the impact of peak electrical loads based on meteorological conditions.
Disclosure of Invention
The invention aims to provide a method for predicting the peak value of the next daily electricity load based on meteorological conditions so as to provide scientific and effective decision basis for power dispatching of a power grid company.
The purpose of the invention is realized in the following way:
the method for predicting the peak value of the next daily electrical load based on meteorological conditions comprises the following steps:
s1, acquiring daily electricity load peak historical data of a tested region after quality control for at least five years from a tested day from an electric power department;
s2, acquiring weather observation historical data in a period corresponding to a measured region from a weather department, wherein the weather observation historical data comprise daily maximum air temperature, daily minimum air temperature, daily average air temperature, 24h temperature change, daily average relative humidity, daily minimum relative humidity, daily maximum wind speed, daily average wind speed, and accumulated precipitation of each period corresponding to 24h, 48h and 72h at 20-20 hours, and calculating daily temperature and humidity indexes according to daily average air temperature and daily average relative humidity;
s3, calculating the correlation between the daily electrical load peak value and the meteorological element of the measured area according to the acquired daily electrical load peak value historical data, the meteorological observation historical data and the calculated daily temperature and humidity index, and performing curve fitting to determine a meteorological element response threshold;
s4, selecting weather elements with obvious correlation from the correlation between the daily electrical load peak value and the weather elements, and calculating the change rate of the daily maximum electrical load peak value change caused by unit change of different weather elements, namely the relative risk of different weather elements to the daily electrical load peak value;
s5, calculating the accumulated change quantity of the peak value of the next-day power consumption load caused by all weather elements with significance according to the change rate of the peak value change of the daily power consumption load and real-time weather element data, namely the change quantity of the peak value of the next-day power consumption load caused by weather conditions;
s6, taking the median of the daily electricity load peak variation of the daily electricity load on the day of obvious decline and daily electricity load on the day of obvious rise in each holiday as the random component of the daily electricity load on the main holiday in the year;
s7, calculating a peak value of the electricity load of the next day according to the random component of the electricity load of the main holiday, the peak value data of the electricity load of the current day, the observation data of the weather elements of the current day and the forecast data of the weather elements of the next day; and when the calculated peak value of the daily electricity load reaches a set early warning level compared with the variation amplitude of the peak value of the daily electricity load, early warning is carried out.
Further, the temperature and humidity index E in the step S2 of the invention t The calculation mode of (a) is as follows:
wherein T is the daily average air temperature (DEG C), and R is the daily average relative humidity (%).
Further, the specific operation mode of the step S3 of the present invention is:
s3-1, calculating the correlation between different meteorological elements and daily electrical load peaks by adopting a Spearman rank correlation method, namely, for observation data (x) between n pairs of meteorological elements and daily electrical load peaks i ,y i ) (i=1, 2, …, n), the rank is increased from small to large according to the size order of n data of each group of variables, and the repeated data is averaged, so that the rank correlation coefficient r between the meteorological element and the daily electricity load peak value s The method comprises the following steps:
wherein n is the logarithm of the observed data, R i For observing meteorological elements i Rank order, Q i Peak data y of daily electrical load i Rank order of (c);
s3-2, checking rank correlation coefficient r by T test method s The significance of (2) is calculated by the following steps:
wherein t is a calculated value of rank correlation coefficient, and n is a logarithm of observation data.
According to the correlation coefficient check table, determining the meteorological elements which can pass the saliency check of alpha=0.01, and eliminating the meteorological elements which do not pass the saliency check.
S3-3, fitting the peak value of the electric load of the next day with the meteorological element of the current day by using a nonlinear least square method:
L＝aX 3 +bX 2 +cX+d
wherein L is the peak value of the power load of the next day, X is a meteorological element, and a, b, c, d is a fitting coefficient.
And drawing a fitting curve according to a fitting formula, wherein if an inflection point appears in the curve, the curve is used as a response threshold value of the meteorological element to a daily electrical load peak value.
Further, the specific operation mode of the step S4 of the present invention is:
s4-1, after eliminating meteorological elements which do not pass the alpha=0.01 saliency test, taking the peak value of the maximum power load of the next day as a dependent variable, taking the current day observed value of the related meteorological elements which pass the saliency test as an independent variable, and fitting the logarithmic relation between the maximum power load of the day and the meteorological factors as follows:
In[E(L)]＝βX+α
wherein E (L) is an expected value of a daily maximum power load, X is a meteorological element, alpha is an intercept, and beta is a coefficient.
For meteorological elements having a response threshold, the intercept α and the coefficient β are calculated in segments by using the response threshold as a dividing point.
S4-2, based on the Poisson distribution theory, calculating the relative change quantity of daily electrical load peak value caused by unit change of different meteorological elements, namely the relative risk RR of the daily electrical load peak value i ：
RR i ＝EXP(β.ΔX i )
Wherein DeltaX i The amount of change in the weather element.
S4-3, calculating different meteorological elements X i Rate of change Δrr of peak daily maximum power load caused by unit change i ：
ΔRR i ＝(RR i -1)×100％。
Further, the specific calculation mode of the step S5 of the present invention is:
wherein DeltaL m To change the power load influenced by the meteorological factors, L max(t+1) For the next daily peak power load, L max(t) To peak current daily electrical load, X i(t+1) For the next day forecast value of a certain meteorological element, X i(t) For the current day live value of the meteorological element, m i The unit change is represented, the air temperature and the air speed are respectively 1, the temperature and humidity index is 100, and the precipitation is 10.
Further, the next-day electric load peak value L in step S7 of the present invention max(t+1) The calculation mode of (a) is as follows:
L max(t+1) ＝L max(t) +ΔL e +ΔL m +ε
wherein L is max(t+1) Delta L is the peak value of the current daily electrical load e Delta L for the change of basic electricity load affected by economic development m For the change of the electricity load influenced by the meteorological factors, epsilon is the random component of the electricity load of the main holiday.
The basic electric load change amount DeltaL affected by the economic development e The calculation mode of (a) is as follows:
ΔL e ＝L (t+1) -L t
wherein L is t For the daily basis of electricity load, L (t+1) The electric load is used for the next day basis.
Further, the calculation method of the change amplitude DeltaLR of the peak value of the next daily electricity load in the step S7 compared with the peak value of the current daily electricity load is as follows:
further, the early warning level in the step S7 of the present invention is set as follows:
when the delta LR is more than 10%, a first-level early warning is issued to remind a power dispatching department that the daily electricity load possibly has amplitude variation of more than 10% caused by meteorological factors;
when the delta LR is 15%, a secondary early warning is issued to remind the power dispatching department that the daily electricity load possibly has amplitude variation of more than 15% due to meteorological factors, and regulation measures need to be taken for coping.
The invention predicts the power consumption load amplitude caused by the meteorological conditions by using the relative risk, and the prediction effect and accuracy can meet the application requirements of the meteorological service business by inspection, thereby providing scientific and effective basis for the power consumption load prediction, especially the reasonable scheduling of short-term power consumption load under the condition of continuous high temperature and high humidity in summer.
Drawings
FIG. 1 is a graph of peak amplitude profile of electrical load.
Fig. 2 is a graph of daily average air temperature versus daily maximum load fit.
Fig. 3 is a graph of daily maximum air temperature versus daily maximum load fit.
FIG. 4 is a graph of daily minimum air temperature versus daily maximum load fit.
FIG. 5 is a graph of daily average wind speed versus daily maximum load fit.
FIG. 6 is a graph of daily maximum wind speed versus daily maximum load fit.
FIG. 7 is a plot of daily precipitation (at 20-20 hours) versus daily maximum load.
FIG. 8 is a graph of temperature and humidity index versus daily maximum load fit.
Fig. 9 is a graph of predicted versus actual peak daily electrical load values for 2022 using the present invention.
Detailed Description
The invention will be further described in detail with reference to the drawings and examples.
Examples
The invention takes certain city of North China as an example to predict the peak value of the power load of the secondary daily use, and the method comprises the following steps:
s1, acquiring daily electricity load peak value historical data of the city in 2013-2021 from the electric company of the city.
S2, weather observation historical data in corresponding time periods are obtained from the city weather station, wherein the weather observation historical data comprise daily maximum air temperature, daily minimum air temperature, daily average air temperature, 24h temperature change, daily minimum relative humidity, daily maximum wind speed, daily average wind speed and accumulated precipitation data in 20-20 hours, corresponding to each time period, of 24h, 48h and 72h, and a daily temperature and humidity index is calculated according to the daily average air temperature and the daily average relative humidity.
Temperature-humidity index E t The calculation mode of (a) is as follows:
wherein T is the daily average air temperature (DEG C), and R is the daily average relative humidity (%).
S3, calculating the correlation between the daily electrical load peak value of the tested area and the meteorological element according to the acquired daily electrical load peak value historical data and meteorological observation historical data in 2013-2021 and the calculated daily temperature and humidity index, and performing curve fitting to determine a meteorological element response threshold. The concrete mode is as follows:
s3-1, calculating the correlation between different meteorological elements and daily electrical load peaks by adopting a Spearman rank correlation method, namely, for observation data (x) between n pairs of meteorological elements and daily electrical load peaks i ,y i ) (i=1, 2, …, n), the rank is increased from small to large according to the size order of n data of each group of variables, and the repeated data is averaged, so that the rank correlation coefficient r between the meteorological element and the daily electricity load peak value s The method comprises the following steps:
wherein n is the logarithm of the observed data, R i For observing meteorological elements i Rank order, Q i Peak data y of daily electrical load i Rank order of (c);
s3-2, checking rank correlation coefficient r by T test method s The significance of (2) is calculated by the following steps:
wherein t is a calculated value of rank correlation coefficient, and n is a logarithm of observation data.
According to the correlation coefficient check table, determining the meteorological elements which can pass the saliency check of alpha=0.01, and eliminating the meteorological elements which do not pass the saliency check.
S3-3, fitting the peak value of the electric load of the next day with the meteorological element of the current day by using a nonlinear least square method:
L＝aX 3 +bX 2 +cX+d
wherein L is the peak value of the power load of the next day, X is a meteorological element, and a, b, c, d is a fitting coefficient.
And S4, calculating the rank correlation coefficient of the daily electricity load peak value and the meteorological element according to the specific mode (see table 1). And selecting the daily average air temperature, the daily highest air temperature, the daily lowest air temperature, the temperature-humidity index, the daily average air speed, the daily maximum air speed and the precipitation amount at 20-20 hours as influence factors, and calculating the change rate of the daily maximum electricity load peak change caused by unit change of different meteorological elements, namely the relative risk of the different meteorological elements to the daily electricity load peak. The specific operation mode is as follows:
s4-1, after eliminating meteorological elements which do not pass the alpha=0.01 saliency test, taking the peak value of the maximum power load of the next day as a dependent variable, taking the current day observed value of the related meteorological elements which pass the saliency test as an independent variable, and fitting the logarithmic relation between the maximum power load of the day and the meteorological factors as follows:
In[E(L)]＝βX+α
wherein E (L) is an expected value of a daily maximum power load, X is a meteorological element, alpha is an intercept, and beta is a coefficient.
For meteorological elements having a response threshold, the intercept α and the coefficient β are calculated in segments by using the response threshold as a dividing point.
S4-2, based on the Poisson distribution theory, calculating the relative change quantity of daily electrical load peak value caused by unit change of different meteorological elements, namely the relative risk RR of the daily electrical load peak value i ：
RR i ＝EXP(β.ΔX i )
Wherein DeltaX i The amount of change in the weather element.
S4-3, calculating different meteorological elements X i Rate of change Δrr of peak daily maximum power load caused by unit change i ：
ΔRR i ＝(RR i -1)×100％。
The calculation results of the rank correlation coefficients of the peak daily electrical load and the meteorological elements shown in table 1 were obtained.
TABLE 1 correlation of daily peak load values of 2013-2021 in certain city of North China and meteorological elements of the same day
Note that: * Represent a significance test by α=0.01.
The weather factors passing the significance test of α=0.01 in table 1 were subjected to curve fitting, and it was found that there was a significant inflection point with the fitted curve of the daily electrical load, namely, there was a threshold effect in the daily average air temperature, the daily maximum air temperature, the daily minimum air temperature, the daily average air speed, the daily maximum air speed, the temperature and humidity index, and the inflection point values were used as thresholds, respectively: the average daily air temperature threshold is 16 ℃, the maximum daily air temperature threshold is 22 ℃, the minimum daily air temperature threshold is 12 ℃, the average daily air speed threshold is 5m/s, the maximum daily air speed threshold is 10m/s, and the temperature and humidity indexes are respectively-1000 and 500. Other meteorological elements that pass the significance test have no threshold effect (fig. 2-8).
S5, calculating the cumulative variation of the peak value of the next-day power consumption load caused by all weather elements with significance according to the variation rate of the peak value variation of the daily power consumption load and real-time weather element data, namely the variation of the peak value of the next-day power consumption load caused by weather conditions. The concrete mode is as follows:
s5-1, selecting historical meteorological elements with obvious correlation, and calculating the corresponding daily electrical load peak value relative risk (see table 2).
TABLE 2 daily Electrical load peak relative Risk for Meteorological elements of certain city 2013-2021 in North China
Note that: * Meaning passing the significance test.
S5-2, calculating the change rate of the peak value of the daily maximum electricity load caused by unit change of different meteorological elements, wherein the specific calculation mode is as follows:
wherein DeltaL m To change the power load influenced by the meteorological factors, L max(t+1) For the next daily peak power load, L max(t) To peak current daily electrical load, X i(t+1) For the next day forecast value of a certain meteorological element, X i(t) For the current day live value of the meteorological element, m i The unit change is represented, the air temperature and the air speed are respectively 1, the temperature and humidity index is 100, and the precipitation is 10.
When the daily average air temperature, the daily maximum air temperature and the daily minimum air temperature are higher than the threshold value, the relative risk of the power load peak value is respectively increased by 2.25 percent, 1.92 percent and 2.07 percent when the temperature rises by 1 ℃; when the temperature is lower than the threshold value, the relative risk of the power load peak value is reduced by 0.62%,0.57% and 0.6% respectively when the temperature rises by 1 ℃. When the daily average wind speed reaches 5m/s, the relative risk of the power load is reduced by 4.31 percent when the daily average wind speed rises by 1 m/s; when the daily average wind speed is less than 5m/s, the relative risk of the electric load is increased by 2.04% every time the daily maximum wind speed is increased by 1m/s, and the relative risk of the electric load is reduced by 0.44%. When the temperature and humidity index is less than or equal to-1000, the relative risk of the electric load is increased by 3.66 percent along with the occurrence of 100 unit changes, when the temperature and humidity index is between-1000 and 500, the relative risk of the electric load is reduced by 0.31 percent, and when the temperature and humidity index exceeds 500, the relative risk of the electric load is increased by 4.22 percent. When the accumulated precipitation amount is increased by 10mm in 24 hours at 20-20 hours, the relative danger degree of the electric load is reduced by 3.47 percent.
S6, the random component of the electricity load basically comprises electricity load variation caused by uncertain factors such as holiday effect, industrial overhaul, power grid price adjustment and the like, and only the holiday effect is considered. According to the existing literature and analysis of the market data materials, the influence of common minor and long holidays and weekends on the maximum value of daily electric load is found to be small, the first-day electric load of holidays such as five-one, eleven, spring festival and the like can be obviously reduced, and the fourth-day electric load can be obviously increased in general. Therefore, the median of the daily electrical load peak variation amounts on two days, i.e., the day when the daily electrical load is significantly reduced and the day when the daily electrical load is significantly increased, in the history data is taken as the main holiday electrical load random component epsilon in that year. According to historical data, the holiday effects of 'five one', 'eleven' and the spring festival are determined to respectively account for 4.1%, 6.5% and 7.8% of the peak value of the electricity load of the previous day.
S7, calculating a peak value of the electricity load of the next day according to the random component of the electricity load of the main holiday, the peak value data of the electricity load of the current day, the observation data of the weather elements of the current day and the forecast data of the weather elements of the next day; and when the calculated peak value of the daily electricity load reaches a set early warning level compared with the variation amplitude of the peak value of the daily electricity load, early warning is carried out.
Peak value L of electric load for the next day in step S7 max(t+1) The calculation mode of (a) is as follows:
L max ( t+1 )＝L max ( t )+ΔL e +ΔL m +ε
wherein L is max(t+1) Delta L is the peak value of the current daily electrical load e Delta L for the change of basic electricity load affected by economic development m For the change of the electricity load influenced by the meteorological factors, epsilon is the random component of the electricity load of the main holiday.
The basic electric load change amount DeltaL affected by the economic development e The calculation mode of (a) is as follows:
ΔL e ＝L( t+1 )-L t
wherein L is t For the daily basis of electricity load, L (t+1) The electric load is used for the next day basis.
The calculation mode of the change amplitude delta LR of the peak value of the daily electricity load in the step S7 compared with the peak value of the current daily electricity load is as follows:
the early warning level in step S7 is set as follows:
1. when the delta LR is more than 10%, a first-level early warning is issued to remind a power dispatching department that the daily electricity load possibly has amplitude variation of more than 10% caused by meteorological factors;
2. when the delta LR is 15%, a secondary early warning is issued to remind the power dispatching department that the daily electricity load possibly has amplitude variation of more than 15% due to meteorological factors, and regulation measures need to be taken for coping.
And calculating the daily electrical load peak predictive value by using the maximum value of the daily electrical load of the city from 1 month to 12 months in 2022 and the corresponding meteorological element. As can be seen from the comparison curve of the predicted value and the actual load value in FIG. 9, the two curves are consistent in height, the predicted error distribution condition is shown in Table 3, the predicted error is between plus and minus 50MW for 101 days, the ratio of the predicted error is 27.7%, and the predicted error of 92.9% is between plus and minus 200MW, which indicates that the predicted has a certain accuracy rate and can be used as weather service business.
TABLE 3 prediction error

Claims (8)

1. The method for predicting the peak value of the next daily electrical load based on the meteorological conditions is characterized by comprising the following steps:
s1, acquiring daily electricity load peak historical data of a tested region after quality control for at least five years from a tested day from an electric power department;
s2, acquiring weather observation historical data in a period corresponding to a measured region from a weather department, wherein the weather observation historical data comprise daily maximum air temperature, daily minimum air temperature, daily average air temperature, 24h temperature change, daily average relative humidity, daily minimum relative humidity, daily maximum wind speed, daily average wind speed, and accumulated precipitation of each period corresponding to 24h, 48h and 72h at 20-20 hours, and calculating daily temperature and humidity indexes according to daily average air temperature and daily average relative humidity;
s3, calculating the correlation between the daily electrical load peak value and the meteorological element of the measured area according to the acquired daily electrical load peak value historical data, the meteorological observation historical data and the calculated daily temperature and humidity index, and performing curve fitting to determine a meteorological element response threshold;
s4, selecting weather elements with obvious correlation from the correlation between the daily electricity load peak value and the weather elements, and calculating the change rate of the daily maximum electricity load peak value change caused by unit change of different weather elements;
s5, calculating the accumulated change quantity of the peak value of the next-day power consumption load caused by all weather elements with significance according to the change rate of the peak value change of the daily power consumption load and the weather element data of the current day, namely the change quantity of the peak value of the next-day power consumption load caused by weather conditions;
s6, taking the median of the daily electricity load peak variation of the daily electricity load on the day of obvious decline and daily electricity load on the day of obvious rise in each holiday as the random component of the daily electricity load on the main holiday in the year;
s7, calculating a peak value of the electricity load of the next day according to the random component of the electricity load of the main holiday, the peak value data of the electricity load of the current day, the observation data of the weather elements of the current day and the forecast data of the weather elements of the next day; and when the calculated peak value of the daily electricity load reaches a set early warning level compared with the variation amplitude of the peak value of the daily electricity load, early warning is carried out.
2. The method for predicting peak electrical loads on next day based on meteorological conditions as recited in claim 1, wherein the temperature and humidity index E in step S2 t The calculation mode of (a) is as follows:
wherein T is the daily average air temperature (DEG C), and R is the daily average relative humidity (%).
3. The method for predicting peak power load on next day based on meteorological conditions according to claim 1, wherein the specific operation manner of step S3 is as follows:
s3-1, calculating different gases by adopting Spearman rank correlation methodCorrelation between meteorological elements and daily electrical load peaks, i.e. for observation data between n pairs of meteorological elements and daily electrical load peaks (x i ,y i ) (i=1, 2, …, n), the rank is increased from small to large according to the size order of n data of each group of variables, and the repeated data is averaged, so that the rank correlation coefficient r between the meteorological element and the daily electricity load peak value s The method comprises the following steps:
wherein n is the logarithm of the observed data, R i For observing meteorological elements i Rank order, Q i Peak data y of daily electrical load i Rank order of (c); s3-2, checking rank correlation coefficient r by T test method s The significance of (2) is calculated by the following steps:
wherein t is a calculated value of a rank correlation coefficient, and n is a logarithm of observed data;
according to the correlation coefficient check list, determining meteorological elements which can pass the saliency check of alpha=0.01, and eliminating meteorological elements which do not pass the saliency check;
s3-3, fitting the peak value of the electric load of the next day with the meteorological element of the current day by using a nonlinear least square method:
L＝aX 3 +bX 2 +cX+d
wherein L is the peak value of the power load of the next day, X is a meteorological element, and a, b, c, d is a fitting coefficient;
and drawing a fitting curve according to a fitting formula, wherein if an inflection point appears in the curve, the curve is used as a response threshold value of the meteorological element to a daily electrical load peak value.
4. The method for predicting peak daily electrical loads based on meteorological conditions of claim 3, wherein the specific operation manner of step S4 is as follows:
s4-1, after eliminating meteorological elements which do not pass the alpha=0.01 saliency test, taking the peak value of the maximum power load of the next day as a dependent variable, taking the current day observed value of the related meteorological elements which pass the saliency test as an independent variable, and fitting the logarithmic relation between the maximum power load of the day and the meteorological factors as follows:
In[E(L)]＝βX+α
wherein E (L) is an expected value of a daily maximum power load, X is a meteorological element, alpha is an intercept, and beta is a coefficient;
for meteorological elements with response thresholds, taking the response thresholds as dividing points, and respectively calculating in sections to obtain an intercept alpha and a coefficient beta;
s4-2, based on the Poisson distribution theory, calculating the relative change quantity of daily electrical load peak value caused by unit change of different meteorological elements, namely the relative risk RR of the daily electrical load peak value i ：
RR i ＝EXP(β.ΔX i )
Wherein DeltaX i The change amount of the meteorological element;
s4-3, calculating different meteorological elements X i Rate of change Δrr of peak daily maximum power load caused by unit change i ：
ΔRR i ＝(RR i -1)×100％。
5. The method for predicting peak power load on next day based on meteorological conditions according to claim 4, wherein the specific calculation mode in step S5 is as follows:
wherein DeltaL m To change the power load influenced by the meteorological factors, L max(t+1) For the next daily peak power load, L max(t) To peak current daily electrical load, X i(t+1) For the next day forecast value of a certain meteorological element, X i(t) For the weatherFactor current day live value, m i The unit change is represented, the air temperature and the air speed are respectively 1, the temperature and humidity index is 100, and the precipitation is 10.
6. The method for predicting peak daily electrical loads based on meteorological conditions of claim 5, wherein peak daily electrical loads L in step S7 max(t+1) The calculation mode of (a) is as follows:
Lmax (t+1) ＝L max (t)+ΔL e +ΔL m +ε
wherein L is max(t+1) Delta L is the peak value of the current daily electrical load e Delta L for the change of basic electricity load affected by economic development m The electricity load variation quantity influenced by the meteorological factors is epsilon which is the random component of the electricity load of the main holiday;
the basic electric load change amount DeltaL affected by the economic development e The calculation mode of (a) is as follows:
ΔL e ＝L( t+1 )-L t
wherein L is t For the daily basis of electricity load, L (t+1) The electric load is used for the next day basis.
7. The method for predicting peak value of next-day electric load based on meteorological conditions according to claim 6, wherein the method for calculating the variation range Δlr of peak value of next-day electric load in step S7 compared with peak value of current-day electric load is as follows:
8. the method for predicting peak power load on next day based on meteorological conditions according to claim 7, wherein the pre-warning level in step S7 is set as follows:
when the delta LR is more than 10%, a first-level early warning is issued to remind a power dispatching department that the daily electricity load possibly has amplitude variation of more than 10% caused by meteorological factors;
when the delta LR is 15%, a secondary early warning is issued to remind the power dispatching department that the daily electricity load possibly has amplitude variation of more than 15% due to meteorological factors, and regulation measures need to be taken for coping.
CN202310217831.3A 2023-03-08 2023-03-08 Meteorological condition-based next-day power load peak prediction method Active CN116432821B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310217831.3A CN116432821B (en) 2023-03-08 2023-03-08 Meteorological condition-based next-day power load peak prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310217831.3A CN116432821B (en) 2023-03-08 2023-03-08 Meteorological condition-based next-day power load peak prediction method

Publications (2)

Publication Number Publication Date
CN116432821A CN116432821A (en) 2023-07-14
CN116432821B true CN116432821B (en) 2023-09-26

Family

ID=87091600

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310217831.3A Active CN116432821B (en) 2023-03-08 2023-03-08 Meteorological condition-based next-day power load peak prediction method

Country Status (1)

Country Link
CN (1) CN116432821B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807508A (en) * 2019-10-31 2020-02-18 国网辽宁省电力有限公司经济技术研究院 Bus peak load prediction method considering complex meteorological influence
CN114266375A (en) * 2021-08-18 2022-04-01 国家电网有限公司 Regional short-term power load peak prediction method based on multi-innovation algorithm
CN114386204A (en) * 2022-01-14 2022-04-22 河北省气象服务中心(河北省气象影视中心) Power grid summer air conditioner load peak value refined prediction method and system based on gray model

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108009938B (en) * 2016-11-02 2021-12-03 中国电力科学研究院 System load clustering and load period pattern recognition method based on shape
CN111222711B (en) * 2020-01-16 2022-04-08 大连理工大学 Index linkage analysis-based multi-objective optimization method for peak shaving scheduling of electric power system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807508A (en) * 2019-10-31 2020-02-18 国网辽宁省电力有限公司经济技术研究院 Bus peak load prediction method considering complex meteorological influence
CN114266375A (en) * 2021-08-18 2022-04-01 国家电网有限公司 Regional short-term power load peak prediction method based on multi-innovation algorithm
CN114386204A (en) * 2022-01-14 2022-04-22 河北省气象服务中心(河北省气象影视中心) Power grid summer air conditioner load peak value refined prediction method and system based on gray model

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
气象因素对淮北市电力负荷的影响及其预测研究;汪付华;周后福;张屏;靳继斌;孙金贺;王苏瑶;;气象与环境学报(第04期);全文 *

Also Published As

Publication number Publication date
CN116432821A (en) 2023-07-14

Similar Documents

Publication Publication Date Title
CN104376510A (en) Method of predicting and accessing level of wildfire-caused trip risk in power transmission lines
CN110532505B (en) Method for calculating misalignment rate of ammeter
CN105678396A (en) Photovoltaic power station super-short-term power prediction device
CN102426674A (en) Power system load prediction method based on Markov chain
CN107423836B (en) Short-term load prediction method based on body sensing temperature
CN107403015B (en) Short-term optical power prediction method based on time series similarity
CN110889527A (en) Electric vehicle charging load prediction method based on LSTM neural network
CN102509173A (en) Markov chain based method for accurately forecasting power system loads
CN107977898B (en) Generated energy insurance pricing evaluation method of photovoltaic power station
CN112634077A (en) Medium-and-long-term power supply and demand situation analysis method
CN110909958A (en) Short-term load prediction method considering photovoltaic grid-connected power
CN114372360A (en) Method, terminal and storage medium for power load prediction
CN105337574A (en) Robust-regression-based distributed photovoltaic generating electricity-stealing identification method
Li et al. Relations of total electricity consumption to climate change in Nanjing
CN109787295A (en) It is a kind of meter and wind power plant state ultrashort-term wind power predict calculation method
CN109636033B (en) Spontaneous self-use comprehensive electricity price prediction method for distributed photovoltaic project
CN116432821B (en) Meteorological condition-based next-day power load peak prediction method
JP2012043857A (en) Bright and clear time determination device and photovoltaic power generation amount prediction system
Jose et al. Weather dependency of electricity demand: A case study in warm humid tropical climate
CN104657787B (en) A kind of wind power time series combination forecasting method
Klimenko et al. Climatic extremes: A new challenge for Russian power systems
Falkoni et al. Linear correlation and regression between the meteorological data and the electricity demand of the Dubrovnik region in a short-term scale
CN103455730A (en) Distributed photovoltaic power generating capacity estimating system and solar radiation data generation method
CN116365500A (en) Wind power plant power generation power prediction method based on special region set prediction
Song Response of Power Load to Meteorological Factors: an Empirical Study from Anhui, China

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