CN107730031A - A kind of ultra-short term peak load forecasting method and its system - Google Patents
A kind of ultra-short term peak load forecasting method and its system Download PDFInfo
- Publication number
- CN107730031A CN107730031A CN201710874560.3A CN201710874560A CN107730031A CN 107730031 A CN107730031 A CN 107730031A CN 201710874560 A CN201710874560 A CN 201710874560A CN 107730031 A CN107730031 A CN 107730031A
- Authority
- CN
- China
- Prior art keywords
- peak load
- load
- peak
- prediction
- ultra
- 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.)
- Granted
Links
- 238000013277 forecasting method Methods 0.000 title claims abstract 3
- 238000000034 method Methods 0.000 claims abstract description 57
- 238000004364 calculation method Methods 0.000 claims abstract description 8
- 239000013598 vector Substances 0.000 claims description 50
- 238000013528 artificial neural network Methods 0.000 claims description 36
- 238000012549 training Methods 0.000 claims description 26
- 230000000630 rising effect Effects 0.000 claims description 24
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000010219 correlation analysis Methods 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 2
- 238000012821 model calculation Methods 0.000 claims description 2
- 238000004393 prognosis Methods 0.000 abstract 2
- 238000004422 calculation algorithm Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 8
- 230000008859 change Effects 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 6
- 230000006872 improvement Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000002596 correlated effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000001816 cooling Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000013213 extrapolation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000017105 transposition 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of ultra-short term peak load forecasting method and its system, gathers weather prognosis information；The weather prognosis information of peak load forecasting model and the collection based on foundation carries out ultra-short term；The peak load forecasting model includes：The correlation of peak load index, the meteorologic factor for influenceing peak load and each meteorologic factor；Meteorologic factor based on the peak load and influence peak load.Example of calculation provided by the invention proves that this method predicts peak the precision of prediction for effectively, improving super short period load.
Description
Technical Field
The invention relates to a prediction method and a prediction system in the field of power system dispatching automation, in particular to an ultra-short-term peak load prediction method and an ultra-short-term peak load prediction system.
Background
With the increase of economy and the improvement of the living standard of people, the power load tends to increase year by year. The ultra-short term load prediction is to predict the power load within 5 minutes to 4 hours in the future in real time by using the latest load information. The load change of the power system can be tracked on line by ultra-short term load prediction, and the method is the basis of dynamic power grid safety monitoring and automatic power generation control. The method has the advantages of accurate and quick ultra-short term load prediction, and has an important supporting function for ensuring the safety and the economical efficiency of the operation of the power grid.
At present, the main methods for ultra-short-term load prediction include extrapolation, support vector machines, neural networks, data mining and the like. Because these prediction models think that the weather changes very little when the time is short, and the accuracy of weather forecast is not high, easily causes the error stack, therefore mostly do not consider the influence of meteorological factor to the load. Therefore, the prediction accuracy of the method is high in the load stable period, and the prediction error is large in the load period with particularly sensitive meteorological changes such as peak load and low-valley load.
The difficulty of ultra-short term load prediction remains peak load prediction. Through research on typical summer loads of a province, the province finds that the urban cooling load is suddenly reduced due to sudden change of weather in summer due to excessive storm and rainstorm, and the total load is reduced. According to a conventional prediction method, the practical requirement of ultra-short term prediction is obviously difficult to meet without considering weather mutation or only considering day characteristic meteorological elements such as the day highest temperature, the day lowest temperature, the weather type and the like. Therefore, accurate real-time weather forecast becomes a key factor for improving the load prediction accuracy.
Numerical weather forecast is a weather forecast which is more precise in time and space and is mainly used in wind power prediction in an electric power system. The system provides abundant meteorological information, and the spatial scale comprises various information such as air pressure, temperature, humidity, wind, cloud, precipitation and the like every dozens of square kilometers; on a time scale, the prediction results are refined to hours or less. The numerical weather forecast provides real-time accurate meteorological data support for power grid ultra-short term load prediction.
Since peak loads are very sensitive to meteorological changes, peak load prediction has been a difficult problem for power systems due to its randomness and complexity.
Disclosure of Invention
In order to solve the above-mentioned deficiencies in the prior art, the present invention aims to provide a method and a system for ultra-short term peak load prediction, which are effective for peak prediction and improve the prediction accuracy of ultra-short term load.
The purpose of the invention is realized by adopting the following technical scheme:
the invention provides a method for predicting peak load in an ultra-short period, which has the improvement that:
collecting weather prediction information;
ultra-short-term load prediction is carried out based on the built peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
Further, the peak load prediction model comprises:
determining a peak load index;
analyzing the peak load and the weather information based on historical weather information to determine the correlation of weather factors influencing the peak load;
and constructing a peak load prediction model by combining an L-M neural network based on the peak load and the meteorological factors influencing the peak load.
Further, the determining a peak load index comprises:
determining peak load based on pre-acquired data and analyzing the peak load to extract a peak load index;
the peak load indicators include:
1) peak top and valley bottom load values:
peak top load value PpeakIs the maximum load value during the day; load value P of valley bottomvalleyIs the minimum load value during the day;
2) the peak top time:
the peak-top time is the time when the peak load peak occurs;
3) peak-to-valley difference:
the expression for the peak-to-valley difference is as follows:
Pdifference＝Ppeak–Pvalley(1)
wherein, PdifferenceIs the peak-to-valley difference;
4) a high end load ramp rate comprising:
system load PhSatisfies the relationship:
Ph≥Ppeak–Pdifference/4 (2)
the load level is at a high level at this time;
the high-end load slope at time t is:
in the formula: s is the high load slope, d is the high load time interval, PtIndicating the high level load value, P, at time ttRepresenting the high-order load value at the time t-1; stIs the high load slope at time t;
the load slope of the high steep rise section must reach a steep rise level, measured as:
ssteep≥0.75smax(4)
in the formula: smaxThe slope of the steepest time period;
5) load high-position steep rise starting time:
the expression is as follows:
Tsteep＝min(Tsteep1,Tsteep2,…,Tsteepn) (5)
wherein s issteepIs at a steep level; t issteep1,Tsteep2,…,TsteepnRespectively as follows: no. 1, No. 2, No. n high steep rising paragraph start time; t issteepDefining the ratio of the load increment and the time increment of the high-position steep rising section satisfying the formula (4) as the high-position steep rising speed of the load for the high-position steep rising starting moment of the sample curve;
6) peak duration:
peak duration is the time span made up of loads reaching high load levels.
Further, said analyzing said peak load and weather information based on historical weather information to determine the relevance of weather factors affecting peak load comprises:
determining related meteorological factors according to the historical meteorological information;
analyzing the correlation of the peak load and meteorological factors;
and verifying the relevance of each meteorological factor influencing the peak load based on a robust regression model.
Further, the meteorological factors include: instantaneous wind speed, instantaneous temperature, instantaneous air pressure, maximum wind direction, rainfall, maximum wind speed, instantaneous wind direction, instantaneous humidity, minimum humidity, maximum temperature, minimum air pressure, maximum wind speed information.
Further, said analyzing the correlation of peak load to meteorological factors includes: calculating the correlation coefficient of each meteorological factor based on the following formula:
in the formula: r is a correlation coefficient; cov (X, Y) is the covariance of X and Y;mean square error of X and Y, respectively;
x is: historical load data; y is: historical temperature data.
Further, the robust regression model includes:
constructing a robust regression model calculation formula as follows:
Y＝Xβ+ε (7)
wherein: y ═ Yi)m×1Historical load data at a certain moment of m days; x ═ Xij)m×nHistorical temperature data at a time of m days, β ═ βj)n×1Is an estimated unknown parameter vector, e ═ ei)m×1Is a random error vector that is not observable;
for β in the robust regression model (β) based on the following formulaj)n×1And (3) performing parameter estimation:
wherein D (β) represents the objective function in the robust regression model, yiRepresents the historical load data at a certain time of the ith day, i is 1, …, m, xijThe temperature data of a certain time on the ith day is shown, and m represents the total number of days;
finding optimal estimation parametersThe key point of (1) is to determine a weighting function matrix; let m-order full-rank diagonal matrix W be (W)i)m×mFor weighting function matrix, obtaining optimal estimation parameters by solvingSolving the formula:
wherein,ri＝yi-x(i,j)βjis a residual term; xTIs transposed data of the historical temperature data X at a certain moment of m days.
Further, said building a peak load prediction model in conjunction with an L-M neural network based on said peak load and said meteorological factors affecting peak load comprises:
selecting a training sample of the L-M neural network based on the peak load and the meteorological factors influencing the peak load;
determining a data type vector of the synthetic data sample;
and constructing a peak load prediction model according to the training samples.
Further, said selecting training samples of the L-M neural network based on said peak load and said meteorological factors affecting the peak load comprises: selecting system load samples and corresponding comprehensive data samples of k working days or holidays before a prediction day;
the synthetic data type vector of the synthetic data sample is:
Dk＝(Dk1Dk2Dk3Dk4Dk5)T(16)
wherein: k denotes the number of days k 1,2, …, p, p being the number of selected days, Dk1Day maximum temperature on day k, Dk2Day minimum temperature on day k, Dk3Weather conditions on day k, Dk4Average humidity of day k, Dk5A day type for day k;
constructing a peak load prediction model according to the training samples, comprising:
performing iterative optimization calculation to obtain vectors belonging to the same class from the comprehensive data type vectors of the comprehensive data samples by a C-means fuzzy clustering method, and selecting the peak load sample curves of the days corresponding to the vectors in the class and respectively recording the curves as Yp1(t),Yp2(t),…,Ypm(t)；
Determining a peak load prediction model according to the peak load sample curve, wherein the expression is as follows:
Mp＝{Yp1(t),Yp2(t),…,Ypm(t)} (17)
wherein: y isp1(t),Yp2(t),…,Ypm(t) a first, second, and/or mth peak load sample curve, respectively; m is the number of peak load sample curves.
Further, ultra-short term load prediction is performed based on the peak load prediction model in combination with near weather prediction, and the method comprises the following steps:
selecting a vector input to a peak load prediction model;
calculating a peak load prediction error until an iteration condition is met;
and outputting a peak load prediction result.
Further, said selecting a vector input to the peak load prediction model comprises:
taking a time threshold as each interval period of peak load, then predicting that the object is a set point load for a day; for a single training, taking the predicted moment loads of m peak samples as load input vectors of a neural network;
according to the correlation analysis between the peak load and the meteorological factors, the simultaneous temperature t of m peak load samples is selectedkAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for a peak load prediction model.
Further, the time threshold is selected to be 5 minutes; the setpoint load takes 288 points.
Further, the iteration condition is represented by the following formula:
wherein: y (t)i) Predicting model output value, A (t), for peak loadi) The peak load prediction model is an ideal value, W is an optimal value which enables the error between the output value of the peak load prediction model and the ideal value to be minimum, and epsilon is a prediction precision threshold value; i denotes day i.
Further, the peak load prediction result comprises:
load forecast power estimation value
Calculating high-order steep rising initial time vectors T of various days for n peak load samples with the prediction types consistent with the prediction days and belonging to the same modesteep(1×n)＝(Tsteep1Tsteep2…Tsteepn)TCalculating the average starting time TAvsteep；
Peak time [ TAVsteep-T2H，TAVsteep+T2F]Wherein T is2HThe value is between 0.5 and 1h, T2FTaking values within 2-3 h;
in the formula: t issteep1,Tsteep2,…,TsteepnRespectively as follows: no. 1, No. 2, No. n high steep rising paragraph start time.
The invention also provides a system for forecasting the peak load in the ultra-short period, and the improvement lies in that: the system comprises:
the acquisition module is used for acquiring meteorological prediction information;
the prediction module is used for carrying out ultra-short-term load prediction based on the built peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
Further: the prediction module further comprises:
the building submodule is used for building a peak load prediction model;
the establishing submodule further includes:
a first determining unit for determining a peak load index;
the second determining unit is used for analyzing the peak load and the weather information to determine the correlation of each weather factor influencing the peak load based on historical weather information;
and the construction unit is used for constructing a peak load prediction model by combining an L-M neural network based on the peak load and the meteorological factors influencing the peak load.
Further, the first determining unit further includes:
and the analysis and extraction subunit is used for determining the peak load based on the pre-acquired data and analyzing the peak load to extract the peak load index.
Further, the second determining unit further includes:
the meteorological factor determining subunit is used for determining related meteorological factors according to the historical meteorological information;
the analysis subunit is used for analyzing the correlation between the peak load and the meteorological factor;
and the verification subunit is used for verifying the correlation of each meteorological factor influencing the peak load based on the robust regression model.
Further, the building unit further includes:
the first selection subunit is used for selecting a training sample of the L-M neural network based on the peak load and the meteorological factors influencing the peak load;
a third determining subunit, configured to determine a data type vector of the synthetic data sample;
and the construction subunit is used for constructing a peak load prediction model according to the training samples.
Further, the first selecting subunit is further configured to: selecting system load samples and corresponding comprehensive data samples of k working days or holidays before a prediction day;
further, the prediction module further includes:
a second selection submodule for selecting a vector input to the peak load prediction model;
the calculation submodule is used for calculating a peak load prediction error until an iteration condition is met;
and the output submodule is used for outputting a peak load prediction result.
Further, the second selecting submodule is further configured to:
taking a time threshold as each interval period of peak load, then predicting that the object is a set point load for a day; for a single training, taking the predicted moment loads of m peak samples as load input vectors of a neural network;
according to the correlation analysis between the peak load and the meteorological factors, the simultaneous temperature t of m peak load samples is selectedkAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for a peak load prediction model.
Further, the time threshold is selected to be 5 minutes; the setpoint load takes 288 points.
Compared with the closest prior art, the technical scheme provided by the invention has the beneficial effects that:
(1) according to the numerical weather forecast information, the weather-sensitive peak load change rule is analyzed, and the peak-valley load and weather factor correlation analysis is carried out to obtain the positive correlation between the peak load and the temperature. For the load data of the power grid system, the load high-order steep rising rate and the load high-order steep rising starting moment are used as peak operation mode characteristics, the contradiction that the response capability of the system is not suitable for the load high-order steep rising rate is reflected and processed, and the two indexes play an important role in improving the peak load prediction accuracy.
(2) On the basis of the peak load mode, an L-M neural network method is adopted to establish a peak load prediction method based on meteorological factors, and the design of a prediction model solving algorithm is completed. The method improves the structure and the learning algorithm of the original neural network aiming at the defects of insufficient data processing capacity and difficult convergence of the original network, replaces the original gradient descent method with the L-M algorithm, improves the dynamic performance of the network, reduces the learning time, has high convergence speed and improves the prediction precision.
Drawings
FIG. 1 is a typical daily load graph in summer of a certain province in 2014 provided by the present invention;
FIG. 2 is a peak load graph of 30 days at 6 months 2014 provided by the present invention;
FIG. 3 is a diagram of peak summer load versus peak time temperature provided by the present invention;
FIG. 4 is a detailed flow chart of the ultra short term peak load prediction provided by the present invention;
FIG. 5 is a diagram of the C-mean fuzzy clustering classification results provided by the present invention;
FIG. 6 is a graph comparing peak load prediction and actual curves for 7/month and 2/day provided by the present invention;
FIG. 7 is a comparison graph of prediction errors of the L-M and BP algorithms for 2 days after 7 months, provided by the present invention;
fig. 8 is a block diagram of the ultra-short term peak load prediction process provided by the present invention.
Detailed Description
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
The following description and the drawings sufficiently illustrate specific embodiments of the invention to enable those skilled in the art to practice them. Other embodiments may incorporate structural, logical, electrical, process, and other changes. The examples merely typify possible variations. Individual components and functions are optional unless explicitly required, and the sequence of operations may vary. Portions and features of some embodiments may be included in or substituted for those of others. The scope of embodiments of the invention encompasses the full ambit of the claims, as well as all available equivalents of the claims. Embodiments of the invention may be referred to herein, individually or collectively, by the term "invention" merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept if more than one is in fact disclosed.
The first embodiment,
The invention provides a method for predicting peak load in an ultra-short period, a flow chart of which is shown in figure 8, and the method comprises the following steps:
collecting weather prediction information;
ultra-short-term load prediction is carried out based on the built peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
Wherein: the peak load prediction model is established and comprises the following steps:
step 1: determining a peak load index;
1.1 Peak load feature analysis
Along with the increase of economy and the improvement of the living standard of people, the power load is in the trend of increasing year by year, and the summer load value is greatly higher than other seasons due to the existence of the cooling load in summer. The summer peak load characteristic sensitive to weather is mainly researched, and the summer load data of 6 months to 8 months in 2014 of a certain power saving network is taken as an example for analysis, so that holiday and double-holiday data are eliminated. Fig. 1 is a load curve of a typical summer day of a power saving network, and it can be seen from the figure that the power saving network is characterized by three peaks and three valleys, namely an early peak (11:15), a late peak (17:00), a late peak (21:15), a night valley (4:30), a rest valley (12:45) in the middle of the day and a low valley (19:00) in the evening and in the next shift.
1.2 Peak load index
Peak load graph of 30 days 6 months 2014 as shown in fig. 2, the object of the study herein is peak load, and the valley load mode can be treated as a peak (negative valley load) mode. And analyzing the load 24 hours a day, and extracting 6 characteristics of a peak top load value, a peak top time, a peak valley difference, a load high position steep rising speed, a load high position steep rising initial time, a peak duration and the like to form a peak load characteristic index vector. The specific meanings of these 6 features are defined as follows:
(1) peak top and valley bottom load values:
the peak top load value being the maximum load value P during the daypeak(ii) a The load value at the bottom of the valley is the minimum load value P in one dayvalley。
(2) The peak top time:
the peak-top time is the time when a peak load peak occurs.
(3) Peak-to-valley difference:
the peak-to-valley difference is the difference between the maximum load value and the minimum load value during a day.
Pdifference＝Ppeak–Pvalley(1)
(4) High-order load steep rising rate:
the "high" criterion needs to be specified here. In general, the intermediate value between the peak load value and the waist load value of the daily load curve is used as a boundary. Namely, the system load Ph satisfies the relationship:
Ph≥Ppeak–Pdifference/4 (2)
the load level is now "high". In a special case, the method is not limited by the above formula.
Defining s as the load slope and d as the load time interval, the load slope at time t is:
the load slope of the high steep rise section must reach a steep rise level, measured as:
ssteep≥0.75smax(4)
in the formula, smaxThe slope of the steepest period. The ratio of the load increment to the time increment of the high steep rising section satisfying the equation (4) is defined as the high steep rising rate of the load.
(5) Load high-position steep rise starting time:
the starting time of the 1 st high steep rise section determined according to equation (2) is referred to as the high steep rise starting time of the peak load curve population. Let a load sample curve determine the 1 st, 2 nd and so forth from equations (2) and (4), and the start time of the nth high steep rising segment is Tsteep1,Tsteep2,…,TsteepnThen, the high-order steep rise starting time of this sample curve is:
Tsteep＝min(Tsteep1,Tsteep2,…,Tsteepn) (5)
(6) peak duration:
peak duration is the time span made up of loads reaching high load levels; wherein: stIs the load slope at time t; ssteepIs at a steep level; pdifferenceIs the peak-to-valley difference; smaxThe slope of the steepest time period; t issteepThe sample curve is at the high steep rise starting time.
Step 2: analyzing and verifying the correlation of peak load and meteorological factors;
in recent years, the maximum load of power supply of a power grid is increased year by year, and the peak load is very sensitive to meteorological changes, particularly to changes of meteorological factors such as temperature and humidity. The change of weather causes the electricity load caused by the change of the comfortable feeling of human bodies, and the increase of the peak load is very sensitive to the change of the climate, which is called as the climate sensitive peak load. In the section, correlation analysis of peak load and meteorological factors is carried out, key factors influencing meteorological sensitive peak load are found out, and a regression model is established through a robust regression method for verification.
2.1 correlation analysis
The correlation coefficient is an important index describing the degree of linear correlation between vectors. The correlation coefficient calculation formula is as follows:
in the formula: r is a correlation coefficient; cov (X, Y) is the covariance of X and Y;mean square error of X and Y, respectively.
Instantaneous wind speed, instantaneous temperature, instantaneous air pressure, maximum wind direction, rainfall, maximum wind speed, instantaneous wind direction, instantaneous humidity, minimum humidity, highest temperature, lowest temperature, minimum air pressure, maximum wind speed and other information provided by numerical meteorological information and three-peak three-valley load of the province in one month are utilized to calculate three-peak three-valley load and correlation coefficients of the meteorological factors, and the meteorological factors with high load correlation are listed in an attached table 1.
As can be seen from the attached Table 1, the peak load in summer of this province is positively correlated with the temperature and negatively correlated with the rainfall. Wherein:
(1) the trimodal load is sensitive to temperature, and the correlation coefficient is above 0.80. The early peak (11:15) has the largest correlation coefficient of load with time temperature (0.855), the late peak (21:15) has the larger correlation coefficient with the highest temperature (0.8069), and the mid-day peak (17:00) has the larger correlation coefficient with the lowest temperature (0.8623).
(2) Secondly, the correlation coefficient of load and temperature of the noon valley (12:45) is larger (0.8236);
(3) peak-to-valley loading is sensitive to rainfall and is negatively correlated, i.e. as rainfall increases, the load decreases. The correlation coefficients of the afternoon peak (17:00) and the afternoon valley (19:00) to the rainfall amount reach-0.537 and-0.7106, respectively.
2.2 robust regression method verification
Considering a robust regression model:
Y＝Xβ+ε (7)
wherein Y is (Y)i)m×1Historical load data at a certain moment of m days; x ═ Xij)m×nHistorical temperature data at a time of m days, β ═ βj)n×1Is an estimated unknown parameter vector, e ═ ei)m×1Is a random error vector that is not observable.
For β ═ in the robust regression model (β)j)n×1The parameter estimation is performed to minimize the following objective function.
Finding optimal estimation parametersThe key to (1) is to determine a weighting function matrix. Let m-order full-rank diagonal matrix W be (W)i)m×mFor weighting function matrix, it can be obtained by solvingSolving the formula:
wherein,ri＝yi-x(i,j)βjare residual terms.
Establishing a peak load regression model by using a robust regression method, and obtaining a regression equation of the peak load at the 11:15 moment and the temperature at the moment as follows:
Y＝132.72X+595.238 (10)
the scatter plot shown in fig. 3 shows that the peak load and temperature relationship fit well with a more pronounced linear correlation, with peak load increasing with increasing temperature.
And step 3: constructing a peak load prediction model of the L-M neural network and carrying out load prediction:
3.1LM neural network
The L-M (Levenberg-Marquardt) neural network method combines the advantages of the gradient descent method and the Gauss-Newton method, and has the advantages that the convergence is very quick when the number of network weights is small, so that the learning time is obviously shortened, and the training frequency and the accuracy are obviously superior to those of a common BP algorithm.
In the process of training the network, the error between the network output and the ideal is solved as follows:
E(ti)＝Y(ti)-A(ti),i＝1,…,m (11)
by adjusting the values in the weight matrix W, the error of the output of the network from the ideal value is minimized, i.e. the following equation is satisfied:
let W denote the vector formed by the weight and the threshold, and the dimension of the vector is l. The L-M algorithm is a modified Gauss-Newton method, and is of the form:
ΔW＝(JTJ+μI)-1JTE (13)
wherein:
a prediction accuracy threshold epsilon is set and the iteration is ended if equation (15) is satisfied. Otherwise, modify the weights of the neurons of each layer until satisfied.
Wherein: e (t)i) For the error of the network output from the ideal, Y (t)i) For network output, i.e. power, A (t)i) Is an ideal value, WoptFor an optimal W value that minimizes the error between the output of the network and the ideal value, Δ W is a vector consisting of a weight obtained by the improved Gauss-Newton method and a threshold, W is an optimal value that minimizes the error between the output of the peak load prediction model and the ideal value, J is a Jacobian matrix, andTfor the transposition of the Jacobian matrix, E is E (W)opt) Representing the minimum value of the error of the output of the network from the ideal value, and m is the number of high peak load samples in the pattern.
3.2 peak load prediction model of L-M neural network:
3.2.1 sample selection:
the system load samples of 8-14 working days (or holidays) before a forecast day are selected, and in order to ensure that enough samples can be selected, the highest daily temperature, the lowest daily temperature, the weather condition, the rainfall, the daily type and the like are selected as comprehensive data samples to establish a load mode.
Setting the cluster sample weather comprehensive data type vector as:
Dk＝(Dk1Dk2Dk3Dk4Dk5)T(16)
where k is 1,2, …, p. p is selected days, Dk1Day maximum temperature on day k, Dk2Day minimum temperature on day k, Dk3Weather conditions on day k, Dk4Average humidity of day k, Dk5Day type of day k.
Calculating the vectors belonging to the same class by iterative optimization calculation through a C-means fuzzy clustering method, selecting the peak load sample curves of the days corresponding to the vectors in the class, and respectively recording the peak load sample curves as Yp1(t),Yp2(t),…,Ypm(t) and thereby forming a peak load mode, namely:
Mp＝{Yp1(t),Yp2(t),…,Ypm(t)} (17)
where m is the number of peak load samples in the mode.
3.2.2 selection of input vectors
5 minutes were taken as each interval period of peak load, so the predicted object was 288 load of the day. For a single training, the predicted moment loads of m peak samples are used as load input vectors of the neural network.
Selecting m peak loads according to the correlation analysis between peak loads and meteorological factors in section 2.1Simultaneous instant temperature t of the samplekAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for neural networks. Where k is 1,2, …, m.
3.3 ultra-short term peak load prediction procedure
The flow of the ultra-short term peak load prediction algorithm is shown in fig. 4, and the steps are as follows:
step 1: selecting a prediction date;
step 2: selecting a composite data type D according to meteorological factorskAnd performing a maximum minimum normalized data type to obtain D'k；
And step 3: adopting C-means fuzzy clustering to select M peak load curves M close to the predicted dayp＝{Yp1(t),Yp2(t),…,Ypm(t) } as samples belonging to the same peak load mode;
and 4, step 4: and inputting the neural network vector according to the selection method of the input vector.
And 5: establishing a multilayer feedforward L-M network prediction model, and calculating a prediction error until the prediction error meets the formula (15);
step 6: output load predicted power estimation
And 7: calculating high-order steep rising initial time vectors T of various days for n peak load samples with the prediction types consistent with the prediction days and belonging to the same modesteep(1×n)＝(Tsteep1Tsteep2…Tsteep n)TThen calculate their average starting time TAvsteep；
And 8: determining peak hours [ TAVsteep-T2H，TAVsteep+T2F]Wherein T is2HCan take a value of between 0.5 and 1 hour, T2FThe value can be taken within 2-3 h.
Example II,
Taking meteorological factor comprehensive data and peak load sample data in summer-to-summer period of 2014 of a certain province network as an example, predicting peak loads of a power network at 7, month and 2 days of 2014 by using an L-M neural network peak prediction algorithm, firstly, selecting 11 working day samples closer to 7, month and 2 days for clustering, wherein the comprehensive data type samples are shown in a table 2, the type 1 in the table 2 represents a working day, and weather condition distinguishing coefficients are shown in a table 3, obtaining results by using a C-mean fuzzy clustering method and are shown in a table 5, selecting load sample curves of days corresponding to ⊕ symbols in the same type from result graphs divided into two types, and forming a load mode Mp{Xp(t),Xp(t),…,Xp(t), wherein the components represent peak load sample curves of 17 days 6 months, 21 days 6 months, 26 days 6 months, 27 days 6 months, 28 days 6 months, and 1 day 7 months, respectively, in this order.
And predicting the ultra-short term peak load of 7, month and 2 days in 2014 according to the ultra-short term peak load prediction process to obtain an L-M prediction result. The curves of predicted and actual values of L-M are shown in FIG. 6.
The L-M prediction results are compared with the prediction results of the BP neural network, and the average relative error is calculated, the error comparison is shown in fig. 7. As can be seen from the figure, the L-M algorithm predicts peak duration intervals [19:30, 22:25 ]; the average relative error is 0.61785%; the maximum relative error is 1.7083%; predicting a peak load value to be 7667MW, actually predicting a peak load value to be 7694.7MW, and obtaining a peak value relative error to be 0.3599%; the predicted high steep rise start time is 19: 35. The number of learning training is 14. And the average relative error of the common BP method is 1.64205%, the maximum relative error is 3.4409%, the peak relative error is 1.765%, and the number of times of network learning training is 1000.
Meter 12014-year summer three-peak three-valley load and meteorological factor related coefficient meter
TABLE 2 comprehensive data type sample
Table 3 weather distinguishing coefficient table
The ultra-short-term peak load prediction method based on the L-M neural network and the numerical weather forecast establishes a peak load mode by using characteristic vectors such as peak starting time, high steep rising rate and the like, performs correlation analysis on peak load and meteorological factors through abundant meteorological information collected by the numerical weather forecast, verifies by using a robust regression method, and finds out key meteorological factors influencing the peak load. Then, ultra-short-term load prediction modeling is carried out through an L-M neural network method. And selecting a peak load sample based on a C-means fuzzy clustering method, and selecting key meteorological factors and load vectors of the peak load sample as the input of a neural network to carry out training prediction. The prediction method and the technical scheme are the core contents of the technology.
A summer peak load versus peak time temperature graph as shown in fig. 3, an ultra-short term peak load prediction flow chart as shown in fig. 4, a C-mean fuzzy clustering classification result graph as shown in fig. 5, a peak load prediction versus actual curve comparison graph as shown in fig. 6, and a prediction error comparison graph of L-M and BP algorithms as shown in fig. 7; and a summer three peak three valley load and weather factor correlation coefficient table as in table 1, a combined data type sample table attached to table 2, and a weather condition distinguishing coefficient table attached to table 3.
EXAMPLE III
The invention also provides a system for forecasting the peak load in the ultra-short period, and the improvement lies in that: the system comprises:
the acquisition module is used for acquiring meteorological prediction information;
the prediction module is used for carrying out ultra-short-term load prediction based on the built peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
The prediction module further comprises:
the building submodule is used for building a peak load prediction model;
the establishing submodule further includes:
a first determining unit for determining a peak load index;
the second determining unit is used for analyzing the peak load and the weather information to determine the correlation of each weather factor influencing the peak load based on historical weather information;
and the construction unit is used for constructing a peak load prediction model by combining an L-M neural network based on the peak load and the meteorological factors influencing the peak load.
The first determination unit further includes:
the peak load detection unit comprises a first determining subunit, a second determining subunit and a third determining subunit, wherein the first determining subunit is used for determining peak load based on pre-acquired data and analyzing the peak load to extract a peak load index.
The second determination unit further includes:
the second determining subunit is used for determining related meteorological factors according to the historical meteorological information;
the analysis subunit is used for analyzing the correlation between the peak load and the meteorological factor;
and the verification subunit is used for verifying the correlation of each meteorological factor influencing the peak load based on the robust regression model.
The building unit further comprises:
the first selection subunit is used for selecting a training sample of the L-M neural network based on the peak load and the meteorological factors influencing the peak load;
a third determining subunit, configured to determine a data type vector of the synthetic data sample;
and the construction subunit is used for constructing a peak load prediction model according to the training samples.
The first selecting subunit is further configured to: selecting system load samples and corresponding comprehensive data samples of k working days or holidays before a prediction day;
the prediction module further comprises:
a second selection unit for selecting a vector input to the peak load prediction model;
a calculation unit for calculating a peak load prediction error until an iteration condition is satisfied;
and the output unit is used for outputting the peak load prediction result.
Further, the second selecting unit is further configured to:
taking a time threshold as each interval period of peak load, then predicting that the object is a set point load for a day; for a single training, taking the predicted moment loads of m peak samples as load input vectors of a neural network;
according to the correlation analysis between the peak load and the meteorological factors, the simultaneous temperature t of m peak load samples is selectedkAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for a peak load prediction model. Selecting the time threshold value for 5 minutes; the setpoint load takes 288 points.
The invention provides a method and a system for predicting peak load in an ultra-short period, wherein a peak load mode is established by using characteristic vectors such as peak starting time, high steep rising rate and the like, correlation analysis is carried out on peak load and meteorological factors through abundant meteorological information collected by numerical weather forecast, and a robust regression method is used for verification to find out key meteorological factors influencing the peak load. Then, ultra-short-term load prediction modeling is carried out through an L-M neural network method. And selecting a peak load sample based on a C-means fuzzy clustering method, and selecting key meteorological factors and load vectors of the peak load sample as the input of a neural network to carry out training prediction. Practical examples prove that the technology is effective for peak prediction, and the prediction precision of ultra-short-term loads is improved.
Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can make modifications and equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention, which is set forth in the claims of the present application.
Claims (23)
1. A method for ultra-short term peak load prediction, characterized by:
collecting weather prediction information;
ultra-short-term load prediction is carried out based on a pre-established peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
2. The ultra-short term peak load prediction method of claim 1, wherein the pre-established peak load prediction model comprises:
determining a peak load index;
analyzing the peak load and the weather information based on historical weather information to determine the correlation of weather factors influencing the peak load;
and constructing a peak load prediction model by combining an L-M neural network based on the peak load and the meteorological factors influencing the peak load.
3. The ultra-short term peak load prediction method of claim 2, wherein said determining a peak load indicator comprises:
determining peak load based on pre-acquired data and analyzing the peak load to extract a peak load index;
the peak load indicators include:
1) peak top and valley bottom load values:
peak top load value PpeakIs the maximum load value during the day; load value P of valley bottomvalleyIs the minimum load value during the day;
2) the peak top time:
the peak-top time is the time when the peak load peak occurs;
3) peak-to-valley difference:
the expression for the peak-to-valley difference is as follows:
Pdifference＝Ppeak–Pvalley(1)
wherein, PdifferenceIs the peak-to-valley difference;
4) a high end load ramp rate comprising:
system load PhSatisfies the relationship:
Ph≥Ppeak–Pdifference/4 (2)
the load level is at a high level at this time;
the high-end load slope at time t is:
in the formula: s is the high load slope, d is the high load time interval, PtIndicating the high level load value, P, at time ttRepresenting the high-order load value at the time t-1; stIs the high load slope at time t;
the load slope of the high steep rise section must reach a steep rise level, measured as:
ssteep≥0.75smax(4)
in the formula: smaxThe slope of the steepest time period;
5) load high-position steep rise starting time:
the expression is as follows:
Tsteep＝min(Tsteep1,Tsteep2,…,Tsteepn) (5)
wherein s issteepIs at a steep level; t issteep1,Tsteep2,…,TsteepnRespectively as follows: no. 1, No. 2, No. n high steep rising paragraph start time; t issteepDefining the ratio of the load increment and the time increment of the high-position steep rising section satisfying the formula (4) as the high-position steep rising speed of the load for the high-position steep rising starting moment of the sample curve;
6) peak duration:
peak duration is the time span made up of loads reaching high load levels.
4. The ultra-short term peak load forecasting method of claim 2, wherein said analyzing the peak load and weather information based on historical weather information to determine the relevance of weather factors affecting peak load comprises:
determining related meteorological factors according to the historical meteorological information;
analyzing the correlation of the peak load and meteorological factors;
and verifying the relevance of each meteorological factor influencing the peak load based on a robust regression model.
5. The ultra-short term peak load prediction method of claim 4, wherein the meteorological factors comprise: instantaneous wind speed, instantaneous temperature, instantaneous air pressure, maximum wind direction, rainfall, maximum wind speed, instantaneous wind direction, instantaneous humidity, minimum humidity, maximum temperature, minimum air pressure, maximum wind speed information.
6. The ultra-short term peak load prediction method of claim 5, wherein said analyzing the correlation of peak load to meteorological factors comprises: calculating the correlation coefficient of each meteorological factor based on the following formula:
in the formula: r is a correlation coefficient; cov (X, Y) is the covariance of X and Y;mean square error of X and Y, respectively; x is historical load data; y is: historical temperature data.
7. The ultra-short term peak load prediction method of claim 4, wherein the robust regression model comprises: constructing a robust regression model calculation formula as follows:
Y＝Xβ+ε (7)
wherein: y ═ Yi)m×1Historical load data at a certain moment of m days; x ═ Xij)m×nHistorical temperature data at a time of m days, β ═ βj)n×1Is an estimated unknown parameter vector, e ═ ei)m×1Is a random error vector that is not observable;
for β in the robust regression model (β) based on the following formulaj)n×1And (3) performing parameter estimation:
wherein D (β) represents the objective function in the robust regression model, yiRepresents the historical load data at a certain time of the ith day, i is 1, …, m, xijThe temperature data of a certain time on the ith day is shown, and m represents the total number of days;
finding optimal estimation parametersThe key point of (1) is to determine a weighting function matrix; let m-order full-rank diagonal matrix W be (W)i)m×mFor weighting function matrix, obtaining optimal estimation parameters by solvingSolving the formula:
wherein,ri＝yi-x(i,j)βjis a residual term; xTIs transposed data of the historical temperature data X at a certain moment of m days.
8. The ultra short term peak load prediction method of claim 2, wherein said building a peak load prediction model in conjunction with an L-M neural network based on said peak load and said meteorological factors affecting peak load comprises:
selecting a training sample of the L-M neural network based on the peak load and the meteorological factors influencing the peak load;
determining a data type vector of the synthetic data sample;
and constructing a peak load prediction model according to the training samples.
9. The method of ultra-short term peak load prediction according to claim 8, wherein said selecting training samples of L-M neural networks based on said peak load and said meteorological factors affecting peak load comprises: selecting system load samples and corresponding comprehensive data samples of k working days or holidays before a prediction day;
the synthetic data type vector of the synthetic data sample is:
Dk＝(Dk1Dk2Dk3Dk4Dk5)T(16)
wherein: k denotes the number of days k 1,2, …, p, p being the number of selected days, Dk1Day maximum temperature on day k, Dk2Day minimum temperature on day k, Dk3Weather conditions on day k, Dk4Average humidity of day k, Dk5A day type for day k;
constructing a peak load prediction model according to the training samples, comprising:
performing iterative optimization calculation to obtain vectors belonging to the same class from the comprehensive data type vectors of the comprehensive data samples by a C-means fuzzy clustering method, and selecting the peak load sample curves of the days corresponding to the vectors in the class and respectively recording the curves as Yp1(t), Yp2(t),…,Ypm(t)；
Determining a peak load prediction model according to the peak load sample curve, wherein the expression is as follows:
Mp＝{Yp1(t),Yp2(t),…,Ypm(t)} (17)
wherein: y isp1(t),Yp2(t),…,Ypm(t) a first, second, and/or mth peak load sample curve, respectively; m is the number of peak load sample curves.
10. The ultra-short term peak load prediction method of claim 1, wherein ultra-short term load prediction based on the peak load prediction model in combination with near weather prediction comprises:
selecting a vector input to a peak load prediction model;
calculating a peak load prediction error until an iteration condition is met;
and outputting a peak load prediction result.
11. The ultra short term peak load prediction method of claim 10, wherein said selecting a vector input to a peak load prediction model comprises:
taking a time threshold as each interval period of peak load, then predicting that the object is a set point load for a day; for a single training, taking the predicted moment loads of m peak samples as load input vectors of a neural network;
according to the correlation analysis between the peak load and the meteorological factors, the simultaneous temperature t of m peak load samples is selectedkAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for a peak load prediction model.
12. The ultra short term peak load prediction method of claim 11, wherein said time threshold is selected to be 5 minutes; the setpoint load takes 288 points.
13. The ultra short term peak load prediction method of claim 10, wherein the iteration condition is represented by the following equation:
wherein: y (t)i) Predicting the model output value, A (t), for peak loadsi) The peak load prediction model is an ideal value, W is an optimal value which enables the error between the output value of the peak load prediction model and the ideal value to be minimum, and epsilon is a prediction precision threshold value; i denotes day i.
14. The ultra-short term peak load prediction method of claim 10, wherein the peak load prediction result comprises:
load forecast power estimation value
Calculating high-order steep rising initial time vectors T of various days for n peak load samples with the prediction types consistent with the prediction days and belonging to the same modesteep(1×n)＝(Tsteep1Tsteep2…Tsteepn)TCalculating the average starting time TAvsteep；
Peak time [ TAVsteep-T2H，TAVsteep+T2F]Wherein T is2HThe value is between 0.5 and 1h, T2FTaking values within 2-3 h;
in the formula: t issteep1,Tsteep2,…,TsteepnRespectively as follows: no. 1, No. 2, No. n high steep rising paragraph start time.
15. An ultra-short term peak load prediction system, comprising: the system comprises:
the acquisition module is used for acquiring meteorological prediction information;
the prediction module is used for carrying out ultra-short-term load prediction based on the built peak load prediction model and the collected meteorological prediction information;
the peak load prediction model comprises: peak load index, meteorological factors affecting peak load and correlation of each meteorological factor; based on the peak load and meteorological factors affecting the peak load.
16. The ultra short term peak load prediction system of claim 15, wherein: the prediction module further comprises:
the building submodule is used for building a peak load prediction model;
the establishing submodule further includes:
a first determining unit for determining a peak load index;
the second determining unit is used for analyzing the peak load and the weather information to determine the correlation of each weather factor influencing the peak load based on historical weather information;
and the construction unit is used for constructing a peak load prediction model by combining an L-M neural network based on the peak load and the meteorological factors influencing the peak load.
17. The ultra-short term peak load prediction system of claim 16, wherein the first determination unit further comprises:
and the analysis and extraction subunit is used for determining the peak load based on the pre-acquired data and analyzing the peak load to extract the peak load index.
18. The ultra-short term peak load prediction system of claim 16, wherein the second determination unit further comprises:
the meteorological factor determining subunit is used for determining related meteorological factors according to the historical meteorological information;
the analysis subunit is used for analyzing the correlation between the peak load and the meteorological factor;
and the verification subunit is used for verifying the correlation of each meteorological factor influencing the peak load based on the robust regression model.
19. The ultra-short term peak load prediction system of claim 16, wherein the building unit further comprises:
the first selection subunit is used for selecting a training sample of the L-M neural network based on the peak load and the meteorological factors influencing the peak load;
a third determining subunit, configured to determine a data type vector of the synthetic data sample;
and the construction subunit is used for constructing a peak load prediction model according to the training samples.
20. The ultra-short term peak load prediction system of claim 19, wherein said first selected subunit is further configured to: and selecting system load samples and corresponding comprehensive data samples of k working days or holidays before the prediction day.
21. The ultra-short term peak load prediction system of claim 15, wherein the prediction module further comprises:
a second selection submodule for selecting a vector input to the peak load prediction model;
the calculation submodule is used for calculating a peak load prediction error until an iteration condition is met;
and the output submodule is used for outputting a peak load prediction result.
22. The ultra-short term peak load prediction system of claim 21, wherein said second selection submodule is further configured to:
taking a time threshold as each interval period of peak load, then predicting that the object is a set point load for a day; for a single training, taking the predicted moment loads of m peak samples as load input vectors of a neural network;
according to the correlation analysis between the peak load and the meteorological factors, the simultaneous temperature t of m peak load samples is selectedkAnd simultaneously the rainfall rkMaximum daily temperature t of the samplemaxDay minimum temperature t of the sampleminAs meteorological factor input for a peak load prediction model.
23. The ultra short term peak load prediction system of claim 22, wherein said time threshold is selected to be 5 minutes; the setpoint load takes 288 points.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710874560.3A CN107730031B (en) | 2017-09-25 | 2017-09-25 | Ultra-short-term peak load prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710874560.3A CN107730031B (en) | 2017-09-25 | 2017-09-25 | Ultra-short-term peak load prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107730031A true CN107730031A (en) | 2018-02-23 |
| CN107730031B CN107730031B (en) | 2022-08-09 |
Family
ID=61207953
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710874560.3A Active CN107730031B (en) | 2017-09-25 | 2017-09-25 | Ultra-short-term peak load prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107730031B (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108898273A (en) * | 2018-05-29 | 2018-11-27 | 国网能源研究院有限公司 | A kind of user side load characteristic clustering evaluation method based on morphological analysis |
| CN108932557A (en) * | 2018-04-28 | 2018-12-04 | 云南电网有限责任公司临沧供电局 | A kind of Short-term Load Forecasting Model based on temperature cumulative effect and grey relational grade |
| CN109409574A (en) * | 2018-09-27 | 2019-03-01 | 深圳供电局有限公司 | Processing method of load fluctuation characteristics |
| CN109935338A (en) * | 2019-03-07 | 2019-06-25 | 平安科技（深圳）有限公司 | Data prediction processing method, device and computer equipment based on machine learning |
| CN110068759A (en) * | 2019-05-22 | 2019-07-30 | 四川华雁信息产业股份有限公司 | A kind of fault type preparation method and device |
| CN110826750A (en) * | 2018-08-08 | 2020-02-21 | 阿里巴巴集团控股有限公司 | Power load prediction method, device, equipment and system |
| CN111144650A (en) * | 2019-12-26 | 2020-05-12 | 南京工程学院 | Power load prediction method, device, computer readable storage medium and equipment |
| CN111210059A (en) * | 2019-12-26 | 2020-05-29 | 国网北京市电力公司 | Bus daily maximum load processing method and device |
| CN111428926A (en) * | 2020-03-23 | 2020-07-17 | 国网江苏省电力有限公司镇江供电分公司 | Regional power load prediction method considering meteorological factors |
| CN111784028A (en) * | 2020-06-08 | 2020-10-16 | 深圳供电局有限公司 | Community load prediction method |
| WO2020224111A1 (en) * | 2019-05-07 | 2020-11-12 | 深圳大学 | Multivariate time series prediction method |
| CN117239739A (en) * | 2023-11-13 | 2023-12-15 | 国网冀北电力有限公司 | Method, device and equipment for predicting user side load by knowledge big model |
| CN117436707A (en) * | 2023-12-18 | 2024-01-23 | 厦门锋联信息技术有限公司 | Fire safety management method and system based on artificial intelligence |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383023A (en) * | 2008-10-22 | 2009-03-11 | 西安交通大学 | Neural network short-term electric load prediction based on sample dynamic organization and temperature compensation |
| US20130110756A1 (en) * | 2011-10-31 | 2013-05-02 | Siemens Corporation | Short-term Load Forecast Using Support Vector Regression and Feature Learning |
| CN105069525A (en) * | 2015-07-30 | 2015-11-18 | 广西大学 | All-weather 96-point daily load curve prediction and optimization correction system |
| CN106971240A (en) * | 2017-03-16 | 2017-07-21 | 河海大学 | The short-term load forecasting method that a kind of variables choice is returned with Gaussian process |
-
2017
- 2017-09-25 CN CN201710874560.3A patent/CN107730031B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383023A (en) * | 2008-10-22 | 2009-03-11 | 西安交通大学 | Neural network short-term electric load prediction based on sample dynamic organization and temperature compensation |
| US20130110756A1 (en) * | 2011-10-31 | 2013-05-02 | Siemens Corporation | Short-term Load Forecast Using Support Vector Regression and Feature Learning |
| CN105069525A (en) * | 2015-07-30 | 2015-11-18 | 广西大学 | All-weather 96-point daily load curve prediction and optimization correction system |
| CN106971240A (en) * | 2017-03-16 | 2017-07-21 | 河海大学 | The short-term load forecasting method that a kind of variables choice is returned with Gaussian process |
Non-Patent Citations (1)
| Title |
|---|
| 刘旭: "基于实时气象因素的短期负荷预测方法研究", 《中国优秀硕士学位论文全文数据库 工程科技辑》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108932557A (en) * | 2018-04-28 | 2018-12-04 | 云南电网有限责任公司临沧供电局 | A kind of Short-term Load Forecasting Model based on temperature cumulative effect and grey relational grade |
| CN108898273A (en) * | 2018-05-29 | 2018-11-27 | 国网能源研究院有限公司 | A kind of user side load characteristic clustering evaluation method based on morphological analysis |
| CN108898273B (en) * | 2018-05-29 | 2022-04-15 | 国网能源研究院有限公司 | User side load characteristic clustering evaluation method based on morphological analysis |
| CN110826750A (en) * | 2018-08-08 | 2020-02-21 | 阿里巴巴集团控股有限公司 | Power load prediction method, device, equipment and system |
| CN110826750B (en) * | 2018-08-08 | 2023-09-26 | 阿里巴巴集团控股有限公司 | Power load prediction method, device, equipment and system |
| CN109409574A (en) * | 2018-09-27 | 2019-03-01 | 深圳供电局有限公司 | Processing method of load fluctuation characteristics |
| CN109409574B (en) * | 2018-09-27 | 2022-02-22 | 深圳供电局有限公司 | Processing method of load fluctuation characteristics |
| CN109935338A (en) * | 2019-03-07 | 2019-06-25 | 平安科技（深圳）有限公司 | Data prediction processing method, device and computer equipment based on machine learning |
| WO2020224111A1 (en) * | 2019-05-07 | 2020-11-12 | 深圳大学 | Multivariate time series prediction method |
| CN110068759A (en) * | 2019-05-22 | 2019-07-30 | 四川华雁信息产业股份有限公司 | A kind of fault type preparation method and device |
| CN111210059A (en) * | 2019-12-26 | 2020-05-29 | 国网北京市电力公司 | Bus daily maximum load processing method and device |
| CN111144650A (en) * | 2019-12-26 | 2020-05-12 | 南京工程学院 | Power load prediction method, device, computer readable storage medium and equipment |
| CN111428926A (en) * | 2020-03-23 | 2020-07-17 | 国网江苏省电力有限公司镇江供电分公司 | Regional power load prediction method considering meteorological factors |
| CN111784028A (en) * | 2020-06-08 | 2020-10-16 | 深圳供电局有限公司 | Community load prediction method |
| CN117239739A (en) * | 2023-11-13 | 2023-12-15 | 国网冀北电力有限公司 | Method, device and equipment for predicting user side load by knowledge big model |
| CN117239739B (en) * | 2023-11-13 | 2024-02-02 | 国网冀北电力有限公司 | Method, device and equipment for predicting user side load by knowledge big model |
| CN117436707A (en) * | 2023-12-18 | 2024-01-23 | 厦门锋联信息技术有限公司 | Fire safety management method and system based on artificial intelligence |
| CN117436707B (en) * | 2023-12-18 | 2024-03-22 | 厦门锋联信息技术有限公司 | Fire safety management method and system based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107730031B (en) | 2022-08-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107730031B (en) | Ultra-short-term peak load prediction method and system | |
| CN111860979B (en) | Short-term load prediction method based on TCN and IPSO-LSSVM combined model | |
| CN111222698B (en) | Internet of things-oriented ponding water level prediction method based on long-time and short-time memory network | |
| CN105069525B (en) | Round-the-clock 96 Day Load Curve Forecastings and optimization update the system | |
| CN112288164B (en) | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast | |
| CN112001439A (en) | GBDT-based shopping mall building air conditioner cold load prediction method, storage medium and equipment | |
| CN101383023B (en) | Neural network short-term electric load prediction based on sample dynamic organization and temperature compensation | |
| CN107967542B (en) | Long-short term memory network-based electricity sales amount prediction method | |
| CN111428926B (en) | Regional power load prediction method considering meteorological factors | |
| CN111105104A (en) | Short-term power load prediction method based on similar day and RBF neural network | |
| CN109359786A (en) | A kind of power station area short-term load forecasting method | |
| CN105117810A (en) | Residential electricity consumption mid-term load prediction method under multistep electricity price mechanism | |
| CN111525587B (en) | Reactive load situation-based power grid reactive voltage control method and system | |
| Sülo et al. | Energy efficient smart buildings: LSTM neural networks for time series prediction | |
| WO2015081660A1 (en) | Method for forecasting residential quarter short-term load | |
| CN109146063B (en) | Multi-segment short-term load prediction method based on important point segmentation | |
| CN113705877A (en) | Real-time monthly runoff forecasting method based on deep learning model | |
| CN115099511A (en) | Photovoltaic power probability estimation method and system based on optimized copula | |
| CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
| CN111915083B (en) | Wind power prediction method and prediction system based on time layered combination | |
| CN117977568A (en) | Power load prediction method based on nested LSTM and quantile calculation | |
| CN115222106A (en) | User day-ahead load prediction method of self-adaptive model | |
| CN116523142A (en) | Photovoltaic cluster interval prediction method considering space-time characteristic clustering | |
| CN111680818B (en) | Short-term reactive load prediction method and system | |
| CN115796915A (en) | Electricity price prediction method and system for electricity trading market |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: No. 15, Qinghe Xiaoying East Road, Haidian District, Beijing Applicant after: CHINA ELECTRIC POWER RESEARCH INSTITUTE Co.,Ltd. Applicant after: STATE GRID CORPORATION OF CHINA Address before: No. 15, Qinghe Xiaoying East Road, Haidian District, Beijing Applicant before: China Electric Power Research Institute Applicant before: State Grid Corporation of China |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190626 Address after: No. 15, Qinghe Xiaoying East Road, Haidian District, Beijing Applicant after: CHINA ELECTRIC POWER RESEARCH INSTITUTE Co.,Ltd. Applicant after: STATE GRID CORPORATION OF CHINA Applicant after: CENTRAL CHINA BRANCH OF STATE GRID CORPORATION OF CHINA Address before: No. 15, Qinghe Xiaoying East Road, Haidian District, Beijing Applicant before: CHINA ELECTRIC POWER RESEARCH INSTITUTE Co.,Ltd. Applicant before: State Grid Corporation of China |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |