CN116050604A - Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity - Google Patents
Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity Download PDFInfo
- Publication number
- CN116050604A CN116050604A CN202211740856.3A CN202211740856A CN116050604A CN 116050604 A CN116050604 A CN 116050604A CN 202211740856 A CN202211740856 A CN 202211740856A CN 116050604 A CN116050604 A CN 116050604A
- Authority
- CN
- China
- Prior art keywords
- power
- wind
- water
- time
- forecasting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013277 forecasting method Methods 0.000 title claims abstract description 30
- 230000000295 complement effect Effects 0.000 claims abstract description 130
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 50
- 238000013136 deep learning model Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 61
- 238000000034 method Methods 0.000 claims description 61
- 241000039077 Copula Species 0.000 claims description 22
- 238000005315 distribution function Methods 0.000 claims description 22
- 238000013507 mapping Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 108090000672 Annexin A5 Proteins 0.000 claims description 9
- 238000013135 deep learning Methods 0.000 claims description 9
- 230000007774 longterm Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 230000001186 cumulative effect Effects 0.000 claims description 5
- 238000013480 data collection Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000001970 hydrokinetic effect Effects 0.000 claims description 3
- 238000013139 quantization Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 description 12
- 230000005611 electricity Effects 0.000 description 9
- 230000005622 photoelectricity Effects 0.000 description 9
- 238000009825 accumulation Methods 0.000 description 8
- 230000001360 synchronised effect Effects 0.000 description 8
- 230000003068 static effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000005291 magnetic effect Effects 0.000 description 3
- 235000019800 disodium phosphate Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005294 ferromagnetic effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Analysis (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Computational Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Evolutionary Computation (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Optimization (AREA)
- Biophysics (AREA)
- Pure & Applied Mathematics (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Databases & Information Systems (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Evolutionary Biology (AREA)
- Development Economics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
Abstract
The invention provides a water and wind power combined forecasting method, device and equipment considering space-time complementarity, which comprises the steps of collecting and arranging water and wind power multi-energy complementary system data, quantifying the complementarity of water and electric power, wind power and photoelectric power in the water and wind power multi-energy complementary system, considering the combined influence of large-scale hydrological weather factors on the water and wind power conditions of corresponding watershed of the water and wind power multi-energy complementary system, considering the autocorrelation of the water and electric power, wind power, photoelectric power or total power in time, obtaining the space-time complementary factors related to the water and wind power multi-energy complementary system power, constructing a point forecasting model and an interval forecasting model based on a deep learning model, and independently forecasting the water and wind power or the photoelectric power by combining the space-time complementary factors, and directly forecasting the total power in a medium-long term by taking the total power as a label.
Description
Technical Field
The invention relates to the technical field of power forecasting of electric power systems, in particular to a water-wind-solar power combined forecasting method, device and equipment considering space-time complementarity.
Background
The water-wind-solar multi-energy complementary operation is an important approach for solving the problem of new energy consumption and realizing the aim of double carbon. In the hybrid system, the hydropower, wind power and the photoelectricity are bundled into total power and are transmitted to the power grid, and the accurate forecast of the total power in the middle and long periods has a non-negligible effect on the establishment of electricity prices, the trade of electric power markets, the arrangement of power generation plans of the power grid, the management of operation modes and the like.
The water-wind-solar multi-energy complementary system fully utilizes the time complementarity of runoff, wind energy and light energy resources, and meets the load demand in a complementary scheduling mode of hydroelectric compensation wind power and photoelectricity, so that the wind power, photoelectricity and hydropower in the water-wind-solar multi-energy complementary system have certain complementarity. For power forecasting of hydropower, wind power and photoelectric single-type power stations, the method can be divided into a physical model and a statistical model. If the medium-long term power forecast of the water-wind-solar multi-energy complementary system is carried out based on a physical model, a large amount of data such as local runoff, wind speed, photovoltaic, temperature and the like need to be collected, and the data need to be converted into power based on a physical relationship, so that the problems that the data collection difficulty is high and the physical mechanism is difficult to be clear are solved; if the power forecast is performed based on the statistical model, the multiple uncertainties of the input of the system bring difficulty to the feature selection because the system is influenced by more various and complex elements such as hydrology, weather and various power stations. In addition, for the total power forecast of the water, wind and light multi-energy complementary system, the research on the total power forecast is relatively less at present, the total power of the system is usually obtained by accumulating the separately forecasted hydropower, wind power and photoelectricity, but the total power forecast accuracy obtained by accumulation is not high due to the strong randomness and volatility of the single power.
However, in the past, no matter for single power forecast or total power forecast, complementarity of hydropower, wind power and photoelectricity of the water-wind-solar energy complementary system in resources and scheduling modes is fully utilized, and the large-scale hydrological meteorological factors are used as meteorological circulation anomalies of remote areas, have obvious combined influence on water-wind conditions of a flow field, can solve forecast difficulty caused by uncertainty of multiple inputs of the water-wind-solar energy complementary system, and do not fully consider the effects of the large-scale hydrological meteorological factors in related technologies.
Disclosure of Invention
According to the defects of the prior art, the invention aims to provide the water-wind-light power combined forecasting method, the device and the equipment which consider space-time complementarity, fully consider complementarity and large-scale hydrological meteorological factors in a water-wind-light multi-energy complementary system, and improve forecasting precision.
In order to solve the technical problems, the invention adopts the following technical scheme:
the water-wind-light power combined forecasting method considering space-time complementarity comprises the following steps:
Step 2, taking hydroelectric power, wind power, photoelectric power or total power as a target power forecast object, and quantifying complementarity of the hydroelectric power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system;
step 3, considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watershed of the water-wind-solar multi-energy complementary system, and screening the large-scale hydrological factors of the corresponding watershed;
step 4, considering the autocorrelation of hydroelectric power, wind power, photoelectric power or total power in time;
step 5, constructing a mapping relation of a power prediction model, wherein the mapping relation considers the complementarity of the water-electricity power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system, and the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, wind power, photoelectric power or total power in time, so that space-time complementary factors related to the power of the water-wind-solar multi-energy complementary system are obtained;
step 6, constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
and 7, based on the point prediction model and the interval prediction model, carrying out independent prediction on the hydroelectric power, the wind power or the photoelectric power according to space-time complementary factors, and directly carrying out medium-long term prediction on the total power by taking the total power as a label.
Further, in step 2, complementarity of the hydroelectric power, the wind power, and the photovoltaic power is quantified based on the Copula function.
Further, the step 2 specifically includes:
step 201, determining the edge distribution of hydroelectric power, wind power and photoelectric power based on nuclear density estimation, wherein the nuclear density estimation formula is as follows:
wherein x is i Sample points with X distributed in the same way; k (t) is a kernel function; h is window width; n is the number of sample points; the kernel function K (t) needs to satisfy in the real domain:
step 202, solving the two-to-two variable joint distribution based on a Copula function, wherein the calculation formula is as follows:
in the method, in the process of the invention,
is an n-dimensional joint distribution function; />
The edge distribution function is the water power, wind power and photoelectric power; c is the Copula function of these variables;
selecting Euclidean distance d 2 And (3) evaluating the fitting degree:
wherein C is n (u, v) is an empirical Copula function, I is an oscillometric function, when F n (x ic ) If u is less than or equal to u, I is 1, otherwise 0;
based on the established joint distribution function, a conditional distribution of X.ltoreq.x when the variable Y=y is known can be established, the variable X, Y is any two of the water electric power, the wind power and the photoelectric power, and the binary conditional probability distribution function is shown as follows:
where f (X, Y) is a joint probability density function of the variables X and Y; f (x), f (y) are probability density functions of the variables X, Y, respectively; u, v are cumulative distribution functions of variables X and Y; c (u, v) is a probability density function of the Copula function;
Step 203, based on the binary conditional probability distribution function, establishing a conditional expectation distribution function as shown in the following formula:
based on equation (7), the average value of variable X when given variable y=y can be found, and the conditional expectation value of another relevant variable corresponding to the known distribution value of hydroelectric power, wind power or photovoltaic power can be found, thereby quantifying the complementarity between two by two in the water-wind-solar multi-energy complementary system based on the expected value, and determining the power station power p= { P complementary to the hydroelectric power, wind power or photovoltaic power based on the expected value 1 ,P 2 ,...,P n1 }。
Further, in step 3, large scale hydrological factors corresponding to the basin are screened out based on the maximum mutual information coefficient
The step 3 specifically comprises the following steps:
step 301, there is an ordered pair dataset D = { (x) mi ,y mi ) I=1, 2, …, n }, by dividing the X-axis into X m The Y axis is divided into Y m Parts, get a x m ×y m Calculating the mutual information value of each grid:
wherein p (x) m ) And p (y) m ) As variable X m 、Y m Is a function of the edge probability density of (2); p (x) m ,y m ) The joint probability density function of two variables is the ratio of the number of points in the current grid to the total number of points of the data;
step 302, normalizing the maximum value row of the mutual information in the grid:
step 303, performing grid division of different schemes, and repeating step 301 and step 302, wherein the maximum normalized mutual information value in all schemes is the maximum mutual information coefficient:
Wherein x is m y m <B is a constraint condition of total grid division, and B is set to be 0.6 th power of the total data.
Further, in step 5, the space-time complementary factor includes the ith power station power P of the t-th period that is complementary to the target power forecast object i (t) jth large scale hydrokinetic factor LC of the tth period j (t) and historical time sequences N (t) of the target power forecast object in the t period, wherein the obtained mapping relation is as follows:
N(t)＝f(P 1 (t-1),...,P n1 (t-m),LC 1 (t-1),...,LC n2 (t-m),N(t-1),...,N(t-m))(
11)
wherein f is a simulation function of a deep learning method; n is n 1 To work with waterThe number of power stations with the complementation of the rate, wind power or photoelectric power; n is n 2 The number of the large-scale hydrological weather forecast factors is the number; m is a lag period of the space-time complementary factor, and the space-time complementary factor of the previous m period is used for forecasting the power of a target power forecasting object of the next period.
Further, in step 6, an interval prediction model is constructed by adopting an upper and lower limit estimation method, and deep learning is trained towards a direction with better prediction interval performance by constructing a loss function, and considering that the prediction interval needs to meet a criterion of smaller interval width on the basis of covering more real values as much as possible, the adopted loss function is shown in the following formula:
Loss＝f 1 +f 2 (12)
f 2 ＝k 2 u q (t)-l q (t) (14)
wherein f 1 The distance between the median value and the actual value of the interval is expressed, the closer the median value is to the actual value of the interval, the more accurate the result of the forecast interval is, and meanwhile, the distance between the median value and the actual value of the interval is f 1 A penalty factor lambda is also arranged in q Punishment is carried out on the condition that the actual value is not in the forecast interval, so that the coverage rate of the interval is improved in the network training process; f (f) 2 The method is used for calculating the interval width, and the narrower the interval width is under the condition that the interval coverage rate is the same, the higher the quality of the corresponding prediction interval is; by adjusting k 1 、k 2 The proportionality coefficient can lead the network to purposefully deviate to a certain index for training, when k is 1 、k 2 When the proportion is proper, the penalty factor lambda is adjusted q The coverage rate of the actual value of the interval can be combined with the width; n (t) represents the actual power value of the t period, and the unit is MW; u (u) q (t)、l q (t) is a forecast result of the model in a t period, and the forecast result respectively represents a forecast upper boundary and a forecast lower boundary; lambda (lambda) q Is a penalty coefficient; k (k) 1 、k 2 Are all proportionality coefficients.
Further, the method also comprises a step 8 of evaluating various indexes of Nash efficiency coefficient NSE, deviation coefficient Bias, root mean square error RMSE and mean absolute error MAE for point prediction;
for the section forecast, an improved coverage width comprehensive criterion CWC, a section coverage rate PICP and a section standardized average width PINAW index are adopted for evaluation, wherein a specific calculation formula is as follows;
where N (t) represents the actual value of power, MW, of the t-th period; t is the total period number of samples; 
Mean value of power, MW; />
Beta and eta are penalty coefficients of CWC; VR is the difference between the maximum and minimum values of the forecast samples; μ is confidence level;
and (3) comparing the long-term result in the total power obtained in the step (7) with the accumulated result of the hydroelectric power, the wind power and the photoelectric power.
The water, wind and light power combined forecasting device considering space-time complementarity comprises:
the data collection module is used for collecting and arranging data of the water, wind and light multi-energy complementary system, and comprises large-scale hydrological factors, total power time sequence data, hydropower power time sequence data, wind power time sequence data and photoelectric power time sequence data of the water, wind and light multi-energy complementary system;
the power complementary quantization module is used for quantizing the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-light multi-energy complementary system by taking the water-electricity power, the wind power, the photoelectric power or the total power as a target power forecast object;
the large-scale hydrological meteorological factor determining module is used for screening the large-scale hydrological meteorological factors of corresponding watersheds by considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watersheds of the water-wind-solar multi-energy complementary system;
the self-correlation determination module is used for considering the self-correlation of hydroelectric power, wind power, photoelectric power or total power in time;
The space-time complementary factor determining module is used for constructing a mapping relation of a power prediction model, the mapping relation considers the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-solar multi-energy complementary system, the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, the wind power, the photoelectric power or the total power in time is used for acquiring the space-time complementary factor related to the power of the water-wind-solar multi-energy complementary system;
the deep learning module is used for constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
the power forecasting module is used for independently forecasting the hydropower power, the wind power or the photoelectric power according to the space-time complementary factors based on the point forecasting model and the interval forecasting model, and directly forecasting the total power for a medium-long term by taking the total power as a label.
The water-wind-solar power combined forecasting device considering space-time complementarity comprises a processor and a memory, wherein the memory is used for storing a computer program capable of running on the processor, and the processor is used for executing the steps of the water-wind-solar power combined forecasting method considering space-time complementarity when running the computer program.
A computer storage medium having a computer program stored therein, wherein the computer program, when executed by a processor, implements the method for joint prediction of water, wind and solar power taking into account space-time complementarity according to any one of the above.
Compared with the prior art, the invention has the following advantages and beneficial effects:
(1) The water-wind-light power combined forecasting method, device and equipment considering space-time complementarity fully considers the complementarity of water electricity, wind electricity and photoelectricity in the water-wind-light multi-energy complementary system in terms of resources and scheduling modes, and improves forecasting precision.
(2) According to the water, wind and light power combined forecasting method, device and equipment considering space-time complementarity, the large-scale hydro-meteorological factors are added as the space-time complementarity factors, the combined influence of the large-scale hydro-meteorological factors on the water, wind and light conditions of the river basin and the long-term forecasting capability are utilized, and the long-term power forecasting effect of the water, wind and light multifunctional complementary system is improved.
(3) According to the water-wind-solar power combined forecasting method, device and equipment considering space-time complementarity, when independent forecasting of water-electric power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system is carried out, the space-time complementarity is considered for forecasting, so that the accuracy of forecasting the power of a single type of power station is improved, and the forecasting accuracy of the total power obtained by a traditional summation method is improved.
(4) According to the water-wind-solar power combined forecasting method, the device and the equipment considering space-time complementarity, when the total power forecasting of the water-wind-solar multi-energy complementary system is carried out, the total power is obtained through the combined forecasting method taking the total power as a label based on space-time complementary factors, the situation that the hydropower power, the wind power and the photoelectric power with obvious fluctuation and randomness are respectively forecasted is avoided, and the method is more direct and simpler, and meanwhile, more accurate total power point forecasting and interval forecasting compared with the traditional accumulation method are realized.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application. The exemplary embodiments of the present invention and the descriptions thereof are for explaining the present invention and do not constitute an undue limitation of the present invention. In the drawings:
FIG. 1 is a flow chart of a water-wind-light power joint forecasting method considering space-time complementarity;
FIG. 2 (a) is a schematic diagram of expected values of hydroelectric power under probability distribution of photoelectric power in the water-wind-solar multi-energy complementary system of the invention;
FIG. 2 (b) is a schematic diagram of expected values of hydroelectric power under the probability distribution of wind power in the water-wind-solar multi-energy complementary system;
FIG. 2 (c) is a schematic diagram of expected value of photoelectric power under probability distribution of wind power in the water-wind-solar multi-energy complementary system;
FIG. 3 (a) is a graph showing the point prediction result of the hydroelectric power of the water-wind-solar energy multi-energy complementary system;
FIG. 3 (b) is a point forecast result of wind power of the water-wind-solar multi-energy complementary system;
FIG. 3 (c) is a point forecast result of the photoelectric power of the water-wind-solar multi-energy complementary system of the invention;
FIG. 4 (a) is a section forecast result of the hydroelectric power of the water-wind-solar multi-energy complementary system of the invention;
FIG. 4 (b) is a section forecast result of wind power of the water-wind-solar multi-energy complementary system;
FIG. 4 (c) is a graph showing the result of forecasting the region of the photoelectric power of the water-wind-solar multi-energy complementary system;
FIG. 5 is a comparison of point forecast results of total power of the water-wind-solar system obtained by a combined forecast method and an accumulation method;
fig. 6 is a comparison of the interval forecast results of the total power of the water-wind-solar system obtained by the combined forecast method and the accumulation method.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The invention provides a water-wind-light power combined forecasting method considering space-time complementarity, which is shown in figure 1 and comprises the following steps:
Step 2, taking hydroelectric power, wind power, photoelectric power or total power as a target power forecast object, and quantifying complementarity of the hydroelectric power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system;
step 3, considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watershed of the water-wind-solar multi-energy complementary system, and screening the large-scale hydrological factors of the corresponding watershed;
step 4, considering the autocorrelation of hydroelectric power, wind power, photoelectric power or total power in time;
step 5, constructing a mapping relation of a power prediction model, wherein the mapping relation considers the complementarity of the water-electricity power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system, and the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, wind power, photoelectric power or total power in time, so that space-time complementary factors related to the power of the water-wind-solar multi-energy complementary system are obtained;
step 6, constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
and 7, based on the point prediction model and the interval prediction model, carrying out independent prediction on the hydroelectric power, the wind power or the photoelectric power according to space-time complementary factors, and directly carrying out medium-long term prediction on the total power by taking the total power as a label.
The water-wind-light power combined forecasting method considering space-time complementarity fully considers the complementarity of water electricity, wind electricity and photoelectricity in the water-wind-light multi-energy complementary system in terms of resources and scheduling modes, and improves forecasting precision.
According to the water, wind and light power combined forecasting method considering space-time complementarity, the large-scale hydrological meteorological factors are added to serve as space-time complementarity factors, and the combined influence and the long-term forecasting capability of the large-scale hydrological meteorological factors on the water, wind and light conditions of the river basin are utilized, so that the long-term power forecasting effect of the water, wind and light multifunctional complementary system is improved.
According to the water-wind-solar power combined forecasting method considering space-time complementarity, when the total power forecasting of the water-wind-solar multi-energy complementary system is carried out, the total power is obtained through the combined forecasting method taking the total power as a label based on space-time complementary factors, the situation that the water-electricity power, the wind-electricity power and the photoelectric power with obvious fluctuation and randomness are respectively forecasted is avoided, the method is more direct and simple, and meanwhile, more accurate total power point forecasting and interval forecasting compared with the traditional accumulation method are realized.
In one embodiment of the invention, taking a two-beach water-wind-solar multi-energy complementary system as an example, due to lack of system power actual data with long time sequence, firstly, runoff, wind speed, photovoltaic and the like are converted into time sequence data of hydroelectric power, wind power and photoelectric power of the water-wind-solar multi-energy complementary system through medium-long term optimized scheduling, and the time sequence data of the total power of the water-wind-solar multi-energy complementary system is obtained through accumulation, so that the next forecast research is carried out on the basis of the time sequence data.
According to the water, wind and light power combined forecasting method considering space-time complementarity, in the step 1, a long-short-term memory network (LSTM) model is selected as a deep learning model for carrying out combined forecasting of medium-long term power, data required by arrangement are collected and collected, the data comprise 96 items of large-scale hydrological meteorological factor data and time sequence data of total power, hydropower power, wind power and photoelectric power of a water, wind and light multifunctional complementary system, and the data used in the embodiment are month data from 1 month in 1959 to 12 months in 2010.
In the invention, in step 2, complementarity of the hydroelectric power, the wind power and the photoelectric power is quantified based on a Copula function.
Specifically, the step 2 specifically includes:
step 201, determining the edge distribution of hydroelectric power, wind power and photoelectric power based on nuclear density estimation, wherein the nuclear density estimation formula is as follows:
wherein x is i Sample points with X distributed in the same way; k (t) is a kernel function; h is window width; n is the same as the sampleThe number of the points; the kernel function K (t) needs to satisfy in the real domain:
step 202, after the single variable edge distribution of the hydroelectric power, the wind power and the photoelectric power is determined, carrying out solution on the two-to-two variable joint distribution based on a Copula function, wherein the calculation formula is as follows:
In the method, in the process of the invention,
is an n-dimensional joint distribution function; />
The edge distribution function is the water power, wind power and photoelectric power; c is a Copula function of the connection variable;
selecting Euclidean distance d 2 And (3) evaluating the fitting degree:
wherein C is n (u, v) is an empirical Copula function, I is an oscillometric function, when F n (x ic ) If u is less than or equal to u, I is 1, otherwise 0;
based on the established joint distribution function, a conditional distribution of x.ltoreq.x when the variable y=y is known can be established, and the variable X, Y is any two of the water electric power, the wind electric power and the photoelectric power, and the binary conditional probability distribution function is shown as the following formula.
Where f (X, Y) is a joint probability density function of the variables X and Y; f (x), f (y) are probability density functions of the variables X, Y, respectively; u, v are cumulative distribution functions of variables X and Y; c (u, v) is a probability density function of the Copula function;
step 203, based on the binary conditional probability distribution function, a conditional expectation distribution function is established as shown in the following formula.
Based on equation (7), the average value of variable X when given variable y=y can be found, and the conditional expectation value of another relevant variable corresponding to the known distribution value of hydroelectric power, wind power or photovoltaic power can be found, thereby quantifying the complementarity between two by two in the water-wind-solar multi-energy complementary system based on the expected value, and determining the power station power p= { P complementary to the hydroelectric power, wind power or photovoltaic power based on the expected value 1 ,P 2 ,...,P n1 }。
According to the method, based on the Copula function, when the independent prediction of the power of each single type of power station of the water-wind-solar multi-energy complementary system is carried out, the complementarity among the water-electricity power, the wind power and the photoelectric power is considered, other significant relevant power station powers in the water-wind-solar multi-energy complementary system are added into the prediction factor, so that the accuracy of the power prediction of the single type of power station is improved, and the prediction accuracy of the total power obtained by the traditional addition method is improved.
In the embodiment of the invention, in step 2, three commonly used Archimedean Copula functions are adopted to fit the joint distribution of hydroelectric power-photoelectric power, hydroelectric power-wind power and photoelectric power-wind power, and the parameter values of the Copula functions are estimated based on a maximum likelihood method MLE, and the calculation results are shown in table 1.
TABLE 1Copula function parameter estimation values
In order to select a proper Copula function corresponding to the relation of the three functions to fit, the Euclidean distance result of the Copula function and the actual experience distribution is calculated as shown in table 2.
TABLE 2 Euclidean distance between wind, light and water output using different Copula functions
As can be seen from Table 2, the d of the Frank Copula function is for both the hydroelectric power-photovoltaic power and the hydroelectric power-wind power 2 Minimum, d of gumme-Hougaard function for photoelectric power-wind power 2 Minimum. Therefore, the Frank Copula function is selected to fit the combined distribution of the hydroelectric power, the photoelectric power and the hydroelectric power, the wind power, and the gumm-Hougaard Copula function is selected to fit the combined distribution of the photoelectric power, the wind power.
Based on the selected Copula function and equation (7), a conditional desired distribution function between different outputs can be established, and the images are plotted as shown in fig. 2 (a) -2 (c).
As can be seen from fig. 2 (a) -2 (c), as the probability distribution value of the photoelectric power and the wind power is larger and larger, that is, the output value of the corresponding photoelectric power and wind power is larger and larger, the expected value of the condition corresponding to the hydroelectric power is smaller and smaller, and the complementarity of the hydroelectric power, the wind power and the photoelectric power is quantized.
In the invention, in step 3, large scale hydrological factors LC= { LC of corresponding watershed are screened out based on the maximum Mutual Information Coefficient (MIC) 1 ,LC 2 ,...,LC n2 }。
The step 3 specifically comprises the following steps:
step 301, there is oneThe ordered pair data set d= { (x) mi ,y mi ) I=1, 2, …, n }, by dividing the X-axis into X m The Y axis is divided into Y m Parts, get a x m ×y m Calculating the mutual information value of each grid:
Wherein p (x) m ) And p (y) m ) As variable X m 、Y m Is a function of the edge probability density of (2); p (x) m ,y m ) The joint probability density function of two variables is the ratio of the number of points in the current grid to the total number of points of the data;
step 302, normalizing the maximum value row of the mutual information in the grid:
step 303, performing grid division of different schemes, and repeating step 301 and step 302, wherein the maximum normalized mutual information value in all schemes is the maximum mutual information coefficient:
wherein x is m y m <B is a constraint condition of total grid division, and B is set to be 0.6 th power of the total data.
In the related art, only a linear relation can be described by a linear relation coefficient, and in the invention, the maximum Mutual Information Coefficient (MIC) is an index for measuring the relation coefficient between two variables based on mutual information. The MIC can measure the linear and nonlinear relations between two variables, is not easily affected by abnormal values in data, and has the characteristics of universality, robustness and fairness.
In the embodiment of the invention, in step 3, large-scale hydrological forecasting factors of the power of each single type of power station of the system can be obtained through correlation analysis, and the large-scale hydrological forecasting factors are shown in table 3.
Table 3 system each single type power station power large scale hydrological weather forecast factor
In the invention, in step 5, the space-time complementary factor comprises the ith power station power P of the t period and complementary with the target power forecast object i (t) jth large scale hydrokinetic factor LC of the tth period j (t) and historical time sequences N (t) of the target power forecast object in the t period, wherein the obtained mapping relation is as follows:
N(t)＝f(P 1 (t-1),...,P n1 (t-m),LC 1 (t-1),...,LC n2 (t-m),N(t-1),...,N(t-m))(
11)
wherein f is a simulation function of a deep learning method; n is n 1 The number of power stations complementary to the target power forecast object; n is n 2 The number of the large-scale hydrological weather forecast factors is the number; m is a lag period of the space-time complementary factor, and the space-time complementary factor of the previous m period is used for forecasting the power of a target power forecasting object of the next period.
In step 6, for the section prediction, an interval prediction model is constructed by adopting an upper and lower limit estimation method, the deep learning is trained towards the direction of better performance of the prediction interval by constructing a loss function, and considering that the prediction interval needs to meet the criterion of smaller interval width on the basis of covering more real values as much as possible, the adopted loss function is shown in the following formula:
Loss＝f 1 +f 2 (12)
f 2 ＝k 2 u q (t)-l q (t) (14)
wherein f 1 The distance between the median value and the actual value of the interval is expressed, the closer the median value is to the actual value of the interval, the more accurate the result of the forecast interval is, and meanwhile, the distance between the median value and the actual value of the interval is f 1 A penalty factor lambda is also arranged in q Punishment is carried out on the condition that the actual value is not in the forecast interval, so that the coverage rate of the interval is improved in the network training process; f (f) 2 The method is used for calculating the interval width, and the narrower the interval width is under the condition that the interval coverage rate is the same, the higher the quality of the corresponding prediction interval is; by adjusting k 1 、k 2 The proportionality coefficient can lead the network to purposefully deviate to a certain index for training, when k is 1 、k 2 When the proportion is proper, the penalty factor lambda is adjusted q The coverage rate of the actual value of the interval can be combined with the width; n (t) represents the actual power value of the t period, and the unit is MW; u (u) q (t)、l q (t) is a forecast result of the model in a t period, and the forecast result respectively represents a forecast upper boundary and a forecast lower boundary; lambda (lambda) q Is a penalty coefficient; k (k) 1 、k 2 Are all proportionality coefficients.
The method also comprises a step 8 of evaluating various indexes of Nash efficiency coefficient NSE, deviation coefficient Bias, root mean square error RMSE and mean absolute error MAE for point prediction;
for interval forecast, evaluating by adopting an improved coverage width comprehensive criterion CWC, an interval coverage PICP and an interval standardized average width PINAW index;
and (3) comparing the long-term result in the total power obtained in the step (7) with the accumulated result of the hydroelectric power, the wind power and the photoelectric power.
Where N (t) represents the actual value of power, MW, of the t-th period;
power forecast, MW, representing a t-th period; t is the total period number of samples; />
Mean value of power, MW; />
Beta and eta are penalty coefficients of CWC; VR is the difference between the maximum and minimum values of the forecast samples; μ is confidence level.
In the section prediction, the higher the coverage rate of the PICP section is, the narrower the normalized average width of the PINAW section is, which means that the better the section prediction effect is, the CWC is used as an index for integrating the coverage rate of the section and the normalized average width of the section, an exponential penalty is given when the PICP is lower than the confidence level mu, and when the PICP exceeds the confidence level mu, the evaluation target is mainly concentrated on the PINAW index of the section width.
However, the original CWC index has a problem in that the influence of the interval width index PINAW on the CWC index is almost ignored when the PICP is lower than the confidence level, and the superiority and inferiority of the model cannot be scientifically evaluated. The improved CWC comprehensive evaluation index is adopted, the proportion of the PINAW index is linearly amplified by adding the beta parameter, and the alpha parameter is used for avoiding the problem that the influence of PICP caused by too small PINAW is ignored, so that the forecast interval is evaluated more reasonably.
In the embodiment of the invention, the period from 5 months in 1959 to 1 month in 2003 is set as the calibration period, and the period from 2 months in 2003 to 4 months in 2010 is set as the inspection period. In order to verify the influence of the space-time complementary factors on the power forecasting effect of the single-type power station, different forecasting schemes are set as shown in table 4, and the point forecasting results of the power of each single-type power station of the system under different schemes are obtained after the parameters in the LSTM model are adjusted as shown in table 5.
Table 4 predictor scheme settings for power of a single type of plant
Table 5 comparison of Point forecast effects for System Single type Power station Power under different forecast schemes
From table 5, it can be seen that, by using the space-time complementary factor obtained by the invention, the forecasting effect of each single type of power station power in the water-wind-solar multi-energy complementary system is generally improved by adopting the water-wind-solar power combined forecasting method considering the space-time complementarity.
Table 6 shows error indexes of point prediction results of power of each single type of power station of the water-wind-solar multi-energy complementary system, fig. 3 (a) -3 (c) show point prediction results of water-wind power, wind power and photoelectric power of the water-wind-solar multi-energy complementary system, table 7 shows error indexes of interval prediction results of power of each single type of power station of the water-wind-solar multi-energy complementary system at 90% confidence level, and fig. 4 (a) -4 (c) show interval prediction results of water-wind power, wind power and photoelectric power of the water-wind-solar multi-energy complementary system.
TABLE 6 error index of Point forecast results for Power of each single type of Power station of Water-wind-solar Multi-energy complementary System
As can be seen from fig. 3 and table 6, the correlation between the hydropower and the corresponding forecasting factors is strong, and the forecasting effect of both are good; the forecasting result of wind power and photoelectric power can only reflect the overall change trend of actual power, and the forecasting effect is relatively poor.
TABLE 7 error index of section forecast result of each single type of power station power of water-wind-solar multi-energy complementary system
As can be seen from fig. 4 and table 7, at the confidence level of 90%, the hydroelectric power, the wind power and the photovoltaic power all achieve better interval forecasting effects, the forecasting interval can be narrower on the basis of better covering the actual value, and in the interval forecasting, the wind power forecasting effect is the best, the hydroelectric power is the next worst, and the photovoltaic power is the worst.
And 8, comparing the result obtained by accumulating the hydroelectric power, the wind power and the photoelectric power with the total power forecasting result directly obtained by the water-wind-solar power combined forecasting method considering the space-time complementarity, and verifying the superiority of the method in the total power forecasting of the water-wind-solar multi-energy complementary system.
Table 8 shows the error index of the total power prediction result obtained by the present invention and the total power point prediction result obtained by the accumulation prediction method, fig. 5 shows the comparison of the point prediction results of the total power obtained by the two methods, table 9 shows the error index of the total power interval prediction result obtained by the two methods, and fig. 6 shows the comparison of the interval prediction results of the total power obtained by the two methods. Compared with the accumulation forecasting method, the total power obtained by the forecasting method provided by the invention has improved accuracy in point forecasting and interval forecasting, and the effectiveness and superiority of the combined forecasting method in total power forecasting are verified.
Table 8 dot forecast results of total power obtained by the combined forecast method and the cumulative forecast method
Table 9 comparison of the section forecast results for the total Power obtained by the Joint forecast method and the cumulative forecast method
The invention also provides a water-wind-light power combined forecasting device considering space-time complementarity, which comprises:
the data collection module is used for collecting and arranging data of the water, wind and light multi-energy complementary system, and comprises large-scale hydrological factors, total power time sequence data, hydropower power time sequence data, wind power time sequence data and photoelectric power time sequence data of the water, wind and light multi-energy complementary system;
the power complementary quantization module is used for quantizing the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-light multi-energy complementary system by taking the water-electricity power, the wind power, the photoelectric power or the total power as a target power forecast object;
the large-scale hydrological meteorological factor determining module is used for screening the large-scale hydrological meteorological factors of corresponding watersheds by considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watersheds of the water-wind-solar multi-energy complementary system;
the self-correlation determination module is used for determining historical time sequences of the hydroelectric power, the wind power, the photoelectric power or the total power by considering the self-correlation of the hydroelectric power, the wind power, the photoelectric power or the total power in time;
The space-time complementary factor determining module is used for constructing a mapping relation of a power prediction model, the mapping relation considers the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-solar multi-energy complementary system, the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, the wind power, the photoelectric power or the total power in time is used for acquiring the space-time complementary factor related to the power of the water-wind-solar multi-energy complementary system;
the deep learning module is used for constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
the power forecasting module is used for independently forecasting the hydroelectric power, the wind power or the photoelectric power according to the space-time complementary factors based on the point forecasting model and the interval forecasting model, and forecasting the medium-long-term total power directly by taking the total power as a label.
The invention also comprises a verification module which is used for evaluating various indexes of Nash efficiency coefficient NSE, deviation coefficient Bias, root mean square error RMSE and mean absolute error MAE for point prediction;
for interval forecast, evaluating by adopting an improved coverage width comprehensive criterion CWC, an interval coverage PICP and an interval standardized average width PINAW index;
And (3) obtaining the total power of the water-wind-solar multi-energy complementary system by accumulating the water electricity, wind electricity and photoelectricity which are individually forecasted, and comparing the total power with the combined forecasting result of the water electricity, wind electricity and photoelectricity for a medium and long time obtained in the step (7).
The invention also provides a water-wind-light power combined forecasting device considering space-time complementarity, which comprises a processor and a memory for storing a computer program capable of running on the processor, wherein the processor is used for executing the steps of the water-wind-light power combined forecasting method considering space-time complementarity when running the computer program.
The memory in the embodiment of the invention is used for storing various types of data so as to support the operation of the water-wind-solar power combined forecasting equipment considering space-time complementarity. Examples of such data include: any computer program for operating on a water wind power joint prediction apparatus taking into account space-time complementarity.
The water-wind-light power combined forecasting method considering space-time complementarity disclosed by the embodiment of the invention can be applied to a processor or realized by the processor. The processor may be an integrated circuit chip having signal processing capabilities. In the implementation process, each step of the water, wind and light power combined forecasting method considering space-time complementarity can be completed through an integrated logic circuit of hardware in a processor or instructions in a software form. The processor may be a general purpose processor, a digital signal processor (DSP, digital SignalProcessor), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like. The processor may implement or perform the methods, steps, and logic blocks disclosed in embodiments of the present invention. The general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the method disclosed in the embodiment of the invention can be directly embodied in the hardware of the decoding processor or can be implemented by combining hardware and software modules in the decoding processor. The software module can be located in a storage medium, the storage medium is located in a memory, the processor reads information in the memory, and the steps of the water, wind and light power combined forecasting method considering space-time complementarity provided by the embodiment of the invention are completed by combining hardware of the software module.
In an exemplary embodiment, the water wind and light power joint prediction apparatus considering space-time complementarity may be implemented by one or more application specific integrated circuits (ASIC, application Specific Integrated Circuit), DSPs, programmable logic devices (PLD, programmable Logic Device), complex programmable logic devices (CPLD, complex Programmable LogicDevice), FPGAs, general purpose processors, controllers, microcontrollers (MCU, micro Controller Unit), microprocessors (Microprocessor), or other electronic components for performing the aforementioned methods.
It will be appreciated that the memory can be either volatile memory or nonvolatile memory, and can include both volatile and nonvolatile memory. Wherein the nonvolatile Memory may be Read Only Memory (ROM), programmable Read Only Memory (PROM, programmable Read-Only Memory), erasable programmable Read Only Memory (EPROM, erasable Programmable Read-Only Memory), electrically erasable programmable Read Only Memory (EEPROM, electrically Erasable Programmable Read-Only Memory), magnetic random access Memory (FRAM, ferromagnetic random access Memory), flash Memory (Flash Memory), magnetic surface Memory, optical disk, or compact disk Read Only Memory (CD-ROM, compact Disc Read-Only Memory); the magnetic surface memory may be a disk memory or a tape memory. The volatile memory may be random access memory (RAM, random AccessMemory), which acts as external cache memory. By way of example, and not limitation, many forms of RAM are available, such as static random access memory (SRAM, static Random Access Memory), synchronous static random access memory (SSRAM, synchronous Static Random Access Memory), dynamic random access memory (DRAM, dynamic Random Access Memory), synchronous dynamic random access memory (SDRAM, synchronousDynamic Random Access Memory), double data rate synchronous dynamic random access memory (ddr sdram, double Data Rate Synchronous Dynamic Random Access Memory), enhanced synchronous dynamic random access memory (ESDRAM, enhanced Synchronous Dynamic Random Access Memory), synchronous link dynamic random access memory (SLDRAM, syncLink Dynamic Random Access Memory), direct memory bus random access memory (DRRAM, direct Rambus Random Access Memory). The memory described by embodiments of the present invention is intended to comprise, without being limited to, these and any other suitable types of memory.
A computer storage medium having a computer program stored therein, wherein the computer program, when executed by a processor, implements the method for joint prediction of water, wind and solar power taking into account space-time complementarity according to any one of the above.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (10)
1. The water-wind-light power combined forecasting method considering space-time complementarity is characterized by comprising the following steps of:
step 1, collecting and arranging data of a water, wind and light multi-energy complementary system, wherein the data comprises large-scale hydrological factors, total power time sequence data, hydropower power time sequence data, wind power time sequence data and photoelectric power time sequence data of the water, wind and light multi-energy complementary system;
step 2, taking hydroelectric power, wind power, photoelectric power or total power as a target power forecast object, and quantifying complementarity of the hydroelectric power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system;
Step 3, considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watershed of the water-wind-solar multi-energy complementary system, and screening the large-scale hydrological factors of the corresponding watershed;
step 4, considering the autocorrelation of hydroelectric power, wind power, photoelectric power or total power in time;
step 5, constructing a mapping relation of a power prediction model, wherein the mapping relation considers the complementarity of the water-electricity power, wind power and photoelectric power in the water-wind-solar multi-energy complementary system, and the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, wind power, photoelectric power or total power in time, so that space-time complementary factors related to the power of the water-wind-solar multi-energy complementary system are obtained;
step 6, constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
and 7, based on the point prediction model and the interval prediction model, carrying out independent prediction on the hydroelectric power, the wind power or the photoelectric power according to space-time complementary factors, and directly carrying out medium-long term prediction on the total power by taking the total power as a label.
2. The water-wind-light power joint prediction method considering space-time complementarity according to claim 1, wherein the method is characterized by:
In step 2, the complementarity of the hydroelectric power, the wind power and the photoelectric power is quantified based on the Copula function.
3. The water-wind-light power joint prediction method considering space-time complementarity according to claim 2, wherein the method is characterized by:
the step 2 specifically includes:
step 201, determining the edge distribution of hydroelectric power, wind power and photoelectric power based on nuclear density estimation, wherein the nuclear density estimation formula is as follows:
wherein x is i Sample points with X distributed in the same way; k (t) is a kernel function; h is window width; n is the number of sample points; the kernel function K (t) needs to satisfy in the real domain:
step 202, solving the two-to-two variable joint distribution based on a Copula function, wherein the calculation formula is as follows:
in the method, in the process of the invention,
is an n-dimensional joint distribution function; />
The edge distribution function is the water power, wind power and photoelectric power; c is a Copula function of the connection variable;
selecting Euclidean distance d 2 And (3) evaluating the fitting degree:
wherein C is n (u, v) is an empirical Copula function, I is an oscillometric function, when F n (x ic ) If u is less than or equal to u, I is 1, otherwise 0;
based on the established joint distribution function, a conditional distribution of X.ltoreq.x when the variable Y=y is known can be established, the variable X, Y is any two of the water electric power, the wind power and the photoelectric power, and the binary conditional probability distribution function is shown as follows:
Where f (X, Y) is a joint probability density function of the variables X and Y; f (x), f (y) are probability density functions of the variables X, Y, respectively; u, v are cumulative distribution functions of variables X and Y; c (u, v) is a probability density function of the Copula function;
step 203, based on the binary conditional probability distribution function, establishing a conditional expectation distribution function as shown in the following formula:
based on equation (7), the average value of variable X when given variable y=y can be found, and the expected value of the condition of another related variable corresponding to the known distribution value of hydroelectric power, wind power or photovoltaic power can be found, thereby quantifying the complementarity between two by two in the water-wind-solar multi-energy complementary system based on the expected value, and determining the power station power complementary to the hydroelectric power, wind power or photovoltaic power based on the expected value
4. The water-wind-light power joint prediction method considering space-time complementarity according to claim 1, wherein the method is characterized by:
in step 3, large scale hydrological factors of the corresponding watershed are screened out based on the maximum mutual information coefficient
The step 3 specifically comprises the following steps:
step 301, there is an ordered pair dataset D = { (x) mi ,y mi ) I=1, 2, …, n }, by dividing the X-axis into X m The Y axis is divided into Y m Parts, get a x m ×y m Calculating the mutual information value of each grid:
Wherein p (x) m ) And p (y) m ) As variable X m 、Y m Is a function of the edge probability density of (2); p (x) m ,y m ) The joint probability density function of two variables is the ratio of the number of points in the current grid to the total number of points of the data;
step 302, normalizing the maximum value row of the mutual information in the grid:
step 303, performing grid division of different schemes, and repeating step 301 and step 302, wherein the maximum normalized mutual information value in all schemes is the maximum mutual information coefficient:
wherein x is m y m <B is a constraint condition of total grid division, and B is set to be 0.6 th power of the total data.
5. The water-wind-light power joint prediction method considering space-time complementarity according to claim 1, wherein the method is characterized by:
in step 5, the space-time complementary factor comprises the ith power station power P of the t period, which is complementary to the target power forecast object i (t) jth large scale hydrokinetic factor LC of the tth period j (t) and historical time sequences N (t) of the target power forecast object in the t period, wherein the obtained mapping relation is as follows:
N(t)＝f(P 1 (t-1),...,P n1 (t-m),LC 1 (t-1),...,LC n2 (t-m),N(t-1),...,N(t-m))(11)
wherein f is a simulation function of a deep learning method; n is n 1 The number of the power stations is complementary to the hydroelectric power, the wind power or the photoelectric power; n is n 2 The number of the large-scale hydrological weather forecast factors is the number; m is a lag period of the space-time complementary factor, and the space-time complementary factor of the previous m period is used for forecasting the power of a target power forecasting object of the next period.
6. The water-wind-light power joint prediction method considering space-time complementarity according to claim 1, wherein the method is characterized by:
in step 6, an interval prediction model is constructed by adopting an upper and lower limit estimation method, the deep learning is trained towards the direction of better prediction interval performance by constructing a loss function, and considering that the prediction interval needs to meet the criterion of smaller interval width on the basis of covering more true values as much as possible, the adopted loss function is shown in the following formula:
Loss＝f 1 +f 2 (12)
f 2 ＝k 2 |u q (t)-l q (t)| (14)
wherein f 1 The distance between the median value and the actual value of the interval is expressed, the closer the median value is to the actual value of the interval, the more accurate the result of the forecast interval is, and meanwhile, the distance between the median value and the actual value of the interval is f 1 A penalty factor lambda is also arranged in q Punishment is carried out on the condition that the actual value is not in the forecast interval, so that the coverage rate of the interval is improved in the network training process; f (f) 2 The method is used for calculating the interval width, and the narrower the interval width is under the condition that the interval coverage rate is the same, the higher the quality of the corresponding prediction interval is; by adjusting k 1 、k 2 The proportionality coefficient can lead the network to purposefully deviate to a certain index for training, when k is 1 、k 2 When the proportion is proper, the penalty factor lambda is adjusted q The coverage rate of the actual value of the interval can be combined with the width; n (t) represents the actual power value of the t period, and the unit is MW; u (u) q (t)、l q (t) is a forecast result of the model in a t period, and the forecast result respectively represents a forecast upper boundary and a forecast lower boundary; lambda (lambda) q Is a penalty coefficient; k (k) 1 、k 2 Are all proportionality coefficients.
7. The water-wind-light power joint prediction method considering space-time complementarity according to claim 1, wherein the method is characterized by:
step 8, for point forecast, evaluating various indexes of Nash efficiency coefficient NSE, deviation coefficient Bias, root mean square error RMSE and mean absolute error MAE;
for the section forecast, an improved coverage width comprehensive criterion CWC, a section coverage rate PICP and a section standardized average width PINAW index are adopted for evaluation, wherein a specific calculation formula is as follows;
where N (t) represents the actual value of power, MW, of the t-th period; t is the total period number of samples;
mean value of power, MW;
beta and eta are penalty coefficients of CWC; VR is the difference between the maximum and minimum values of the forecast samples; μ is confidence level;
and (3) comparing the long-term result in the total power obtained in the step (7) with the accumulated result of the hydroelectric power, the wind power and the photoelectric power.
8. The water, wind and light power combined forecasting device considering space-time complementarity is characterized by comprising:
the data collection module is used for collecting and arranging data of the water, wind and light multi-energy complementary system, and comprises large-scale hydrological factors, total power time sequence data, hydropower power time sequence data, wind power time sequence data and photoelectric power time sequence data of the water, wind and light multi-energy complementary system;
The power complementary quantization module is used for quantizing the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-light multi-energy complementary system by taking the water-electricity power, the wind power, the photoelectric power or the total power as a target power forecast object;
the large-scale hydrological meteorological factor determining module is used for screening the large-scale hydrological meteorological factors of corresponding watersheds by considering the combined influence of the large-scale hydrological factors on the water-wind-solar conditions of the corresponding watersheds of the water-wind-solar multi-energy complementary system;
the self-correlation determination module is used for considering the self-correlation of hydroelectric power, wind power, photoelectric power or total power in time;
the space-time complementary factor determining module is used for constructing a mapping relation of a power prediction model, the mapping relation considers the complementarity of the water-electricity power, the wind power and the photoelectric power in the water-wind-solar multi-energy complementary system, the large-scale hydrological meteorological factors have combined influence on the water-wind condition of the corresponding river basin of the water-wind-solar multi-energy complementary system, and the autocorrelation of the water-electricity power, the wind power, the photoelectric power or the total power in time is used for acquiring the space-time complementary factor related to the power of the water-wind-solar multi-energy complementary system;
the deep learning module is used for constructing a point forecasting model and an interval forecasting model of the power based on the deep learning model;
The power forecasting module is used for independently forecasting the hydropower power, the wind power or the photoelectric power according to the space-time complementary factors based on the point forecasting model and the interval forecasting model, and directly forecasting the total power for a medium-long term by taking the total power as a label.
9. A combined water-wind-solar power forecasting device taking into account space-time complementarity, characterized by comprising a processor and a memory for storing a computer program capable of running on the processor, the processor being adapted to execute the steps of the combined water-wind-solar power forecasting method taking into account space-time complementarity according to any one of the preceding claims 1-7 when running the computer program.
10. A computer storage medium, characterized by: the computer storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the method for joint prediction of water, wind and light power taking into account space-time complementarity according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211740856.3A CN116050604A (en) | 2022-12-30 | 2022-12-30 | Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211740856.3A CN116050604A (en) | 2022-12-30 | 2022-12-30 | Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116050604A true CN116050604A (en) | 2023-05-02 |
Family
ID=86126873
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211740856.3A Pending CN116050604A (en) | 2022-12-30 | 2022-12-30 | Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116050604A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116581755A (en) * | 2023-07-12 | 2023-08-11 | 长江水利委员会水文局 | Power prediction method, device, equipment and storage medium |
-
2022
- 2022-12-30 CN CN202211740856.3A patent/CN116050604A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116581755A (en) * | 2023-07-12 | 2023-08-11 | 长江水利委员会水文局 | Power prediction method, device, equipment and storage medium |
| CN116581755B (en) * | 2023-07-12 | 2023-09-29 | 长江水利委员会水文局 | Power prediction method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dong et al. | Storage sizing with peak-shaving policy for wind farm based on cyclic Markov chain model | |
| Zhang et al. | Wind speed interval prediction based on lorenz disturbance distribution | |
| CN107194495B (en) | Photovoltaic power longitudinal prediction method based on historical data mining | |
| Fatemi et al. | Parametric methods for probabilistic forecasting of solar irradiance | |
| Ramakrishna et al. | A model for joint probabilistic forecast of solar photovoltaic power and outdoor temperature | |
| CN116050604A (en) | Water-wind-light power combined forecasting method, device and equipment considering space-time complementarity | |
| CN111275571A (en) | Resident load probability prediction deep learning method considering microclimate and user mode | |
| CN114819322B (en) | Forecasting method for flow of lake entering lake | |
| CN114266416A (en) | Photovoltaic power generation power short-term prediction method and device based on similar days and storage medium | |
| CN116050595A (en) | Attention mechanism and decomposition mechanism coupled runoff amount prediction method | |
| CN113159102A (en) | Multi-time-scale photovoltaic power prediction method and system | |
| CN113761019A (en) | Statistical post-processing correction method for meteorological hydrological ensemble prediction | |
| Souhe et al. | Forecasting of electrical energy consumption of households in a smart grid | |
| CN116581755B (en) | Power prediction method, device, equipment and storage medium | |
| WO2024093493A1 (en) | Joint random simulation method and apparatus for hydro-wind-photovoltaic resources in drainage basin | |
| CN111476411B (en) | Photovoltaic power station generating capacity calculation method and server | |
| CN110909916A (en) | Entropy weight method based wind power generation monthly electric quantity interval prediction method | |
| CN115860165A (en) | Neural network basin rainfall runoff forecasting method and system considering initial loss | |
| Prusty et al. | Comparison of photovoltaic generation uncertainty models for power system planning using regression framework | |
| El Genidy et al. | Algorithms of solar energy prediction combined with percentile root estimation of three-parameters distributions | |
| Yang et al. | A hybrid vmd-based ARIMA-LSTM model for day-ahead pv prediction and uncertainty analysis | |
| Cui et al. | Quantify and reduce flood forecast uncertainty by the CHUP-BMA method | |
| Sotiropoulos et al. | Modeling of electricity load for forward contract pricing | |
| De Felice | Hydropower information for power system modelling: the JRC-EFAS-Hydropower dataset | |
| Takilalte et al. | Prediction of direct normal irradiation using a new empirical sunshine duration-based model |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230809 Address after: 430072 Hubei Province, Wuhan city Wuchang District of Wuhan University Luojiashan Applicant after: WUHAN University Applicant after: HUANENG LANCANG RIVER HYDROPOWER Inc. Applicant after: POWERCHINA GUIYANG ENGINEERING CORPORATION LIMITED Address before: 430072 Hubei Province, Wuhan city Wuchang District of Wuhan University Luojiashan Applicant before: WUHAN University |