CN118094486A - Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system - Google Patents
Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system Download PDFInfo
- Publication number
- CN118094486A CN118094486A CN202410508595.5A CN202410508595A CN118094486A CN 118094486 A CN118094486 A CN 118094486A CN 202410508595 A CN202410508595 A CN 202410508595A CN 118094486 A CN118094486 A CN 118094486A
- Authority
- CN
- China
- Prior art keywords
- data
- photovoltaic power
- time
- model
- power generation
- 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
- 238000010248 power generation Methods 0.000 title claims abstract description 87
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 45
- 238000013528 artificial neural network Methods 0.000 claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 24
- 238000004422 calculation algorithm Methods 0.000 claims description 37
- 238000004590 computer program Methods 0.000 claims description 25
- 230000000694 effects Effects 0.000 claims description 21
- 230000008859 change Effects 0.000 claims description 19
- 230000002159 abnormal effect Effects 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 14
- 230000007774 longterm Effects 0.000 claims description 13
- 238000010606 normalization Methods 0.000 claims description 13
- 230000003749 cleanliness Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 11
- 238000004140 cleaning Methods 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000009825 accumulation Methods 0.000 claims description 4
- 239000000428 dust Substances 0.000 claims description 4
- 238000005286 illumination Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 5
- 238000010801 machine learning Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 238000009434 installation Methods 0.000 description 4
- ADTDNFFHPRZSOT-PVFUSPOPSA-N ram-330 Chemical compound C([C@H]1N(CC2)C)C3=CC=C(OC)C(OC)=C3[C@]32[C@@]1(O)CC[C@@H](OC(=O)OCC)C3 ADTDNFFHPRZSOT-PVFUSPOPSA-N 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000001174 ascending effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a photovoltaic power generation power prediction method and a system based on full scene time sequence decomposition, and relates to the technical field of photovoltaic power generation, wherein the method comprises the following steps: establishing a time sequence model of photovoltaic power generation power data decomposition; embedding an autoregressive neural network into the time sequence model, and constructing an n-order autoregressive model; acquiring historical data and weather data of a photovoltaic power station, taking the historical data and the weather data as training data of an n-order autoregressive model, and training the n-order autoregressive model to obtain a prediction model; and predicting the future generation power of the photovoltaic power station through a prediction model. The method and the system can improve the accuracy of photovoltaic power generation power prediction, and are beneficial to power grid dispatching and load distribution.
Description
Technical Field
The embodiment of the invention relates to the technical field of photovoltaic power generation, in particular to a photovoltaic power generation power prediction method and system based on full scene time sequence decomposition.
Background
Solar photovoltaic power generation is a clean novel energy source, and attention is paid in recent years, however, the photovoltaic power generation is easily influenced by the surrounding environment, and the output has strong fluctuation and intermittence. The prediction accuracy of the generated power is improved, so that the photovoltaic power station can be helped to make a production plan, the operation mode is reasonably arranged, and the income and efficiency are increased; and the method can be applied in cooperation with a power grid dispatching system, so that the safety and reliability of the power grid are improved, and the capacity of the power grid for absorbing photovoltaic power is improved.
The commonly used photovoltaic power generation output power prediction method comprises a physical method, a machine learning method and a time sequence method. The physical method does not depend on historical data, uses parameters such as geographic information, photovoltaic equipment parameters, solar irradiation intensity, environmental temperature and the like to construct a physical prediction model, uses few parameters, is simple, and has poor generalization capability. The machine learning method and the time series method model based on meteorological data and historical photovoltaic power generation data. The time sequence method aims at summing up trend rules of the photovoltaic power generation power from the historical data and carries out subsequent prediction; the machine learning method uses numerous weather parameter value modeling, and can effectively reduce the deviation of the prediction result. However, in general, a single model algorithm cannot meet the increasing precision requirement, and currently, a deep learning algorithm of a combination of a neural network, machine learning, time series and the like is mostly adopted.
The method for short-term prediction of the photovoltaic power generation power is various, the photovoltaic power data is subjected to time series decomposition, weather factors are comprehensively considered on the basis of the time series method, various algorithms are fused, main weather characteristics are selected, the KNN and the isolated forest algorithm are used for repairing missing values and abnormal values, the reliability of training data is improved, the AR autoregressive neural network is used for modeling the time series correlation, and therefore the predicted photovoltaic power station short-term power is more accurate.
The paper (Jin Shangzhu, xue Run, chongqing academy of science and technology) discloses a photovoltaic power prediction method, which introduces a Prophet algorithm into the field of photovoltaic power generation prediction, uses solar power plant time-power data as a basis, and fits historical time-power data through the Prophet model to predict future power generation trend, thereby obtaining better effect on short-term prediction.
However, the research results still have a larger gap from the actual conditions, mainly:
1. the actual data missing value and outlier handling problems are not considered.
2. The influence of weather factors on photovoltaic power generation is not considered. The influence of weather factors on photovoltaic power generation is greatest, the clear weather power generation amount is greatest, the storm weather power generation amount is smallest, prediction is carried out only by means of time regularity of photovoltaic historical data, and the situation of real data is difficult to be met.
3. Correlation between time series data is not considered. Modeling is not performed sufficiently considering correlation between time-series data, resulting in still further improvement in prediction accuracy.
Disclosure of Invention
In order to overcome the defects of the prior art, the embodiment of the invention aims to provide a photovoltaic power generation power prediction method and a system based on full scene time sequence decomposition, which can improve the accuracy of photovoltaic power generation power prediction and are beneficial to power grid dispatching and load distribution.
In order to solve the above problems, a first aspect of the embodiments of the present invention discloses a photovoltaic power generation power prediction method based on full scene time sequence decomposition, including:
establishing a time sequence model of photovoltaic power generation power data decomposition;
Embedding an autoregressive neural network into the time sequence model to construct an n-order autoregressive model;
Acquiring historical data and weather data of a photovoltaic power station, and training the n-order autoregressive model to obtain a prediction model as training data of the n-order autoregressive model;
And predicting the future power generation of the photovoltaic power station through the prediction model.
In a first aspect of the embodiment of the present invention, a timing model for decomposing photovoltaic power generation power data is established, including:
decomposing the photovoltaic power generation power data into a data item, a short-term change item, a long-term change item, a weather influence factor, a cleanliness influence factor, an artificial activity influence factor and an angle influence factor to obtain a time sequence model:
y(t)=d(t)+s(t)+l(t)+w(t)+c(t)+h(t)+a;
Wherein y (t) is photovoltaic power generation power at the t-th time, d (t) is a data item at the t-th time, s (t) is a short-term variation item at the t-th time, l (t) is a long-term variation item at the t-th time, w (t) is a weather influence factor at the t-th time, c (t) is a cleanliness influence factor at the t-th time, h (t) is an artificial activity influence factor at the t-th time, and a is an angle influence factor at the t-th time;
And: ；
Wherein, C (t) represents a preset upper limit power data value, phi is a data growth rate, d 0 is a constant term, and t0 is a time offset;
；
；
Wherein T1, T2, N 1, and N 2 represent a preset short time period base, long time period base, and short time period value and long time period value, respectively; a n and b n represent the cosine term and sine term coefficients, respectively, of the short-term variation term, and c n and d n represent the cosine term and sine term coefficients, respectively, of the long-term variation term;
w(t)=f(x1,x2,x3,x4)=R·exp(p1·x1+p2·x2+p3·x3+p4·x4);
Wherein x1, x2, x3, x4 respectively represent irradiation intensity, temperature, humidity and wind speed; p1, p2, p3, p4 respectively represent the influence weights of irradiation intensity, temperature, humidity, wind speed on photovoltaic power, p1+p2+p3+p4=1;
c(t)=C0/φ(t)[u(t-t0)-u(t-t0-T3)];
φ(t)=φ(t+T3);
c(t)=c(t+T3);
Wherein T3 is the time period for cleaning the photovoltaic panel, T 0 is the time offset, u (T) is a step function, phi (T) is the increasing function of dust accumulation in the period, and C 0 is the preset photovoltaic panel cleaning threshold;
h(t)~N(μ0,σ0 2);
Wherein μ 0 is the mean of the influence intensities and σ 0 2 is the variance;
a=S·L·T·cosθ;
Wherein S is the area of the photovoltaic module, L is the illumination intensity, T is the photoelectric conversion efficiency of the photovoltaic panel, and θ is the inclination angle of the photovoltaic panel.
In a first aspect of the embodiment of the present invention, embedding an autoregressive neural network into the time sequence model to construct an n-order autoregressive model includes:
constructing an n-order autoregressive neural network:
y(t)=θ0+θ1y(t-1)+θ2y(t-2)+......+θny(t-n)+et；
Wherein θ 0 is a constant term, θ 1~θn is a model parameter, e t represents white gaussian noise with a mean value of μ 1 =0 and a variance of σ 1 2, y (t) is photovoltaic power generated at the t-th time, and y (t-n) is photovoltaic power generated at the t-n-th time;
and embedding the n-order autoregressive neural network into the time sequence model to obtain an n-order autoregressive model.
In a first aspect of the embodiment of the present invention, the obtaining historical data and weather data of the photovoltaic power station as training data of the n-order autoregressive model, training the n-order autoregressive model to obtain a prediction model includes:
Historical data and weather data of the photovoltaic power station are extracted in time sequence and combined into time-power-weather data in DATAFRAME data format.
In a first aspect of the embodiment of the present invention, as a preferred embodiment, the method further includes:
Repairing the time-power-weather data, comprising:
normalizing the weather data;
detecting an abnormal value through an isolated forest algorithm based on the weather data after normalization processing, and setting the abnormal value as a missing value;
and repairing the missing value in the power data of the photovoltaic power station by using a KNN algorithm.
In a first aspect of the embodiment of the present invention, as a preferred embodiment, the method further includes:
And evaluating the prediction effect of the prediction model through a root mean square error (RMS) or/and an average absolute error (MAE) so as to continuously train the prediction model when the RMS or/and the average absolute error (MAE) is greater than a corresponding preset error threshold.
The second aspect of the embodiment of the invention discloses a photovoltaic power generation power prediction system based on full scene time sequence decomposition, which comprises the following components:
The building unit is used for building a time sequence model of the data decomposition of the photovoltaic power generation power;
the embedding unit is used for embedding the autoregressive neural network into the time sequence model to construct an n-order autoregressive model;
The acquisition unit is used for acquiring historical data and weather data of the photovoltaic power station, and training the n-order autoregressive model to obtain a prediction model as training data of the n-order autoregressive model;
And the prediction unit is used for predicting the future generation power of the photovoltaic power station through the prediction model.
In a second aspect of the embodiment of the present invention, the acquiring unit includes:
And the combining subunit is used for extracting historical data of the photovoltaic power station and weather data in time sequence and combining the historical data and the weather data into time-power-weather data in DATAFRAME data format.
The normalization subunit is used for carrying out normalization processing on the weather data;
the abnormal detection subunit is used for detecting abnormal values through an isolated forest algorithm based on the weather data after normalization processing, and setting the abnormal values as missing values;
and the repairing subunit is used for repairing the missing value in the power data of the photovoltaic power station by using a KNN algorithm.
A third aspect of the embodiment of the present invention discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the steps of the photovoltaic power generation power prediction method based on full scene time sequence decomposition disclosed in the first aspect of the embodiment of the present invention are implemented when the processor executes the computer program.
A fourth aspect of the embodiment of the present invention discloses a computer-readable storage medium storing a computer program, where the computer program causes a computer to execute the steps of the photovoltaic power generation power prediction method based on full-scene time-series decomposition disclosed in the first aspect of the embodiment of the present invention.
A fifth aspect of the embodiments of the present invention discloses a computer program product, which when run on a computer causes the computer to execute the steps of the photovoltaic power generation power prediction method based on full scene time sequence decomposition disclosed in the first aspect of the embodiments of the present invention.
A sixth aspect of the embodiment of the present invention discloses an application publishing platform, which is configured to publish a computer program product, where when the computer program product runs on a computer, the computer is caused to execute the steps of the photovoltaic power generation power prediction method based on full scene time sequence decomposition disclosed in the first aspect of the embodiment of the present invention.
Compared with the prior art, the embodiment of the invention has the beneficial effects that:
According to the embodiment of the invention, the photovoltaic power generation power is predicted based on a time sequence decomposition algorithm, and a time sequence method of data decomposition is taken as a main body to decompose a power value; and the weather features are used as variables affecting the power values, the main weather features are screened out, the missing data are repaired by using an isolated forest and KNN algorithm, the AR-Net is embedded into the algorithm, and the accuracy of predicting the photovoltaic power by using the combined algorithm is higher than that by using a single time sequence, machine learning and neural network prediction algorithm, so that the method is beneficial to power grid dispatching and load distribution.
Drawings
Fig. 1 is a schematic flow chart of a photovoltaic power generation power prediction method based on full scene time sequence decomposition according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of time-power-weather data provided in accordance with an embodiment of the present invention;
Fig. 3 is a schematic structural diagram of a photovoltaic power generation power prediction system based on full scene time sequence decomposition according to the second embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention.
Detailed Description
This detailed description is merely illustrative of the embodiments of the invention and is not intended to limit the embodiments of the invention, since modifications of the embodiments can be made by those skilled in the art without creative contribution as required after reading the specification, but are protected by the patent laws within the scope of the claims of the embodiments of the invention.
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of 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, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the embodiments of the present invention.
The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
In embodiments of the invention, words such as "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "e.g." in an embodiment should not be taken as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.
According to the embodiment of the invention, the photovoltaic power generation power is predicted based on a time sequence decomposition algorithm, and a time sequence method of data decomposition is taken as a main body to decompose a power value; and the weather features are used as variables affecting the power values, the main weather features are screened out, the missing data are repaired by using an isolated forest and KNN algorithm, the AR-Net is embedded into the algorithm, and the accuracy of predicting the photovoltaic power by using the combined algorithm is higher than that of a single time sequence, machine learning and neural network prediction algorithm, so that the method is beneficial to power grid dispatching and load distribution, and the method is described in detail below with reference to the accompanying drawings.
Example 1
Referring to fig. 1, fig. 1 is a flow chart of a photovoltaic power generation power prediction method based on full scene time sequence decomposition according to an embodiment of the present invention. As shown in fig. 1, the photovoltaic power generation power prediction method based on full scene time sequence decomposition includes:
S110, establishing a time sequence model of the data decomposition of the photovoltaic power generation power.
The photovoltaic power generation is regular, and the photovoltaic power data is decomposed into a data item, a short-term change item, a long-term change item, a weather influence factor, a cleanliness influence factor, an artificial activity influence factor and an angle influence factor, so that a time sequence model for decomposing the photovoltaic power generation power data is established:
y(t)=d(t)+s(t)+l(t)+w(t)+c(t)+h(t)+a；
Where y (t) is the photovoltaic power generation power of the tth (the time length of t may be set as needed, for example, 1 day may be used as a unit, or an hour may be used as a unit), d (t) is a data item of the tth, s (t) is a short-term change item of the tth, l (t) is a long-term change item of the tth, w (t) is a weather influence factor of the tth, c (t) is a cleanliness influence factor of the tth, h (t) is an artificial activity influence factor of the tth, and a is an angle influence factor of the tth.
(1) Data item: by data item, we mean fitting analysis of the non-periodic trend of variation of photovoltaic power. For photovoltaic power, the lower limit power data value is set to 0, and the upper limit power data value is observed according to historical empirical data.
The expression is as follows:
；
the preset upper limit power data value is represented, the data increment rate is a constant term, and the time offset is represented.
(2) Short term variation term: the short-term change term means that the photovoltaic power has obvious daily periodicity in real-time minute, daytime, night and day data statistics units, the power is in an ascending trend from the morning to the noon, the power is in a descending trend from the noon to the evening, and the change of the photovoltaic power generation power is simulated by the Fourier series.
The expression is as follows:
t1 and N 1 represent a preset short-time period base and a short-time period value, respectively, and when t1=24, for example, T 1 =3 can be set.
(3) Long-term change term: the term of long-term change refers to a change in which photovoltaic power generation power is changed in four seasons in data statistics units such as week, month, year, etc., the amount of power generated in summer is high, the amount of power generated in winter is low, and the change in photovoltaic power generation power is simulated by fourier series.
The expression is as follows:
T2 and N2 represent a preset long period base and a long period value, respectively, and for example, t2=365 days, N 2 =10 may be set.
(4) Weather influencing factors: the influence of the external weather variable on the photovoltaic power generation power is represented. The acquired weather data of the weather station generally comprise irradiation intensity, temperature, humidity, wind direction, wind speed, pressure intensity, cloud cover, rainfall, weather type and the like, and in order to avoid influence of irrelevant features on model training and prediction, the irradiation intensity, temperature, humidity and wind speed are selected in the embodiment of the invention, and data normalization is carried out to eliminate dimensional differences among different weather features.
The expression is as follows:
w(t)=f(x1,x2,x3,x4)=R·exp(p1·x1+p2·x2+p3·x3+p4·x4)。
wherein x1, x2, x3, x4 respectively represent irradiation intensity, temperature, humidity and wind speed; p1, p2, p3, p4 represent the influence weights of irradiation intensity, temperature, humidity, wind speed on photovoltaic power, p1+p2+p3+p4=1, respectively.
The weather factor items include:
a. Irradiation intensity factor x1: the irradiation intensity directly influences the output power of the photovoltaic cell, and the larger the daily irradiation intensity is, the larger the generated energy is.
B. Temperature factor x2: the temperature rise may result in reduced conversion efficiency, reduced lifetime, and reduced reliability of the photovoltaic module.
C. Humidity factor x3: humidity mainly affects the heat dissipation effect and the atmospheric transparency of the photovoltaic system.
D. Wind speed factor x4: the moderate wind speed can reduce the temperature of the photovoltaic cell panel and improve the efficiency of the photovoltaic system.
(5) Cleanliness factor c (t): the method has the advantages that the influence of the cleanliness of the photovoltaic panel on the photovoltaic power generation power is shown, the power generation efficiency of the photovoltaic module is reduced due to the low cleanliness of the photovoltaic panel, the temperature of the photovoltaic module is increased, the power generation efficiency is reduced, and the service life of the photovoltaic module is influenced.
The expression is as follows:
c(t)=C0/φ(t)[u(t-t0)-u(t-t0-T3)];
φ(t)=φ(t+T3);
c(t)=c(t+T3);
Wherein T3 is the time period for cleaning the photovoltaic panel, T 0 is the time offset, u (T) is a step function, phi (T) is the dust accumulation amount increasing function in the period, and phi (T) and c (T) are approximate periodic functions.
(6) Human activity impact factor h (t): other special human activities are shown to have randomness in influencing the photovoltaic power generation power, such as overhauling, cleaning a photovoltaic panel and the like, and the expression is as follows under the assumption of Gaussian distribution:
h(t)~N(μ0,σ0 2)；
Wherein μ 0 is the mean of the influence intensities and σ 0 2 is the variance, i.e. h (t) represents the influence of different time artificial activities on photovoltaic power generation.
(7) Angle influence factor a: the influence of the sun angle on the photovoltaic power generation power is represented, the installation inclination angle of the photovoltaic panel can influence the power generation efficiency, and the expression is as follows:
a=S·L·T·cosθ；
Wherein S is the area of the photovoltaic module, L is the illumination intensity, T is the photoelectric conversion efficiency of the photovoltaic panel, and θ is the inclination angle of the photovoltaic panel.
Each region has an optimum inclination angle for assembly installation, which is related to the latitude of the installation site, and when the inclination angle of assembly installation changes, the amount of electricity generated changes accordingly. The photovoltaic power generation efficiency is highest under the conditions of normal south orientation and optimal inclination angle.
The time sequence method of data decomposition is taken as a main body, and the power value is decomposed into a data item, a short-term change item, a long-term change item, a weather influence factor, a cleanliness influence factor, an artificial activity influence factor and an angle influence factor, so that the power value can be more matched with a photovoltaic power prediction task.
S120, embedding the autoregressive neural network into the time sequence model to construct an n-order autoregressive model.
The photovoltaic power generation power has obvious change trend along with time, the time sequence prediction method is taken as a main body, the photovoltaic power data is decomposed in time sequence, and the machine learning algorithm and the neural network algorithm are fused in the time sequence method to predict the photovoltaic power, so that the problem of insufficient prediction precision of a single algorithm is effectively avoided, the accuracy of predicting the photovoltaic power is higher than that of a single algorithm, and the method is favorable for power grid dispatching and load distribution.
In a preferred embodiment of the present invention, taking AR-Net as an example, AR-Net (autoregressive neural network) is embedded in a time sequence model, and a combination algorithm (a combination model of AR-Net and time sequence, for example, can be used to substitute the time sequence into the autoregressive neural network to obtain a combination model, which is denoted as an n-order autoregressive model).
The autoregressive neural network is embedded into an algorithm, an n-order autoregressive model shows that the power expected value at the current moment not only depends on the current-period values of other variables, but also depends on the current hysteresis values of the previous n moments, and the expression is as follows:
y(t)=θ0+θ1y(t-1)+θ2y(t-2)+......+θny(t-n)+et；
Wherein θ 0 is a constant term, θ 1~θn is a model parameter, e t is gaussian white noise with a mean value of μ 1 =0 and a variance of σ 1 2, and n is an order, which can be adjusted according to an actual effect, for example, when t represents 1 day, n can be 7, that is, an autoregressive neural network is modeled with data lagging behind for 7 days, and is embedded into a time sequence model.
And S130, acquiring historical data and weather data of the photovoltaic power station, and training the n-order autoregressive model to obtain a prediction model as training data of the n-order autoregressive model.
The historical power data of the photovoltaic power station and the high-precision weather data of the weather station are collected hour by hour and are used as training data of a model, the historical data and the weather data of the photovoltaic power station are arranged in time sequence, and time-power-weather data in DATAFRAME data format shown in figure 2 are combined.
Because the photovoltaic power station stores and transmits data through a network in the operation process, the data is easy to lose or the data transmission is incomplete, and therefore, larger errors exist when the model is directly trained by the original data.
In the embodiment of the invention, the correlation between time sequences is considered, the AR-Net is adopted for modeling, and the missing values are filled. On the basis, the embodiment of the invention adopts a KNN multiple regression method based on time sequence to model because the data generated by the photovoltaic power station in the operation process has volatility.
Specifically, firstly, the data are ordered according to ascending sequence, the time is detected, the time points with the defects are sampled and supplemented, the time interval between the data is ensured, and the power value of the time points with the defects is a missing value; the weather features (namely weather data) are subjected to minimum-maximum normalization, so that dimension differences among different features are eliminated, comparability among the different features is achieved, a calculation formula is X '= (X-X min)/(Xmax-Xmin), X' is normalized data, and X max and X min are maximum values and minimum values in original data.
Then, based on the normalized weather characteristics, firstly detecting an abnormal value by using an isolated forest algorithm, and setting the abnormal value as a missing value;
And finally, establishing a KNN multiple regression model based on the screened weather features (namely, the weather features related to the missing values, which comprise the weather features corresponding to the missing values and the related weather features of the weather features corresponding to the missing values), wherein the KNN regression method can comprehensively complement missing data by using the correlation coefficient after the missing values are detected.
The use of the isolated forest algorithm and the KNN algorithm to repair abnormal values and missing values in actual data based on weather features plays an important role in improving prediction accuracy.
The processed time-weather-power data is used as training data to be input into a model for training, and errors are gradually reduced through multiple training rounds, so that each parameter of the n-order autoregressive model is suitable, and a prediction model is obtained.
For the data of the prediction model predicted to be non-zero value at night, the data is modified to be zero value so as to meet the actual situation, and the error can be reduced.
Then, the prediction result of the prediction model is verified, which may specifically be: obtaining weather variables known in the future for a plurality of days, predicting the power generated in the future for a plurality of days by using a trained prediction model, evaluating model effects by using MAE (maximum energy analysis) or/and RMSE (maximum energy analysis) and using the fitting degree of a prediction curve and a real curve as indexes, and evaluating the prediction effects of the prediction model:
；
Where n is the number of training data, y actual is the actual power value, and y pred is the model predictive value.
And if the Root Mean Square Error (RMSE) or/and the average absolute error (MAE) is greater than the corresponding preset error threshold, continuing to train the prediction model until the Root Mean Square Error (RMSE) or/and the average absolute error (MAE) is less than or equal to the corresponding preset error threshold.
The validity of the present prediction model may be verified by comparing the prediction model with an existing single algorithm, for example, a single algorithm such as time series, machine learning, or neural network prediction, with indexes such as Root Mean Square Error (RMSE), mean Absolute Error (MAE), and curve fitting degree.
And S140, predicting the future power generation of the photovoltaic power station through the prediction model.
When the prediction effect of the prediction model reaches the preset requirement, the future power generation power can be predicted based on the historical data, for example, weather variables known in the future for 1 day are obtained, and the trained model is used for predicting the future power generation power for 1 day.
Example two
Referring to fig. 3, fig. 3 is a schematic structural diagram of a photovoltaic power generation power prediction system based on full scene time sequence decomposition according to an embodiment of the present invention. As shown in fig. 3, the full scene time sequence decomposition-based photovoltaic power generation power prediction system may include:
A building unit 210, configured to build a timing model of the photovoltaic power generation power data decomposition;
an embedding unit 220, configured to embed an autoregressive neural network into the time sequence model, and construct an n-order autoregressive model;
An obtaining unit 230, configured to obtain historical data and weather data of the photovoltaic power plant, and train the n-order autoregressive model as training data of the n-order autoregressive model to obtain a prediction model;
And the prediction unit 240 is used for predicting the future generation power of the photovoltaic power station through the prediction model.
Preferably, the establishing unit 210 may include:
decomposing the photovoltaic power generation power data into a data item, a short-term change item, a long-term change item, a weather influence factor, a cleanliness influence factor, an artificial activity influence factor and an angle influence factor to obtain a time sequence model:
y(t)=d(t)+s(t)+l(t)+w(t)+c(t)+h(t)+a；
Wherein y (t) is photovoltaic power generation power at the t-th time, d (t) is a data item at the t-th time, s (t) is a short-term variation item at the t-th time, l (t) is a long-term variation item at the t-th time, w (t) is a weather influence factor at the t-th time, c (t) is a cleanliness influence factor at the t-th time, h (t) is an artificial activity influence factor at the t-th time, and a is an angle influence factor at the t-th time;
And: ；
Wherein, C (t) represents a preset upper limit power data value, phi is a data growth rate, d 0 is a constant term, and t0 is a time offset;
；
；
Wherein T1, T2, N 1, and N 2 represent a preset short time period base, long time period base, and short time period value and long time period value, respectively;
w(t)=f(x1,x2,x3,x4)=R·exp(p1·x1+p2·x2+p3·x3+p4·x4)；
Wherein x1, x2, x3, x4 respectively represent irradiation intensity, temperature, humidity and wind speed; p1, p2, p3, p4 respectively represent the influence weights of irradiation intensity, temperature, humidity, wind speed on photovoltaic power, p1+p2+p3+p4=1;
c(t)=C0/φ(t)[u(t-t0)-u(t-t0-T3)];
φ(t)=φ(t+T3);
c(t)=c(t+T3)；
Wherein T3 is the time period for cleaning the photovoltaic panel, T 0 is the time offset, u (T) is a step function, phi (T) is the increasing function of dust accumulation in the period, and C 0 is the preset photovoltaic panel cleaning threshold;
h(t)~N(μ0,σ0 2)；
Wherein μ 0 is the mean of the influence intensities and σ 0 2 is the variance;
a=S·L·T·cosθ；
Wherein S is the area of the photovoltaic module, L is the illumination intensity, T is the photoelectric conversion efficiency of the photovoltaic panel, and θ is the inclination angle of the photovoltaic panel.
Preferably, the embedding unit 220 may include:
constructing an n-order autoregressive neural network:
y(t)=θ0+θ1y(t-1)+θ2y(t-2)+......+θny(t-n)+et；
Wherein, θ 0 is a constant term, θ 1~θn is a model parameter, e t represents Gaussian white noise with a mean value of mu 1=0、σ1 2 and a variance, y (t) is photovoltaic power generation power at the t-th time, and y (t-n) is photovoltaic power generation power at the t-n-th time;
and embedding the n-order autoregressive neural network into the time sequence model to obtain an n-order autoregressive model.
Preferably, the acquisition unit 230 includes:
And the combining subunit is used for extracting historical data of the photovoltaic power station and weather data in time sequence and combining the historical data and the weather data into time-power-weather data in DATAFRAME data format.
The normalization subunit is used for carrying out normalization processing on the weather data;
the abnormal detection subunit is used for detecting abnormal values through an isolated forest algorithm based on the weather data after normalization processing, and setting the abnormal values as missing values;
and the repairing subunit is used for repairing the missing value in the power data of the photovoltaic power station by using a KNN algorithm.
Preferably, the system further comprises: and the evaluation unit is used for evaluating the prediction effect of the prediction model through root mean square error (RMS) or/and Mean Absolute Error (MAE) so as to continuously train the prediction model when the RMS error or/and the Mean Absolute Error (MAE) is greater than a corresponding preset error threshold.
Example III
Referring to fig. 4, fig. 4 is a schematic diagram of an electronic device that may be used to implement an embodiment of the present invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the embodiments of the invention described and/or claimed herein.
As shown in fig. 4, the electronic device includes at least one processor 310, and a memory, such as a ROM (read only memory) 320, a RAM (random access memory) 330, etc., communicatively connected to the at least one processor 310, wherein the memory stores computer programs executable by the at least one processor, and the processor 310 can perform various suitable actions and processes according to the computer programs stored in the ROM 320 or the computer programs loaded from the storage unit 380 into the random access memory RAM 330. In the RAM 330, various programs and data required for the operation of the electronic device may also be stored. The processor 310, ROM 320, and RAM 330 are connected to each other by a bus 340. An I/O (input/output) interface 350 is also connected to bus 340.
A number of components in the electronic device are connected to the I/O interface 350, including: an input unit 360 such as a keyboard, a mouse, etc.; an output unit 370 such as various types of displays, speakers, and the like; a storage unit 380 such as a magnetic disk, an optical disk, or the like; and a communication unit 390, such as a network card, modem, wireless communication transceiver, etc. The communication unit 390 allows the electronic device to exchange information/data with other devices via a computer network such as the internet or/and various telecommunications networks.
Processor 310 may be a variety of general-purpose or/and special-purpose processing components having processing and computing capabilities. Some examples of processor 310 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 310 performs one or more steps of a full scene time series decomposition based photovoltaic power generation power prediction method described in embodiment one above.
In some embodiments, a photovoltaic power generation power prediction method based on full scene time series decomposition may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as the storage unit 380. In some embodiments, part or all of the computer program may be loaded onto or/and installed onto the electronic device via ROM 320 or/and communication unit 390. When the computer program is loaded into RAM 330 and executed by processor 310, one or more steps of a photovoltaic power generation power prediction method based on full scene timing decomposition as described in embodiment one above may be performed. Alternatively, in other embodiments, the processor 310 may be configured to perform a full scene time-sequential decomposition-based photovoltaic power generation power prediction method in any other suitable manner (e.g., by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, or/and combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed or/and interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
A computer program for implementing the methods of embodiments of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of embodiments of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
The photovoltaic power generation power prediction method, the device, the electronic equipment and the storage medium based on full scene time sequence decomposition disclosed by the invention are described in detail, and specific examples are applied to explain the principle and the implementation mode of the invention, and the description of the above examples is only used for helping to understand the method and the core idea of the invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.
Claims (10)
1. The full scene time sequence decomposition-based photovoltaic power generation power prediction method is characterized by comprising the following steps of:
establishing a time sequence model of photovoltaic power generation power data decomposition;
Embedding an autoregressive neural network into the time sequence model to construct an n-order autoregressive model;
Acquiring historical data and weather data of a photovoltaic power station, and training the n-order autoregressive model to obtain a prediction model as training data of the n-order autoregressive model;
And predicting the future power generation of the photovoltaic power station through the prediction model.
2. The full scene time-series decomposition-based photovoltaic power generation power prediction method according to claim 1, wherein establishing a time-series model of the photovoltaic power generation power data decomposition comprises:
decomposing the photovoltaic power generation power data into a data item, a short-term change item, a long-term change item, a weather influence factor, a cleanliness influence factor, an artificial activity influence factor and an angle influence factor to obtain a time sequence model:
y(t)=d(t)+s(t)+l(t)+w(t)+c(t)+h(t)+a;
Wherein y (t) is photovoltaic power generation power at the t-th time, d (t) is a data item at the t-th time, s (t) is a short-term variation item at the t-th time, l (t) is a long-term variation item at the t-th time, w (t) is a weather influence factor at the t-th time, c (t) is a cleanliness influence factor at the t-th time, h (t) is an artificial activity influence factor at the t-th time, and a is an angle influence factor at the t-th time;
And: ；
Wherein, C (t) represents a preset upper limit power data value, phi is a data growth rate, d 0 is a constant term, and t0 is a time offset;
;
;
Wherein T1, T2, N 1, and N 2 represent a preset short time period base, long time period base, and short time period value and long time period value, respectively; a n and b n represent the cosine term and sine term coefficients, respectively, of the short-term variation term, and c n and d n represent the cosine term and sine term coefficients, respectively, of the long-term variation term;
w(t)=f(x1,x2,x3,x4)=R·exp(p1·x1+p2·x2+p3·x3+p4·x4);
Wherein x1, x2, x3, x4 respectively represent irradiation intensity, temperature, humidity and wind speed; p1, p2, p3, p4 respectively represent the influence weights of irradiation intensity, temperature, humidity, wind speed on photovoltaic power, p1+p2+p3+p4=1;
c(t)=C0/φ(t)[u(t-t0)-u(t-t0-T3)];
φ(t)=φ(t+T3);
c(t)=c(t+T3);
Wherein T3 is the time period for cleaning the photovoltaic panel, T 0 is the time offset, u (T) is a step function, phi (T) is the increasing function of dust accumulation in the period, and C 0 is the preset photovoltaic panel cleaning threshold;
h(t)~N(μ0,σ0 2);
Wherein μ 0 is the mean of the influence intensities and σ 0 2 is the variance;
a=S·L·T·cosθ;
Wherein S is the area of the photovoltaic module, L is the illumination intensity, T is the photoelectric conversion efficiency of the photovoltaic panel, and θ is the inclination angle of the photovoltaic panel.
3. The full scene time sequence decomposition based photovoltaic power generation prediction method according to claim 1, wherein embedding an autoregressive neural network into the time sequence model to construct an n-order autoregressive model comprises:
constructing an n-order autoregressive neural network:
y(t)=θ0+θ1y(t-1)+θ2y(t-2)+......+θny(t-n)+et；
Wherein θ 0 is a constant term, θ 1~θn is a model parameter, e t represents white gaussian noise with a mean value of μ 1 =0 and a variance of σ 1 2, y (t) is photovoltaic power generated at the t-th time, and y (t-n) is photovoltaic power generated at the t-n-th time;
and embedding the n-order autoregressive neural network into the time sequence model to obtain an n-order autoregressive model.
4. The full scene time-series decomposition-based photovoltaic power generation power prediction method according to claim 1, wherein obtaining historical data and weather data of a photovoltaic power plant as training data of the n-order autoregressive model, training the n-order autoregressive model to obtain a prediction model, comprises:
Historical data and weather data of the photovoltaic power station are extracted in time sequence and combined into time-power-weather data in DATAFRAME data format.
5. The full scene time-series decomposition based photovoltaic power generation power prediction method according to claim 4, further comprising:
Repairing the time-power-weather data, comprising:
normalizing the weather data;
detecting an abnormal value through an isolated forest algorithm based on the weather data after normalization processing, and setting the abnormal value as a missing value;
and repairing the missing value in the power data of the photovoltaic power station by using a KNN algorithm.
6. The full scene time based power prediction method for photovoltaic generation of any of claims 1 to 5, further comprising:
And evaluating the prediction effect of the prediction model through the root mean square error or/and the average absolute error, so as to continuously train the prediction model when the root mean square error or/and the average absolute error is greater than a corresponding preset error threshold.
7. The utility model provides a photovoltaic power generation power prediction system based on full scene time sequence decomposition which characterized in that it includes:
The building unit is used for building a time sequence model of the data decomposition of the photovoltaic power generation power;
the embedding unit is used for embedding the autoregressive neural network into the time sequence model to construct an n-order autoregressive model;
The acquisition unit is used for acquiring historical data and weather data of the photovoltaic power station, and training the n-order autoregressive model to obtain a prediction model as training data of the n-order autoregressive model;
And the prediction unit is used for predicting the future generation power of the photovoltaic power station through the prediction model.
8. The full scene time-series decomposition based photovoltaic power generation prediction system according to claim 7, wherein said acquisition unit comprises:
a combining subunit, configured to extract historical data and weather data of the photovoltaic power station in time sequence, and combine the historical data and the weather data into time-power-weather data in DATAFRAME data format;
the normalization subunit is used for carrying out normalization processing on the weather data;
the abnormal detection subunit is used for detecting abnormal values through an isolated forest algorithm based on the weather data after normalization processing, and setting the abnormal values as missing values;
and the repairing subunit is used for repairing the missing value in the power data of the photovoltaic power station by using a KNN algorithm.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the full scene time based decomposition photovoltaic power generation power prediction method according to any of claims 1-6 when the computer program is executed.
10. A computer-readable storage medium, characterized in that it stores a computer program, wherein the computer program causes a computer to execute the steps of the full scene time-series decomposition-based photovoltaic power generation power prediction method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410508595.5A CN118094486B (en) | 2024-04-26 | Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410508595.5A CN118094486B (en) | 2024-04-26 | Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118094486A true CN118094486A (en) | 2024-05-28 |
| CN118094486B CN118094486B (en) | 2024-07-05 |
Family
ID=
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013246746A (en) * | 2012-05-29 | 2013-12-09 | Hitachi Ltd | Photovoltaic power generation amount estimation system, estimation method and photovoltaic power generation system control system |
| US20180034312A1 (en) * | 2016-07-28 | 2018-02-01 | International Business Machines Corporation | Optimal Distributed Energy Resource Management System |
| CN113298297A (en) * | 2021-05-10 | 2021-08-24 | 内蒙古工业大学 | Wind power output power prediction method based on isolated forest and WGAN network |
| CN114329985A (en) * | 2021-12-30 | 2022-04-12 | 山东浪潮科学研究院有限公司 | Photovoltaic power generation prediction method, device, equipment and storage medium |
| US20220373984A1 (en) * | 2021-05-19 | 2022-11-24 | Shandong University | Hybrid photovoltaic power prediction method and system based on multi-source data fusion |
| CN115564126A (en) * | 2022-10-18 | 2023-01-03 | 珠海许继芝电网自动化有限公司 | Photovoltaic power generation prediction method, system, device and storage medium |
| CN116050627A (en) * | 2023-01-30 | 2023-05-02 | 固德威技术股份有限公司 | Photovoltaic power prediction method and photovoltaic power prediction model training method |
| CN116565864A (en) * | 2023-07-11 | 2023-08-08 | 上海融和元储能源有限公司 | Photovoltaic power generation power forecasting method based on PCA-RBF algorithm |
| CN117846894A (en) * | 2024-01-11 | 2024-04-09 | 上海电气风电集团股份有限公司 | Data processing method, early warning device, equipment and medium of wind turbine generator |
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013246746A (en) * | 2012-05-29 | 2013-12-09 | Hitachi Ltd | Photovoltaic power generation amount estimation system, estimation method and photovoltaic power generation system control system |
| US20180034312A1 (en) * | 2016-07-28 | 2018-02-01 | International Business Machines Corporation | Optimal Distributed Energy Resource Management System |
| CN113298297A (en) * | 2021-05-10 | 2021-08-24 | 内蒙古工业大学 | Wind power output power prediction method based on isolated forest and WGAN network |
| US20220373984A1 (en) * | 2021-05-19 | 2022-11-24 | Shandong University | Hybrid photovoltaic power prediction method and system based on multi-source data fusion |
| CN114329985A (en) * | 2021-12-30 | 2022-04-12 | 山东浪潮科学研究院有限公司 | Photovoltaic power generation prediction method, device, equipment and storage medium |
| CN115564126A (en) * | 2022-10-18 | 2023-01-03 | 珠海许继芝电网自动化有限公司 | Photovoltaic power generation prediction method, system, device and storage medium |
| CN116050627A (en) * | 2023-01-30 | 2023-05-02 | 固德威技术股份有限公司 | Photovoltaic power prediction method and photovoltaic power prediction model training method |
| CN116565864A (en) * | 2023-07-11 | 2023-08-08 | 上海融和元储能源有限公司 | Photovoltaic power generation power forecasting method based on PCA-RBF algorithm |
| CN117846894A (en) * | 2024-01-11 | 2024-04-09 | 上海电气风电集团股份有限公司 | Data processing method, early warning device, equipment and medium of wind turbine generator |
Non-Patent Citations (2)
| Title |
|---|
| YURUI XIA等: "Short-term PV power forecasting based on time series expansion and high-order fuzzy cognitive maps", APPLIED SOFT COMPUTING, 18 January 2023 (2023-01-18), pages 1 - 15 * |
| 朱想等: "基于改进NARX神经网络算法的光伏发电功率短期预测", 武汉大学学报(理学版), vol. 66, no. 5, 30 September 2020 (2020-09-30), pages 505 - 510 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102930358B (en) | A kind of neural net prediction method of photovoltaic power station power generation power | |
| WO2021043180A1 (en) | Regional short-term energy power prediction method and system that take weather into consideration | |
| Sun et al. | Research on ultra-short-term wind power prediction considering source relevance | |
| CN106803128B (en) | Photovoltaic power generation prediction method and device | |
| CN108053048A (en) | A kind of gradual photovoltaic plant ultra-short term power forecasting method of single step and system | |
| US9857778B1 (en) | Forecasting solar power generation using real-time power data, weather data, and complexity-based similarity factors | |
| Wang et al. | Performance assessment of photovoltaic modules based on daily energy generation estimation | |
| US10819116B2 (en) | Forecasting solar power generation using real-time power data | |
| US20180101147A1 (en) | Forecasting solar power generation using weather forecasts | |
| CN113497445A (en) | Combined prediction method and system for output of regional multi-scale new energy power station | |
| Chen et al. | Research on wind power prediction method based on convolutional neural network and genetic algorithm | |
| CN115759467A (en) | Time-division integrated learning photovoltaic prediction method for error correction | |
| Mukherjee et al. | Solar irradiance prediction from historical trends using deep neural networks | |
| CN109961190A (en) | A kind of photovoltaic plant power forecasting method a few days ago based on study day degree of fitting cluster | |
| CN118094486B (en) | Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system | |
| CN109657874A (en) | A kind of electric power Mid-long term load forecasting method based on long memory models in short-term | |
| CN105574619A (en) | Solar photovoltaic power generation output prediction system, and prediction method thereof | |
| CN118094486A (en) | Full scene time sequence decomposition-based photovoltaic power generation power prediction method and system | |
| Rashkovska et al. | Online short-term forecasting of photovoltaic energy production | |
| TW201740296A (en) | Method and system for predicting power generation capacity of renewable energy using a neural network to accurately calculate the power generation capacity of renewable energy | |
| CN116307511A (en) | Energy storage configuration method, device, equipment and medium for park power grid | |
| CN115912334A (en) | Method for establishing prediction model of output guarantee rate of wind power plant and prediction method | |
| Zhang et al. | Power forecasting of solar photovoltaic power systems based on similar day and M5'model trees | |
| Li et al. | Short-term Wind Power Forecast based on Teleconnected NWP Information | |
| CN116822658B (en) | Early warning method and device for wind-solar resources, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |