CN117374956A - Short-term prediction method for photovoltaic power generation of comprehensive energy station - Google Patents

Short-term prediction method for photovoltaic power generation of comprehensive energy station Download PDF

Info

Publication number
CN117374956A
CN117374956A CN202311364613.9A CN202311364613A CN117374956A CN 117374956 A CN117374956 A CN 117374956A CN 202311364613 A CN202311364613 A CN 202311364613A CN 117374956 A CN117374956 A CN 117374956A
Authority
CN
China
Prior art keywords
irradiance
model
power generation
photovoltaic power
numerical simulation
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
Application number
CN202311364613.9A
Other languages
Chinese (zh)
Inventor
肖永威
刘博�
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Harbin Tianyuan Petrochemical Engineering Design Co ltd
Original Assignee
Harbin Tianyuan Petrochemical Engineering Design Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Harbin Tianyuan Petrochemical Engineering Design Co ltd filed Critical Harbin Tianyuan Petrochemical Engineering Design Co ltd
Priority to CN202311364613.9A priority Critical patent/CN117374956A/en
Publication of CN117374956A publication Critical patent/CN117374956A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • 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
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • 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
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
    • 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
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin
    • H02J2300/22—The renewable source being solar energy

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Business, Economics & Management (AREA)
  • Molecular Biology (AREA)
  • Economics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Power Engineering (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • General Business, Economics & Management (AREA)
  • Algebra (AREA)
  • Public Health (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Primary Health Care (AREA)
  • Marketing (AREA)
  • Human Resources & Organizations (AREA)
  • Probability & Statistics with Applications (AREA)
  • Water Supply & Treatment (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The application discloses a short-term prediction method for comprehensive energy station photovoltaic power generation, which belongs to the field of photovoltaic power generation and comprises the following steps: constructing a photovoltaic power generation prediction physical model, wherein the photovoltaic power generation prediction physical model comprises a photovoltaic system, a clear sky irradiance model, a meteorological data characteristic model, a local meteorological irradiance model, an environment shielding irradiance model and a simulated photovoltaic power generation model; according to the longitude, latitude and time, obtaining irradiance numerical simulation of the sun through a clear sky irradiance model; optimizing and correcting irradiance numerical simulation through a meteorological data characteristic model, a local meteorological irradiance model and an environmental shielding irradiance model according to meteorological forecast data and environmental parameters; acquiring installation parameters of a photovoltaic system, and calculating and outputting predicted power generation through simulating a photovoltaic power generation model by combining optimized and corrected irradiance numerical simulation; and predicting the final photovoltaic power generation capacity through Bayesian ridge regression and BP neural network. The method is suitable for photovoltaic power generation prediction of the micro-grid application scene.

Description

Short-term prediction method for photovoltaic power generation of comprehensive energy station
Technical Field
The application relates to a short-term prediction method for photovoltaic power generation of a comprehensive energy station, and belongs to the field of photovoltaic power generation.
Background
The photovoltaic micro-grid is a group of small local power grids consisting of distributed photovoltaic, energy storage devices and local loads. Smart micro-grids have great potential in the popularization of new energy sources. Because the photovoltaic is used as clean and environment-friendly energy and has the characteristics of uncontrollable and intermittent, the intelligent micro-grid can use the photovoltaic energy more safely and reliably through the prediction and optimization technology, and stable energy supply is realized.
New energy photovoltaic gas stations, oil reservoirs, and especially stations operated mainly by comprehensive energy are typical micro-grid application scenarios. The research on the photovoltaic power generation prediction plays an important role in the operation of the oil station, and can help the oil station to better manage and utilize photovoltaic power generation resources, and optimize the operation efficiency and economic benefit.
In recent years, photovoltaic power generation is widely used, and photovoltaic power generation prediction is realized by a big data artificial intelligence algorithm. The photovoltaic power generation prediction of the method is firstly to rely on historical data, and uses a big data machine learning and deep learning model to predict. Secondly, the main application scene is a large-scale photovoltaic power station, the site selection condition is that meteorological conditions and surrounding environment are very suitable for photovoltaic power generation, if the surrounding environment is less in shielding, the air environment is stable. However, since the oil adding stations are built on the sides of cities and roads in the micro-grid application scene, and the micro-grid application environment with too few data samples and poor environment such as a new energy station and an oil depot environment is not suitable for the existing photovoltaic power generation prediction due to the reasons of small photovoltaic scale, large environmental difference, pollution to the oil depot environment, simple facilities, lack of historical data, small data dimension and quantity and the like.
Disclosure of Invention
Aiming at the application scenes of micro-grids such as new energy photovoltaic gas stations and oil reservoirs, a short-term prediction method for comprehensive energy station photovoltaic power generation, which can be rapidly applied, is established, and the prediction of photovoltaic power generation capacity is rapidly realized under the condition of no historical data.
To achieve the above object, a first aspect of the present application provides a short-term prediction method for photovoltaic power generation of an integrated energy station, including:
constructing a photovoltaic power generation prediction physical model, wherein the photovoltaic power generation prediction physical model comprises a photovoltaic system, a clear sky irradiance model, a meteorological data characteristic model, a local meteorological irradiance model, an environment shielding irradiance model and a simulated photovoltaic power generation model;
according to longitude, latitude and time, obtaining irradiance numerical simulation of the sun through the clear sky irradiance model;
optimizing and correcting the irradiance numerical simulation through the meteorological data characteristic model, the local meteorological irradiance model and the environmental shielding irradiance model according to meteorological forecast data and environmental parameters;
acquiring installation parameters of a photovoltaic system, and calculating and outputting predicted power generation through simulating a photovoltaic power generation model by combining optimized and corrected irradiance numerical simulation;
and predicting the final photovoltaic power generation amount through Bayesian ridge regression and BP neural network according to the predicted power generation amount.
In one embodiment, the obtaining the irradiance numerical simulation of the sun through the clear sky irradiance model includes:
according to longitude, latitude and time, a solar positioning algorithm is used for obtaining a solar positioning numerical simulation, wherein the solar positioning numerical simulation comprises a zenith angle, an azimuth angle and an altitude angle of the sun;
according to the solar positioning numerical simulation, calculating an irradiance numerical simulation of the sun by using an Ineichen model, wherein the irradiance numerical simulation comprises direct radiation, scattered radiation and total horizontal radiation.
In one embodiment, the optimizing correction of the irradiance numerical simulation by the meteorological data feature model, the local meteorological irradiance model, and the ambient occlusion irradiance model comprises:
optimizing the weather forecast data through the weather data characteristic model, wherein the weather forecast data is 24-hour grid weather forecast, and comprises cloud cover, temperature, wind speed, relative humidity, accumulated precipitation in the current hour and atmospheric pressure;
according to the optimized weather forecast data, performing sunny piecewise linear optimization on the irradiance numerical simulation through a local weather irradiance model;
and according to the environmental parameters, carrying out shielding piecewise linear optimization on the irradiance numerical simulation through an environmental shielding irradiance model.
In one embodiment, the optimizing the weather forecast data by the weather data feature model includes:
according to the cloud cover, aiming at the characteristic of the hour-by-hour prediction of the 24-hour grid weather forecast, smooth data are obtained through linear interpolation in the middle half hour, and the optimized cloud cover is obtained by combining with weather forecast data except the cloud cover.
In one embodiment, the performing a sunny piecewise linear optimization of the irradiance numerical simulation by a local meteorological irradiance model includes:
taking a predicted power generation value and an actual power generation value when the cloud cover is less than or equal to 25% as samples, counting a sectional residual error change rate mean value through a sectional linear function and a Bayesian parameter adjustment method, and obtaining the irradiance numerical simulation of the fine sectional linear optimization through the sectional residual error change rate mean value.
In one embodiment, the performing the occlusion piecewise linear optimization of the irradiance numerical simulation by the ambient occlusion irradiance model includes:
respectively acquiring included angles between the shielding object and the far end of the photovoltaic field, between the shielding object and the central point of the photovoltaic field and between the shielding object and the near end of the photovoltaic field according to environmental parameters, and respectively calculating height angles corresponding to the included angles as a far-end height angle, a central height angle and a near-end height angle of the sun through a plane geometry method;
and carrying out sectional reduction on irradiance corresponding to the far-end altitude angle, the central altitude angle and the near-end altitude angle, and simultaneously obtaining the irradiance numerical simulation of shielding sectional linear optimization by combining a Bayesian parameter adjustment method.
In one embodiment, the predicting the final photovoltaic power generation by bayesian ridge regression and BP neural network further comprises:
constructing a deep learning CNN model, wherein the deep learning CNN model adopts an improved LeNet-5 network structure, and the improved LeNet-5 network structure comprises an input layer, a convolution layer, a pooling layer, a full connection layer and an output layer;
and when the photovoltaic power generation amount is predicted in a mode of the Bayesian ridge regression and the BP neural network to exceed a preset period, predicting the final photovoltaic power generation amount by using the deep learning CNN model by taking prediction data in the preset period as input, wherein the preset period is more than or equal to one year.
In one embodiment, the photovoltaic system, the clear sky irradiance model, and the simulated photovoltaic power generation model are designed based on pvlib open source projects.
A second aspect of the present application provides 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 first aspect or any implementation of the first aspect as described above when the computer program is executed.
A third aspect of the present application provides a computer readable storage medium storing a computer program which when executed by a processor performs the steps of the first aspect or any implementation of the first aspect.
From the above, the application provides a short-term prediction method for photovoltaic power generation of a comprehensive energy station, which is used for directly using the photovoltaic power generation prediction physical model in the stage without actual operation data in the initial online stage of the energy station by constructing the photovoltaic power generation prediction physical model, does not need historical data, and is suitable for new and old environments; meanwhile, the influence of factors such as local weather, pollution and surrounding environment shielding on solar radiation energy is considered, a weather data characteristic model, a local weather irradiance model and an environment shielding irradiance model are innovatively added, so that irradiance numerical simulation of the sun has personalized characteristics, the localization requirements are met, the micro-grid environment of a non-special solar power station such as a new energy station and an oil depot is more suitable, and the accuracy of a photovoltaic power generation numerical simulation local physical model is improved; on the basis of the photovoltaic power generation prediction physical model, the method is combined with weather forecast data, under the condition of a small sample, bayesian ridge regression and BP neural network algorithm which are wide in adaptability and good in return capability are used, the condition of overfitting due to few samples is reduced, the prediction accuracy is improved, a short-term prediction method for comprehensive energy station photovoltaic power generation is established, and the prediction capability of photovoltaic power generation is rapidly realized under the condition of no historical data.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following description will briefly introduce the drawings that are needed in the embodiments or the description of the prior art, it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of the overall structure of a photovoltaic power generation prediction physical model according to an embodiment of the present application;
fig. 2 is a schematic view of solar positioning according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of a residual mean between a predicted power generation value and an actual power generation value according to an embodiment of the present disclosure;
FIG. 4 is a schematic view of selecting a height angle as a segment optimization angle according to an embodiment of the present application;
FIG. 5 is a schematic diagram of an environmental occlusion effect on power generation according to an embodiment of the present disclosure;
FIG. 6 is a plan geometry of solar altitude under a shade according to an embodiment of the present application;
fig. 7 is a schematic diagram of filtering influence of shielding on power generation according to an embodiment of the present application;
fig. 8 is a structural diagram of a photovoltaic system according to an embodiment of the present application;
fig. 9 is a schematic diagram of a BP neural network according to an embodiment of the present application;
fig. 10 is a diagram of an improved LeNet-5 network according to an embodiment of the present application.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It should be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The following description of the embodiments of the present application, taken in conjunction with the accompanying drawings, clearly and fully describes the technical solutions of the embodiments of the present application, and it is evident that the described embodiments are only some embodiments of the present application, 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 within the scope of the present disclosure.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application, but the present application may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present application is not limited to the specific embodiments disclosed below.
Example 1
The embodiment of the application provides a short-term prediction method for photovoltaic power generation of a comprehensive energy station, as shown in fig. 1, the method comprises the following steps:
s100, constructing a photovoltaic power generation prediction physical model, wherein the photovoltaic power generation prediction physical model comprises a photovoltaic system, a clear sky irradiance model, a meteorological data characteristic model, a local meteorological irradiance model, an environment shielding irradiance model and a simulated photovoltaic power generation model;
optionally, the photovoltaic system, the clear sky irradiance model and the simulated photovoltaic power generation model are designed by combining a pvlib open source project, influences of factors such as local weather, pollution and surrounding environment on solar radiation energy are considered, and a meteorological data characteristic model, a local weather irradiance model and an environment shielding irradiance model are added to obtain the photovoltaic power generation prediction physical model. And then acquiring 24H grid weather forecast, longitude, latitude, time, environmental parameters and installation parameters of a photovoltaic system in a photovoltaic power generation area of the energy station as input parameters, and inputting the input parameters into a photovoltaic power generation prediction physical model to obtain output predicted power generation. S200, according to longitude, latitude and time, obtaining irradiance numerical simulation of the sun through the clear sky irradiance model;
optionally, the obtaining the irradiance numerical simulation of the sun through the clear sky irradiance model includes:
according to longitude, latitude and time, a solar positioning algorithm is used for obtaining a solar positioning numerical simulation, wherein the solar positioning numerical simulation comprises a zenith angle, an azimuth angle and an altitude angle of the sun;
according to the solar positioning numerical simulation, calculating an irradiance numerical simulation of the sun by using an Ineichen model, wherein the irradiance numerical simulation comprises direct radiation, scattered radiation and total horizontal radiation.
In one embodiment, when obtaining the irradiance numerical simulation of the sun, firstly, a Solar Positioning Algorithm (SPA) of NREL is used to perform solar positioning, the algorithm calculates the solar zenith angle and azimuth angle during the year 2000 to 6000 according to the date, time and position on the earth, the uncertainty is +/-0.0003 degrees, the solar positioning algorithm of NREL is implemented in pvlib, and longitude (lon), latitude (lat), altitude (alt) and time (t) are input, so that solar zenith angle (z), azimuth angle (a) and altitude angle (h) can be obtained, as shown in fig. 2, the solar zenith angle and azimuth angle are expressed as:
z,A＝SPA(lon,lat,alt,t)
the solar altitude is:
h＝90°-z
in one embodiment, the clear sky irradiance model is based on astronomical and atmospheric principles, and is calculated to predict the theoretical maximum radiation value for theoretical direct and scattered radiation without total cloud occlusion. The model is based on the absorption and scattering effects of atmospheric gases and aerosols on radiation, and on the rotation, revolution and thickness of the earth. In pvlib, the Ineichen model was used, and the resulting irradiance numerical simulations included direct radiation (DirectNormal Irradiance, DNI), scattered radiation (Diffuse Horizontal Irradiance, DHI), and total horizontal radiation (Global Horizontal Irradiance, GHI). Wherein DNI refers to direct sunlight from the solar panel to a surface orthogonal to the optical path, DHI refers to direct ground-reaching sunlight scattered in the atmosphere, GHI refers to total radiation of DHI and DNI of the sun reaching a horizontal surface, and the relationship between three radiation parameters: ghi=dhi+dni×cosz.
S300, optimizing and correcting irradiance numerical simulation through the meteorological data characteristic model, the local meteorological irradiance model and the environmental shielding irradiance model according to weather forecast data and environmental parameters;
optionally, the optimizing and correcting the irradiance numerical simulation through the meteorological data feature model, the local meteorological irradiance model and the environmental shielding irradiance model includes:
optimizing the weather forecast data through the weather data feature model, wherein the weather forecast data is a 24-hour grid weather forecast, and comprises cloud cover (cloud), temperature (temp), wind speed (windSpeed), relative humidity (hub), current hour accumulated precipitation (pre) and atmospheric pressure (pressure), and further comprises a weather condition icon code (icon) and a weather condition text description (text); the weather forecast data are collected time by time through an Internet third party (wind) API;
according to the optimized weather forecast data, performing sunny piecewise linear optimization on the irradiance numerical simulation through a local weather irradiance model;
and according to the environmental parameters, carrying out shielding piecewise linear optimization on the irradiance numerical simulation through an environmental shielding irradiance model.
Optionally, the optimizing the weather forecast data through the weather data feature model includes:
according to the cloud cover, aiming at the characteristic of time-by-time forecasting of 24-hour grid weather forecast, smooth data are obtained through linear interpolation in the middle half hour, and the optimized cloud cover is obtained by combining weather forecast data except the cloud cover and temperature.
In one implementation mode, cloud computing in weather forecast data is used as a local solar irradiance prediction core, cloud computing and text description fusion are used in a weather data feature model, and smooth data are obtained through linear interpolation in the middle half hour according to the characteristics of weather forecast time-by-time forecast, and meanwhile, cloud computing is simulated and optimized through actual measurement according to comprehensive forecast content. Wherein, the linear interpolation formula:
alternatively, other optimized weather parameters such as temperature, wind speed, relative humidity, accumulated precipitation of the current hour and atmospheric pressure can be obtained according to the above linear interpolation formula, and linear interpolation is performed according to the above linear interpolation formula.
Optionally, the performing sunny piecewise linear optimization on the irradiance numerical simulation through a local meteorological irradiance model includes:
taking a predicted power generation value and an actual power generation value when the cloud cover is less than or equal to 25% as samples, counting a sectional residual error change rate mean value through a sectional linear function and a Bayesian parameter adjustment method, and obtaining the irradiance numerical simulation of the fine sectional linear optimization through the sectional residual error change rate mean value.
In one embodiment, the solar energy resource is used under the influence of illumination intensity and meteorological environment, has stronger volatility and randomness, and is more easily influenced by air pollution especially in cities and the periphery thereof. Although researches are carried out to predict the generated power of the photovoltaic power station in severely polluted regions and time, the method is favorable for the construction planning and the scheduling planning of the regional power grid, but is a complex method based on a large amount of historical data, deep learning and the like. Because the gas station is in the environment, the gas station is more easily affected by local environment, such as local air environment (dust), and after the sun rises, the gas station is affected by traffic and peripheral production activities, so that the solar irradiance is greatly affected. Aiming at the universality characteristic of solar irradiance numerical simulation, the embodiment of the application adopts sunny piecewise linear optimization irradiance by combining with the local environment characteristic, so that the solar irradiance numerical simulation has personalized characteristics, and the localization requirement is met. The specific method comprises the following steps: when the cloud cover is smaller than or equal to a preset value (for example, the cloud cover is smaller than or equal to 25%), recent photovoltaic power generation data are counted, wherein the recent photovoltaic power generation data are counted for 15 to 30 days, and residual average values between the predicted power generation value and the actual power generation value are counted according to the recent photovoltaic power generation data and are divided into peak sections and average sections. Taking fig. 3 as a local air environment reference, the residual error mean value is:
where y is the actual power generation amount,is a predicted power generation amount, and t represents a certain period.
In order to facilitate the operation and quick response of the gas station, the predicted power generation value calculated by the comparison theory and the actual power generation value are adopted to correct the GHI method by adopting a simple piecewise linear function, so that the characteristics of a local small-range environment are compensated. Aiming at the characteristic of the operation period of the photovoltaic power generation sun, the height angle theta is selected a For the segment optimization angle, the segment height angle is shown in fig. 4, and the optimized irradiance numerical simulation is:
the local meteorological irradiance peak and flat parameters are obtained through statistics of a mean value of a sectional residual error change rate, the local meteorological irradiance peak and flat parameters participate in optimization approximate solution through a Bayesian parameter adjustment method to obtain approximate values peak=0.23, flat=0.11, peak is a parameter defining a photovoltaic power generation peak period, and flat is a parameter defining a power generation amount higher than an initial period and smaller than a peak period;
GHI＝GHI L ·(offset+(1-offset)·(1-Cloud))
wherein offset=0.35, which means a cloud amount deviation coefficient;
DNI was calculated using GHI, using the DISC model, which is an empirical algorithm implemented in pvlib. :
DNI＝DISC(GHI,θ,Cloud,pressure)
DHI＝GHI-DNI·cos(θ)
optionally, the performing the shielding piecewise linear optimization on the irradiance numerical simulation through the ambient shielding irradiance model includes:
respectively acquiring included angles between the shielding object and the far end of the photovoltaic field, between the shielding object and the central point of the photovoltaic field and between the shielding object and the near end of the photovoltaic field according to environmental parameters, and respectively calculating height angles corresponding to the included angles as a far-end height angle, a central height angle and a near-end height angle of the sun through a plane geometry method;
and carrying out sectional reduction on irradiance corresponding to the far-end altitude angle, the central altitude angle and the near-end altitude angle, and simultaneously obtaining the irradiance numerical simulation of shielding sectional linear optimization by combining a Bayesian parameter adjustment method.
In one embodiment, since the new energy photovoltaic gas station and the oil depot are mostly located on cities and traffic lines, the new energy photovoltaic gas station and the oil depot are always influenced by shielding of surrounding mountain bodies and building bodies, and shielding influences irradiance, so that larger errors exist between actual generated energy and predicted generated energy, and the shielding influence is shown in fig. 5. Aiming at the problem of complex shielding of the environment, a local shielding and solar irradiance influence approximation algorithm model and method are established, namely, the environment shielding irradiance model, the corresponding angle is calculated by a plane geometry method according to the included angles of a shielding scene and the far end, the central point and the near end of the photovoltaic field, the plane geometry schematic diagram is shown in fig. 6, and the far end altitude angle is calculated:
center height angle:
proximal height angle:
wherein h is pv For shielding the height of the object, l pv Distance d is the distance between the far end and the near end of the photovoltaic field pv Is the distance between the near end of the photovoltaic field and the shielding object. Solar irradiance corresponding to the far-end, center point and near-end altitude angle is reduced in a segmentation mode, and a linear function is approximated:
GHI＝GHI·P 0 ,θ≥θ bn andθ<θ bo
DNI＝DNI·P 0 ,θ≥θ bn andθ<θ bo
DHI＝DHI·P 0 ,θ≥θ bn andθ<θ bo
GHI＝GHI·P 1 ,θ≥θ bo andθ<θ bf
DNI＝DNI·P 1 ,θ≥θ bo andθ<θ bf
DHI＝DHI·0.5,θ≥θ bo andθ<θ bf
DHI＝0,θ≥θ bf
GHI＝GHI·P 2 ·0.5,θ≥θ bf
DNI＝DNI·P 2 ,θ≥θ bf
wherein P is 0 P 1 P 2 Defined as param0, param1 and param2, can participate in optimization and P by a Bayesian parameter adjustment method 0 P 1 P 2 And (5) solving. Filtering out the occlusion result is shown in fig. 7, and it can be seen that the predicted power generation amount matches the actual power generation amount.
In one embodiment, the bayesian parameter tuning method adopts a bayesian optimization idea, namely an approximation idea, and the basic idea of parameter optimization is to estimate posterior distribution of an objective function based on data by using bayesian theorem, and then select the next sampled super-parameter combination according to the distribution. The method fully utilizes the information of the previous sampling point, and the optimized working mode is to learn the shape of the objective function and find the parameter which makes the result promoted to the global maximum. Bayesian optimization algorithm framework:
for t＝1,2,...do
maximizing the acquisition function to obtain the next evaluation point: x is x i ＝argmax x∈X a(x|D 1:t-1 )
Evaluating the objective function value: y is i ＝f(x i )+ε i
Integrating data: d (D) t ＝D t-1 ∪{x i ,y i -and updating the probabilistic proxy model;
end for
optimizing parameter adjustment parameters: peak, flat; praram0, param1, param2
Objective function:
where y is the actual value, and where,is the numerical analog predictor p_mp.
Executing optimized parameter adjusting results:
{'target':-105.0,'params':{'flat':0.5,'param0':0.5,'param1':0.95,'param2':0.95,'peak':0.5}}。
s400, acquiring installation parameters of a photovoltaic system, and calculating and outputting predicted power generation through simulating a photovoltaic power generation model by combining optimized and corrected irradiance numerical simulation;
optionally, as shown in fig. 8, the structure diagram of the photovoltaic system is composed of each photovoltaic cell Array (Array) and an inverter (inverter), wherein the installation parameters of the photovoltaic cell Array include an inclination angle, an azimuth angle and the like, and other parameters such as a module temperature model, photovoltaic module configuration and the like. Configuration parameters of the inverter include efficiency and power.
In one implementation, the embodiment of the application constructs a photovoltaic system according to a pvlib power model and obtains installation parameters, wherein the pvlib power model parameters include:
photovoltaic array:
'pdc0': the peak power of the photovoltaic module is W;
'gamma_pdc': a temperature coefficient;
'temp_ref': the reference temperature of the photovoltaic module is expressed as the unit of the temperature;
an inverter:
'pdc0': peak power of the inverter, the unit is W;
'eta_inv_nom': inverter efficiency.
In one embodiment, the local meteorological irradiance model may be calculated based on the data from which the photovoltaic system operates without historical data.
In one embodiment, the predicted photovoltaic power generation is simulated based on 24-hour grid weather predictions, local environmental information, and the like, according to a simulated photovoltaic power generation model that has been configured. The simulated photovoltaic power generation model has a module from the SAPM database and an inverter from the CEC database, and the AC model and the AOI loss model are selected according to settings. According to the embodiment of the application, a power generation model is formed according to a photovoltaic system in the Pvlib, a clear sky irradiance model, in-process attribute values and the like, and the photovoltaic power generation prediction physical model is further combined with a meteorological data characteristic model, a local meteorological irradiance model and an environment shielding irradiance model to obtain photovoltaic power generation capacity through simulation calculation. The output content of the photovoltaic power generation prediction physical model comprises: sun position information, sun clear sky irradiance information, sun ground irradiance information, weather forecast information, predicted power generation, time and the like. Wherein, the predicted power generation amount:
p_mp＝G(GHI,DHI,DNI,temp,windSpeed)
s500, predicting the final photovoltaic power generation amount through Bayesian ridge regression and BP neural network according to the predicted power generation amount.
In one embodiment, when the system operates for about 15 days and is subjected to overcast, rainy and sunny days, a small amount of data is accumulated, and at this time, only short-term operation data is used as an input sample, so that a machine learning algorithm Bayesian ridge regression and BP neural network with wider adaptability and better return capability are adopted, and the situation of overfitting due to fewer samples is reduced. The Bayesian ridge regression is a regression method based on Bayesian statistical theory and is used for processing linear regression problems. The back propagation neural algorithm (Back Propagation NeuralNetwork) is called BP neural network for short, and is a multi-layer feedforward network based on error back propagation. The schematic structural diagram of the BP neural network is shown in fig. 9, and the BP neural network comprises an input layer, a hidden layer and an output layer, wherein the BP neural network hidden layer adopts a three-layer structure, and is 35 multiplied by 35. The training process consists of forward transmission of network signals and reverse transmission of errors, wherein the forward transmission of signals is forward transmission of forward direction calculation output values; the latter uses a gradient descent method to adjust weights and thresholds to calculate the optimum value of the objective function.
Further, the output content of the photovoltaic power generation prediction physical model is input to an input layer, so that bayesian ridge regression and BP neural network are performed, machine learning is performed, and further predicted photovoltaic power generation is output, and input characteristics of the machine learning are shown in table 1:
table 1 machine learning input features
Based on the above input features, pearson correlation coefficient (Pearson) is taken as the main basis for feature selection. The pearson correlation coefficient method measures the linear correlation between variables, and the value interval of the result is [ -1,1], -1 represents the complete negative correlation, +1 represents the complete positive correlation, and 0 represents no linear correlation.
Optionally, the predicting the final photovoltaic power generation through bayesian ridge regression and the BP neural network further includes:
constructing a deep learning CNN model, wherein the deep learning CNN model adopts an improved LeNet-5 network structure, and the improved LeNet-5 network structure comprises an input layer, a convolution layer, a pooling layer, a full connection layer and an output layer;
and when the photovoltaic power generation amount is predicted in a mode of the Bayesian ridge regression and the BP neural network to exceed a preset period, predicting the final photovoltaic power generation amount by using the deep learning CNN model by taking prediction data in the preset period as input, wherein the preset period is more than or equal to one year.
In one embodiment, to further increase the prediction accuracy, when the photovoltaic system operates for more than one year, the data is accumulated for more than one year, and the accumulated data of one year period may be used, with the whole day data as input, to predict the photovoltaic power generation amount using a deep learning CNN algorithm. The core idea of adopting deep learning CNN algorithm prediction is as follows: based on data such as 48 groups of solar irradiation and solar positioning time sequence data thereof, all-day grid point weather forecast and the like every 30 minutes in the whole day, a CNN neural network model is trained to predict the all-day photovoltaic power generation capacity, and the situation that the prediction accuracy of a machine learning algorithm is insufficient is made up. Convolutional neural networks (Convolutional Neural Network, CNN) are a type of feedforward neural network that is widely used in the fields of image, speech, and natural language processing, among others. The CNN is based on convolution operation of neurons, and can effectively identify and extract the characteristics of input data. Lenet-5 is a classical convolutional neural network structure that was first used in handwriting character recognition applications. The embodiment of the application improves the LeNet-5 network structure, and considers the solar irradiance, solar positioning and weather forecast realization data all the day as a two-dimensional image, wherein the dimension data of the rows are time sequences, and the dimension data of the columns are all characteristic data items.
Specifically, the modified LeNet-5 network structure comprises 7 layers, the structure of which is shown in FIG. 10, mainly comprising a convolution layer, a pooling layer and a full connection layer, and finally an output layer uses an identity activation function.
Wherein, input layer: 20 feature columns, 50 line data points;
convolution layer 1: generating 16 feature maps by using 16 convolution kernels with the size of 3×3;
pooling layer 1: downsampling each feature map using a 2 x 2 max pooling kernel, outputting 6 50 x 10 subgraphs;
convolution layer 2: generating 32 feature maps by using 32 convolution kernels with the size of 3×3;
pooling layer 2: downsampling each feature map using a 2 x 2 max pooling kernel, outputting 32 25 x 5 subgraphs; the method comprises the steps of carrying out a first treatment on the surface of the
Full tie layer 1: expanding 32 sub-images 25 multiplied by 5 into one-dimensional vectors, and processing through the full-connection layer of 256 neurons;
full tie layer 2: a fully connected layer of 128 neurons was used.
In one implementation mode, regression model evaluation indexes R2, RMSE and MAE are used for respectively testing and verifying numerical simulation, bayesian ridge regression, random forest and BP neural network models. Wherein, the value range of the R2 (decision coefficient) evaluation model is 0 to 1, and the closer to 1 the value is, the better the fitting degree of the model is (score used for the sklearn regression model evaluation is the calculation formula of R2 by default). The calculation formula of R2 is as follows:
the root mean square error (Root Mean Squared Error, RMSE) is the square root of the MSE, with smaller values indicating higher accuracy of model predictions. The calculation formula is as follows:
the mean absolute error (Mean Absolute Error, MAE) is the average of the absolute values of the differences between the predicted and real values, with smaller values representing higher accuracy of the model predictions. The calculation formula is as follows:
the evaluation results are shown in table 2:
table 2 model evaluation comparative table
As shown in table 2, in the machine learning algorithm with few samples, the generalization ability of bayesian ridge regression is better, the BP neural network has better effect on the training set, the fitting is easy, and the random forest algorithm has various items in between.
In the practical environment application process, the result shows that the photovoltaic power generation prediction model is rapidly deployed under the condition of no historical data, photovoltaic prediction is realized, if the online time is long enough, the data accumulation is realized, the machine learning method plays a larger role, and the prediction is more accurate.
In an application scenario, python development language is used, models are built based on sklearn, pvlib, pandas, tensorflow and other tools, and a photovoltaic power generation prediction machine learning model is built according to the photovoltaic power generation prediction physical model, bayesian ridge regression, BP neural network and deep learning CNN model and is used for predicting the electric quantity of the photovoltaic system for directly converting solar radiation energy into electric energy within 24 hours in the future. And taking the 24H grid point weather forecast, longitude, latitude, time, environmental parameters and photovoltaic installation parameters as inputs to obtain predicted photovoltaic power generation quantity as output, so as to realize the prediction of the photovoltaic power generation quantity.
From the above, the embodiment of the application provides a short-term prediction method for photovoltaic power generation of a comprehensive energy station, which uses pvlib numerical simulation, does not need historical data, and is suitable for new and old environments; the method is combined with a meteorological data characteristic model, a local meteorological irradiance model and an environment shielding irradiance model, and is more suitable for the micro-grid environment of a non-special solar power station such as a new energy station, an oil depot and the like; the piecewise linear regression algorithm model is adopted, and the established engineering algorithm model is simple and easy to use and has strong interpretation according to engineering experience and by combining with Bayesian optimization parameter adjustment; the environmental shielding irradiance model adopts a solar positioning technology, irradiance numerical simulation optimization and corrects irradiance in a shielding environment by a simplified plane geometry method; the LeNet-5 network is improved, the convolutional neural network is identified from the image to be used for regression prediction, the traditional big data prediction analysis is improved and applicable, the characteristics of simplicity, rapidness and accuracy in identification are brought into play, the photovoltaic power generation prediction accuracy is improved, and the application threshold of deep learning is reduced; big data are not used, the requirement on the hardware environment is low, and the implementation period is short. The Bayesian ridge regression algorithm is used, so that the method is simple, has strong generalization capability and is suitable for the conditions of few samples and complex environment.
Example two
The embodiment of the application provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the memory is used for storing the software program and a module, and the processor executes various functional applications and data processing by running the software program and the module stored in the memory. The memory and the processor are connected by a bus. In particular, the processor implements any of the steps of the above-described embodiment by running the above-described computer program stored in the memory.
It should be appreciated that in embodiments of the present application, the processor may be a central processing unit (Central Processing Unit, CPU), which may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf Programmable gate arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may include read-only memory, flash memory, and random access memory, and provides instructions and data to the processor. Some or all of the memory may also include non-volatile random access memory.
It should be appreciated that the above-described integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer-readable storage medium. Based on such understanding, the present application may implement all or part of the flow of the method of the above embodiment, or may be implemented by instructing related hardware by a computer program, where the computer program may be stored in a computer readable storage medium, and the computer program may implement the steps of each method embodiment described above when executed by a processor. The computer program comprises computer program code, and the computer program code can be in a source code form, an object code form, an executable file or some intermediate form and the like. The computer readable medium may include: any entity or device capable of carrying the computer program code described above, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth. The content of the computer readable storage medium can be appropriately increased or decreased according to the requirements of the legislation and the patent practice in the jurisdiction.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.
It should be noted that, the method and the details thereof provided in the foregoing embodiments may be combined into the apparatus and the device provided in the embodiments, and are referred to each other and are not described in detail.
Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus/terminal device and method may be implemented in other manners. For example, the apparatus/device embodiments described above are merely illustrative, e.g., the division of modules or elements described above is merely a logical functional division, and may be implemented in other ways, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed.
The above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.

Claims (10)

1. A short-term prediction method for photovoltaic power generation of an integrated energy station, comprising:
constructing a photovoltaic power generation prediction physical model, wherein the photovoltaic power generation prediction physical model comprises a photovoltaic system, a clear sky irradiance model, a meteorological data characteristic model, a local meteorological irradiance model, an environment shielding irradiance model and a simulated photovoltaic power generation model;
according to longitude, latitude and time, obtaining irradiance numerical simulation of the sun through the clear sky irradiance model;
optimizing and correcting the irradiance numerical simulation through the meteorological data characteristic model, the local meteorological irradiance model and the environmental shielding irradiance model according to meteorological forecast data and environmental parameters;
acquiring installation parameters of a photovoltaic system, and calculating and outputting predicted power generation through simulating a photovoltaic power generation model by combining optimized and corrected irradiance numerical simulation;
and predicting the final photovoltaic power generation amount through Bayesian ridge regression and BP neural network according to the predicted power generation amount.
2. The short-term prediction method of claim 1, wherein the deriving a numerical simulation of irradiance of the sun from the clear sky irradiance model comprises:
according to longitude, latitude and time, a solar positioning algorithm is used for obtaining a solar positioning numerical simulation, wherein the solar positioning numerical simulation comprises a zenith angle, an azimuth angle and an altitude angle of the sun;
according to the solar positioning numerical simulation, calculating an irradiance numerical simulation of the sun by using an Ineichen model, wherein the irradiance numerical simulation comprises direct radiation, scattered radiation and total horizontal radiation.
3. The short-term prediction method of claim 1 or 2, wherein said optimizing modification of said irradiance numerical simulation by said meteorological data feature model, said local meteorological irradiance model, and said ambient occlusion irradiance model comprises:
optimizing the weather forecast data through the weather data characteristic model, wherein the weather forecast data is 24-hour grid weather forecast, and comprises cloud cover, temperature, wind speed, relative humidity, accumulated precipitation in the current hour and atmospheric pressure;
according to the optimized weather forecast data, performing sunny piecewise linear optimization on the irradiance numerical simulation through a local weather irradiance model;
and according to the environmental parameters, carrying out shielding piecewise linear optimization on the irradiance numerical simulation through an environmental shielding irradiance model.
4. A short-term prediction method according to claim 3, wherein said optimizing said weather forecast data by said weather data feature model comprises:
according to the cloud cover, aiming at the characteristic of the hour-by-hour prediction of the 24-hour grid weather forecast, smooth data are obtained through linear interpolation in the middle half hour, and the optimized cloud cover is obtained by combining with weather forecast data except the cloud cover.
5. The short-term prediction method of claim 3, wherein said performing a sunny piecewise linear optimization of said irradiance numerical simulation by a local meteorological irradiance model comprises:
taking a predicted power generation value and an actual power generation value when the cloud cover is less than or equal to 25% as samples, counting a sectional residual error change rate mean value through a sectional linear function and a Bayesian parameter adjustment method, and obtaining the irradiance numerical simulation of the fine sectional linear optimization through the sectional residual error change rate mean value.
6. The short-term prediction method of claim 3, wherein the performing the occlusion piecewise linear optimization of the irradiance numerical simulation by the ambient occlusion irradiance model comprises:
respectively acquiring included angles between the shielding object and the far end of the photovoltaic field, between the shielding object and the central point of the photovoltaic field and between the shielding object and the near end of the photovoltaic field according to environmental parameters, and respectively calculating height angles corresponding to the included angles as a far-end height angle, a central height angle and a near-end height angle of the sun through a plane geometry method;
and carrying out sectional reduction on irradiance corresponding to the far-end altitude angle, the central altitude angle and the near-end altitude angle, and simultaneously obtaining the irradiance numerical simulation of shielding sectional linear optimization by combining a Bayesian parameter adjustment method.
7. The short-term prediction method according to claim 1 or 2, characterized in that the predicting the final photovoltaic power generation by bayesian ridge regression and BP neural network further comprises:
constructing a deep learning CNN model, wherein the deep learning CNN model adopts an improved LeNet-5 network structure, and the improved LeNet-5 network structure comprises an input layer, a convolution layer, a pooling layer, a full connection layer and an output layer;
and when the photovoltaic power generation amount is predicted in a mode of the Bayesian ridge regression and the BP neural network to exceed a preset period, predicting the final photovoltaic power generation amount by using the deep learning CNN model by taking prediction data in the preset period as input, wherein the preset period is more than or equal to one year.
8. The short-term prediction method according to claim 1 or 2, characterized in that the photovoltaic system, the clear sky irradiance model and the simulated photovoltaic power generation model are designed based on pvlib open source project.
9. An electronic device, comprising: memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 8 when the computer program is executed.
10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 8.
CN202311364613.9A 2023-10-20 2023-10-20 Short-term prediction method for photovoltaic power generation of comprehensive energy station Pending CN117374956A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311364613.9A CN117374956A (en) 2023-10-20 2023-10-20 Short-term prediction method for photovoltaic power generation of comprehensive energy station

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311364613.9A CN117374956A (en) 2023-10-20 2023-10-20 Short-term prediction method for photovoltaic power generation of comprehensive energy station

Publications (1)

Publication Number Publication Date
CN117374956A true CN117374956A (en) 2024-01-09

Family

ID=89392378

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311364613.9A Pending CN117374956A (en) 2023-10-20 2023-10-20 Short-term prediction method for photovoltaic power generation of comprehensive energy station

Country Status (1)

Country Link
CN (1) CN117374956A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117791687A (en) * 2024-02-28 2024-03-29 长峡数字能源科技(湖北)有限公司 Energy management method of photovoltaic energy storage system
CN117937478A (en) * 2024-03-22 2024-04-26 长江三峡集团实业发展(北京)有限公司 Photovoltaic power prediction method, device, computer equipment and storage medium
CN117937478B (en) * 2024-03-22 2024-05-28 长江三峡集团实业发展(北京)有限公司 Photovoltaic power prediction method, device, computer equipment and storage medium

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117791687A (en) * 2024-02-28 2024-03-29 长峡数字能源科技(湖北)有限公司 Energy management method of photovoltaic energy storage system
CN117791687B (en) * 2024-02-28 2024-05-14 长峡数字能源科技(湖北)有限公司 Energy management method of photovoltaic energy storage system
CN117937478A (en) * 2024-03-22 2024-04-26 长江三峡集团实业发展(北京)有限公司 Photovoltaic power prediction method, device, computer equipment and storage medium
CN117937478B (en) * 2024-03-22 2024-05-28 长江三峡集团实业发展(北京)有限公司 Photovoltaic power prediction method, device, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
JP6759966B2 (en) How to operate the photovoltaic power generation system
CN108388956B (en) Photovoltaic power prediction method considering radiation attenuation
CN113128793A (en) Photovoltaic power combination prediction method and system based on multi-source data fusion
Yesilbudak et al. A review of data mining and solar power prediction
Carneiro et al. Review on photovoltaic power and solar resource forecasting: current status and trends
JP6693330B2 (en) Solar power system operation
Alomari et al. A predictive model for solar photovoltaic power using the Levenberg-Marquardt and Bayesian regularization algorithms and real-time weather data
CN110766134A (en) Photovoltaic power station short-term power prediction method based on cyclic neural network
CN110781458B (en) Method for predicting surface solar irradiance based on mixed regression model
CN117374956A (en) Short-term prediction method for photovoltaic power generation of comprehensive energy station
CN114492941A (en) Whole-county photovoltaic prediction method based on cluster division and data enhancement
CN115759467A (en) Time-division integrated learning photovoltaic prediction method for error correction
CN114882373A (en) Multi-feature fusion sandstorm prediction method based on deep neural network
Tajjour et al. A novel strategy for solar irradiance forecasting using deep learning techniques and validation for a himalayan location in india as a case study
CN116722544B (en) Distributed photovoltaic short-term prediction method and device, electronic equipment and storage medium
Ehsan et al. Artificial neural network predictor for grid-connected solar photovoltaic installations at atmospheric temperature
CN117526274A (en) New energy power prediction method, electronic equipment and storage medium in extreme climate
Gupta et al. Short Term Solar Irradiation Forecasting using CEEMDAN Decomposition Based BiLSTM Model Optimized by Genetic Algorithm Approach
CN116029440A (en) Ultra-short-term power prediction method and device for photovoltaic power station
CN112734073A (en) Photovoltaic power generation short-term prediction method based on long and short-term memory network
CN117175569B (en) Photovoltaic prediction method and system based on refined weather typing
Pham et al. LS-SPP: A LSTM-Based Solar Power Prediction Method from Weather Forecast Information
Kyliashkina et al. Intelligent Systems as a Tool for Predicting Electrical Energy and Power Generation
Zhu et al. Solar Radiation Prediction Based on Convolution Neural Network and Long Short-Term Memory. Energies 2021, 14, 8498
Zhang et al. A novel intra-hour PV power forecasting technique based on total-sky images

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