CN112215428A - Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic - Google Patents

Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic Download PDF

Info

Publication number
CN112215428A
CN112215428A CN202011126378.8A CN202011126378A CN112215428A CN 112215428 A CN112215428 A CN 112215428A CN 202011126378 A CN202011126378 A CN 202011126378A CN 112215428 A CN112215428 A CN 112215428A
Authority
CN
China
Prior art keywords
day
power generation
photovoltaic power
error correction
generation power
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
Application number
CN202011126378.8A
Other languages
Chinese (zh)
Other versions
CN112215428B (en
Inventor
王士柏
王玥娇
孙树敏
程艳
于芃
张用
滕玮
王楠
游大宁
袁森
张元鹏
徐征
李俊恩
袁帅
张兴友
魏大钧
邢家维
赵帅
张永明
郭永超
李庆华
王彦卓
常万拯
张志豪
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.)
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd
State Grid Shandong Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd
State Grid Shandong Electric Power 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 State Grid Corp of China SGCC, Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd, State Grid Shandong Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202011126378.8A priority Critical patent/CN112215428B/en
Publication of CN112215428A publication Critical patent/CN112215428A/en
Application granted granted Critical
Publication of CN112215428B publication Critical patent/CN112215428B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06N3/045—Combinations of 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
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Evolutionary Computation (AREA)
  • Strategic Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • General Business, Economics & Management (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Water Supply & Treatment (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A photovoltaic power generation power prediction method based on error correction and fuzzy logic comprises the following steps: step 1, acquiring historical photovoltaic power generation power data and historical meteorological data of a forecast day M days ago, and meteorological data of the forecast day; step 2, using two of the time and the time meteorological data as the input of a fuzzy controller, defining the output of the fuzzy controller as a cloud amount coefficient of the time, and step 3, calculating an error correction factor according to a photovoltaic power generation power predicted value and a photovoltaic power generation power true value; step 4, taking meteorological historical data which are not used for calculating the cloud amount coefficient, the cloud amount coefficient and the error correction factor as the input of a neural network, and taking the photovoltaic power generation power predicted value as the output to train the neural network; and 5, predicting the photovoltaic power generation power through the neural network trained in the step 4 by using meteorological data and time data of the day of the prediction day.

Description

Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic
Technical Field
The invention belongs to the field of photovoltaic power generation, and particularly relates to a photovoltaic power generation power prediction method and system based on a neural network.
Background
At present, the traditional coal energy is increasingly exhausted, the price of petroleum is continuously increased, and meanwhile, people pay more attention to environmental protection, so that people have urgent needs for renewable energy. Photovoltaic power generation is to convert solar energy into electric energy, and solar energy is clean, environment-friendly and renewable clean energy. Under the condition of conventional energy shortage at present, the development of the photovoltaic industry can prevent people from depending on non-renewable energy such as petroleum, coal and the like, so that the effects of maintaining ecological balance and adjusting energy structures are achieved.
In view of the current development situation of the global photovoltaic power generation industry, due to the increasing importance of the world countries on the sustainable development concept, the scale of global photovoltaic power generation is rapidly expanding. With the continuous development of electric power technology, the cost of photovoltaic power generation is remarkably reduced, and the price of photovoltaic power generation products is also continuously reduced. At present, photovoltaic power generation projects are actively promoted in countries in many regions in the world, more and more investors participate in the photovoltaic market, and the global photovoltaic market is developing towards diversification. From the overseas market loading perspective, there are an increasing number of projects loading in excess of one billion watts per year. The competitiveness of photovoltaic power generation in the market is gradually improved, and the photovoltaic power generation is likely to become the most popular new energy technology in the future. One of the key problems limiting the development of photovoltaic power generation at present is the problem of predicting the power of photovoltaic power generation.
Firstly, the accurate prediction of the photovoltaic power can improve the stability of the power grid and increase the photoelectric capacity of the power grid. The photovoltaic power generation has intermittence, randomness and fluctuation, so that a series of problems are brought to the safe operation of a power grid, and the traditional method of a power grid dispatching department can only adopt the action of pulling a gate and limiting the power. With the increase of the proportion of the power structure of the power grid of the photovoltaic power station, a photovoltaic power prediction system becomes more important, the more accurate the photovoltaic power prediction is, the smaller the influence of the photovoltaic grid connection on the safe operation of the power grid is, and the scheduling plan of various power supplies can be effectively made by a power grid scheduling department.
And secondly, the photovoltaic power station is helped to reduce economic loss caused by power limiting, and the operation management efficiency of the photovoltaic power station is improved. The more accurate the photovoltaic power prediction is, the more the photovoltaic power is, the less the photovoltaic power limitation is, so that the sunlight absorption capacity of the power grid is greatly improved, the economic loss of photovoltaic owners caused by power limitation is reduced, and the investment return rate of photovoltaic power stations is increased.
Disclosure of Invention
In order to solve the defects in the prior art, the invention aims to provide a method for predicting the power based on an artificial neural network, adding an error correction factor and a fuzzy preprocessing method and predicting the photovoltaic output power more accurately.
The invention adopts the following technical scheme. A photovoltaic power generation power prediction method based on error correction and fuzzy logic comprises the following steps:
step 1, acquiring historical photovoltaic power generation power data and historical meteorological data of a forecast day M days ago, and meteorological data of the forecast day;
step 2, using two of the time and the time meteorological data as the input of the fuzzy controller, defining the output of the fuzzy controller as the cloud cover coefficient of the time,
step 3, calculating an error correction factor according to the predicted photovoltaic power generation power value and the true photovoltaic power generation power value;
step 4, taking meteorological historical data which are not used for calculating the cloud amount coefficient, the cloud amount coefficient and the error correction factor as the input of a neural network, and taking the photovoltaic power generation power predicted value as the output to train the neural network;
and 5, predicting the photovoltaic power generation power through the neural network trained in the step 4 by using meteorological data and time data of the day of the prediction day.
Preferably, in step 1, the step of predicting photovoltaic power generation power historical data and meteorological historical data M days before the day comprises: predicting photovoltaic power generation power and meteorological historical data at the j-th time of the ith day before the day, wherein i is 1, 2., M, i is 1, represents the day before the predicted day, and j is 1, 2., N and N represent the number of sampling points per day;
predicting weather data for the day includes: the meteorological data of the j-th moment of the day before the day are predicted, wherein j is 1, 2.
Preferably, the meteorological data comprises: irradiance vector Ix＝[Ix1，Ix2，...，IxN]Temperature vector Tx＝[Tx1，Tx2，...，TxN]Vector of wind speed WSx＝[WSx1，WSx2，...，WSxN]Wind direction vector WDx＝[WDx1，WDx2，...，WDxN]Air pressure vector Ax＝[Ax1，Ax2，...，AxN]Humidity vector Hx＝[Hx1，Hx2，...，HxN]Vector of rainfall Rx＝[Rx1，Rx2，...，RxN]Relative humidity vector RHx＝[RHx1，RHx2，...，RHxN]When x is i, the day before the prediction day is represented, and when x is 0, the day before the prediction day is represented.
Preferably, in step 2, the rainfall R at the j time of the ith day before the day is predictedijRelative humidity RHijAnd the sum time ij is used as input and input into a fuzzy controller to predict the cloud cover coefficient C at the j time of the ith day before the dayijAs outputs, namely:
Figure BDA0002733740490000031
in the formula:
Xfc_inan input of the fuzzy controller is represented and,
Yfc_outrepresenting the output of the fuzzy controller.
Preferably, step 2 specifically comprises: and calling a fuzzy processing toolbox in the MATLAB, using a three-input single-output control structure to fuzzify three inputs into { low, normal and high }, fuzzifying an output into {1, 2 and 3}, and setting a membership function.
Preferably, the step 2 fuzzy controller uses a fuzzy triangular membership function.
Preferably, step 3 specifically comprises: calculating an error correction factor for predicting the jth time of the ith day by the following formula to obtain the error correction factor EiThe indicated error correction factor vector for the i-th day before the prediction day,
Figure BDA0002733740490000032
in the formula:
Eijerror correction factor representing the j time of the ith day predicted, Ei＝[Ei1，Ei2，...，EiN]，
PijRepresenting the photovoltaic power generation power, P, at the jth moment of the ith day before the day to be predictedf_ijAnd the photovoltaic power generation power predicted value at the j time of the ith day before the day to be predicted is shown.
Preferably, step 4, training the neural network with historical data, with Xnet_ijRepresents the input of the neural network and,
Figure BDA0002733740490000033
with Ynet_ijThe output of the neural network is represented as,
Ynet_ij＝Pf_ij
in the formula:
Eijerror correction factor representing the j time of the ith day predicted, Ei＝[Ei1，Ei2，...，EiN]，
PijRepresenting the photovoltaic power generation power, P, at the jth moment of the ith day before the day to be predictedf_ijAnd the photovoltaic power generation power predicted value at the j time of the ith day before the day to be predicted is shown.
Preferably, the neural network uses a BP neural network model, which is expressed in the following formula,
Figure BDA0002733740490000041
in the formula:
a1βrepresents the output of the beta-th neuron of the hidden layer,
m represents the number of hidden layer neurons,
f1(s) represents a transfer function of the transfer,
s represents the intermediate variable(s) and,
wθβrepresents the connection weight of the theta input unit at the beta neuron of the hidden layer,
xθit indicates the theta-th input unit,
b1βrepresents the bias of the beta neuron of the hidden layer;
Figure BDA0002733740490000042
a2the output of the output layer is represented,
f2(s) represents a transfer function of the transfer,
wβdenotes a1βThe connection weight of (a) is set,
b2indicating the bias of the output layer.
Preferably, a Levenberg-Marquardt optimization method is used as the neural network training algorithm.
Preferably, step 5 specifically includes:
step 5.1, inputting the rainfall and relative humidity data of the forecast day into a fuzzy controller to obtain the cloud cover coefficient of the forecast day,
Figure BDA0002733740490000051
and 5.2, if the predicted day has no error predicted by the previous day, taking the default error as the input of the neural network with a value of 0.
Step 5.3, inputting the meteorological data, the cloud cover coefficient and the error correction factor of the predicted day into the trained neural network,
Figure BDA0002733740490000052
obtaining the output Y of the neural networknet_0i，
Ynet_0j＝Pf_0j
Namely, a prediction result of the predicted solar photovoltaic generating power is obtained.
The invention also provides a photovoltaic power generation power prediction system of the photovoltaic power generation power prediction method based on the error correction and the fuzzy logic, which comprises the following modules:
the data acquisition module is used for acquiring historical photovoltaic power generation power data and historical meteorological data of M days before the forecast day and the meteorological data of the forecast day;
the first data preprocessing module comprises a fuzzy controller unit, uses two of the time acquired by the data acquisition module and the time meteorological data as the input of the fuzzy controller, defines the output of the fuzzy controller as the cloud amount coefficient of the time,
the second data preprocessing module is used for calculating an error correction factor according to the photovoltaic power generation power predicted value and the photovoltaic power generation power true value acquired by the data acquisition module;
the photovoltaic power generation power prediction module is internally provided with a neural network unit, the neural network unit takes meteorological historical data which are not used for calculating a cloud coefficient, the cloud coefficient and an error correction factor as the input of the neural network, and takes a photovoltaic power generation power prediction value as the output to be obtained through training; the photovoltaic power generation power prediction module predicts the photovoltaic power generation power through the trained neural network unit by using meteorological data and time data of the day of prediction;
and the data output module is used for outputting and displaying the prediction result of the photovoltaic power generation power prediction module.
Preferably, the data acquisition module randomly selects 15 days each in each season of the year, and the number of sampling points per day is N-288.
Preferably, the second data preprocessing module includes at least one of a mean square error calculation unit, a root mean square error calculation unit, a mean absolute percentage error calculation unit, or a symmetric mean absolute percentage error calculation unit.
Preferably, the built-in neural network unit is at least one of a convolutional neural network unit, a bayesian neural network unit or a BP neural network unit.
Compared with the prior art, the method and the device have the advantages that the method and the device can be used for predicting the output power of a single photovoltaic panel and also can be used for predicting the output power of a photovoltaic station. Namely, a prediction result of the predicted solar photovoltaic generating power is obtained. The method comprises the specific processes of firstly using historical data, taking irradiance, temperature, humidity, air pressure, wind speed and wind direction as one to six inputs of a neural network input layer, inputting a seventh input as an error factor predicted in the first five minutes to carry out network correction, introducing a fuzzy preprocessing tool kit into a neural network system to search data correlation among relative humidity, rainfall and the time of the day, and classifying a cloud cover coefficient as the eighth input of the neural network. The output of the neural network is photovoltaic output power. And carrying out network training. After training is completed, the neural network can be used for more accurately predicting the photovoltaic output power.
The beneficial effects of the invention at least comprise:
1. and calculating a prediction error based on prediction data obtained in the first five minutes according to an error calculation formula, and returning the prediction error to the input layer of the neural network to be used as the input of prediction at the next moment and used as an error correction factor for correcting the neural network. The neural network can monitor the prediction error at a moment, so that the prediction at the next moment is more accurate.
2. The cloud covering amount has great correlation with irradiance, so that the correlation between a rainfall coefficient and three data of relative temperature, rainfall and time is found by taking the fuzzy logic theory into consideration and utilizing a fuzzy preprocessing tool box carried by MATLAB, the cloud coefficient is obtained and used as the input quantity of the neural network, and the prediction of the neural network on the photovoltaic power is further accurate.
Drawings
FIG. 1 is a flow chart of a photovoltaic power generation power prediction method based on error correction and fuzzy logic in accordance with the present invention;
FIG. 2 is a schematic diagram of a neural network of the error correction and fuzzy logic based photovoltaic power generation power prediction method of the present invention;
FIG. 3 is a schematic diagram of fuzzy logic of the error correction and fuzzy logic based photovoltaic power generation power prediction method of the present invention;
FIG. 4 is a fuzzy logic processing block diagram of the photovoltaic power generation power prediction method based on error correction and fuzzy logic of the present invention.
Detailed Description
The present application is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present application is not limited thereby.
As shown in fig. 1, the invention provides a photovoltaic power generation power prediction method based on error correction and fuzzy logic, which comprises the following specific steps:
step 1, acquiring historical photovoltaic power generation power data and historical meteorological data of a forecast day M days ago, and acquiring meteorological data of the forecast day.
The photovoltaic power generation power historical data and the meteorological historical data for M days before the day are predicted to comprise: predicting photovoltaic power generation power and meteorological historical data at the jth moment on the ith day before the day, wherein i is 1, 2.
Correspondingly, the weather data of the day of the forecast day comprises: the meteorological data of the j-th moment of the day before the day are predicted, wherein j is 1, 2.
In particular, the amount of the solvent to be used,
i denotes irradiance, IiRepresents the irradiance vector, I, of the predicted day I beforeijRepresents the irradiance at the jth time of the ith day, Ii＝[Ii1，Ii2，...，IiN]，I0Represents the irradiance vector for the predicted day, I0jRepresents the irradiance at the jth moment of the predicted day, I0＝[I01，I02，...，I0N]。
T represents the temperature, TiDenotes the temperature vector, T, of the i-th day before the predicted dayiiDenotes the temperature, T, at the j-th time of the i-th day before the predicted dayi＝[Ti1，Ti2，...，TiN]，T0Temperature vector, T, representing the day of the forecast day0jIndicating the temperature, T, at the jth moment of the predicted day0＝[T01，T02，...，T0N]。
WS denotes wind speed, WSiRepresenting the wind velocity vector, WS, predicted day i before dayijRepresents the predicted wind speed at the jth time of day i before the day, WSi＝[WSi1，WSi2，...，WSiN]，WS0Representing the wind velocity vector, WS, of the predicted day0jRepresenting the wind speed, WS, at the jth time of the day of the forecast0＝[WS01，WS02，...，WS0N]。
WD denotes wind direction, WDiIndicating the wind direction vector, WD, of the i-th day before the predicted dayijIndicating the predicted direction of the wind at time j on day i before day, WDi＝[WDi1，WDi2，...，WDiN]，WD0Representing wind direction vectors, WD, of the predicted day0jIndicating the wind direction at the jth moment of the predicted day, WD0＝[WD01，WD02，...，WD0N]。
A represents air pressure, AiDenotes the barometric pressure vector on day i before the predicted day, AijIndicating the predicted pressure at time j on day i before the day, Ai＝[Ai1，Ai2，...，AiN]，A0Indicating the barometric vector for the day of the forecast day, A0jIndicating the barometric pressure at the jth moment of the day predicted, A0＝[A01，A02，...，A0N]。
H denotes humidity, HiDenotes the humidity vector, H, of the i th day before the predicted dayijDenotes the humidity at the j-th time on the ith day before the predicted day, Hi＝[Hi1，Hi2，...，HiN]，H0Denotes the humidity vector of the day of the forecast day, H0jIndicating the humidity at the jth moment of the predicted day, H0＝[H01，H02，...，H0N]。
R represents rainfall, RiRepresenting the rainfall vector, R, of the i th day before the predicted dayijIndicating the predicted rainfall at the jth time of day i before the day, Ri＝[Ri1，Ri2，...，RiN]，R0Representing the rainfall vector of the predicted day, R0jIndicating the amount of rainfall at the jth moment of the predicted day, R0＝[R01，R02，...，R0N]。
RH denotes the relative humidity, RHiRelative humidity vector, RH, representing the i day before the predicted dayijIndicates the relative humidity, RH, at the j-th time of the ith day before the predicted dayi＝[RHi1，RHi2，...，RHiN]，RH0Relative humidity vector, RH, representing the day of the forecast day0jIndicating the relative humidity, RH, at the jth moment of the day of the forecast0＝[RH01，RH02，...，RH0N]。
P represents the photovoltaic power generation power, Pi represents the photovoltaic power generation power vector of the ith day before the day to be predicted, and PijRepresenting the photovoltaic power generation power, P, at the jth moment of the ith day before the day to be predictedi＝[Pi1，Pi2，...，PiN]，Pf_iRepresenting a photovoltaic power generation power prediction vector, P, of the day i before the day to be predictedf_ijRepresenting the predicted value of the photovoltaic power generation power at the j time of the ith day before the day to be predicted, Pf_i＝[Pf_i1，Pf_i2，...，Pf_iN]，Pf_0Representing the photovoltaic power generation power prediction vector, P, for the day to be predictedf_0jRepresenting the photovoltaic power generation power predicted value, P, of the jth moment of the day to be predictedf_0＝[Pf_01，Pf_02，...，Pf_0N]。
It is noted that those skilled in the art can select the type and number of the types of meteorological data at will, and the eight types of meteorological data used in the preferred embodiment of the present invention are only used as non-limiting preferred choices for predicting the photovoltaic power generation power, and those skilled in the art can use more or less meteorological data or other types of meteorological data for prediction.
According to the correlation definition, the closer the result is to 1, the higher the correlation, and vice versa. The result is positive correlation and negative correlation. And according to the calculation of various meteorological data and correlations, selecting the meteorological data with higher correlation as the input of the neural network to predict the photovoltaic power.
The correlation between the meteorological data and the photovoltaic power generation power is analyzed according to the data of a certain photovoltaic station, and the results are as follows:
weather factors Coefficient of correlation
Irradiance of 0.9840
Temperature of 0.7615
Air pressure 0.2151
Humidity -0.4918
Wind speed 0.1970
Wind direction 0.1652
It can be seen that the correlation coefficient values of different meteorological factors are different, the more the input number of the neural network is, the more the network is complex, and the longer the training time is. The selection of meteorological data is limited based on the correlation. The accuracy can be improved, and meanwhile, the network training time can be ensured.
As a preferred option, the meteorological data characteristics vary significantly over the seasons of the year. However, if the meteorological data of each day of a year are taken as samples, the data are huge, the memory is large, and the network training time is reduced, so that the span of the historical data is preferably randomly selected for 15 days in each season of the year according to the common consideration of the network precision and the training time, the number of sampling points per day is preferably N-288, that is, the data are sampled every 5min and the photovoltaic power is predicted.
Step 2, as shown in fig. 3, fuzzy preprocessing, the present invention proposes to use the complexity of fuzzy processing of the existing weather data input. The fuzzy processing is a branch of artificial intelligence. Traditional artificial intelligence is based on "clean" rules. The fuzzy processing is used to simulate human thinking. As fuzzy logic and probability theory are proposed and studied intensively, they show more and more powerful advantages in uncertainty inference and multi-sensor information fusion.
A fuzzy pre-processing toolbox is introduced into the nervous system to look for data correlations between relative humidity, rainfall and time of day, classifying the cloud index as another input to the neural network (i 8). The fuzzy preprocessing comprehensively considers the influence of relative humidity, rainfall and time on irradiance, simplifies the input of a neural network, and simultaneously obtains the relation between the common relation among the relative humidity, the rainfall and the time and the irradiance more accurately.
Three input variables selected: and (3) selecting a triangular membership function for the three variables, carrying out fuzzy partition according to the corresponding maximum and minimum values in the sample data, wherein each partition corresponds to a fuzzy subset. 3 fuzzy language variable values are taken for humidity, rainfall and time: low, normal, high. The output of the meteorological factor after fuzzification processing is also a triangular membership function, and 3 fuzzy language variable values are selected: low, normal, high.
As shown in fig. 4, more specifically, the cloud coverage has a large correlation with irradiance, so that the correlation between the rainfall coefficient and three data of relative temperature, rainfall and time is found by using a fuzzy preprocessing toolbox of MATLAB in consideration of fuzzy logic theory.
The fuzzy processing toolbox is first invoked using the fuzzy command, first selecting (Add Variable) to implement a three-input single-output control structure. And step two, fuzzifying input and output according to the number of the divided sets, fuzzifying three inputs into { low, normal and high }, fuzzifying the output into {1, 2 and 3}, and meanwhile, setting a triangular membership Function in a (Member Function edition) window.
The specific expression of the fuzzified triangular membership function is as follows:
Figure BDA0002733740490000101
definition C denotes the cloud coefficient, CiRepresenting the cloud coefficient vector, C, of the i-th day before the prediction dayijRepresenting the cloud cover coefficient at the j time of the ith day before the predicted day, Ci＝[Ci1，Ci2，...，GiN]，C0Representing the cloud coefficient vector, C, for the day of the forecast day0jRepresenting the cloud cover coefficient, C, at the jth moment of the predicted day0＝[C01，C02，...，C0N]；
To predict the rainfall R at the jth moment on the ith day before the dayijRelative humidity RHijAnd the sum time ij is used as input and input into a fuzzy controller to predict the cloud cover coefficient C at the j time of the ith day before the dayijAs outputs, namely:
Figure BDA0002733740490000111
in the formula:
Xfc_inan input of the fuzzy controller is represented and,
Yfc_outrepresenting the output of the fuzzy controller.
And step 3, calculating an error correction factor,
the definition E denotes an error correction factor,Eirepresenting the error correction factor vector for the i day before the predicted day, EijError correction factor representing the j time of the ith day predicted, Ei＝[Ei1，Ei2，...，EiN]，E0Error correction factor vector representing the day of the predicted day, E0jError correction factor representing the jth moment of the predicted day, E0＝[E01，E02，…，E0N]；
An error correction factor for predicting the jth time on the ith day before the day is calculated by the following formula,
Figure BDA0002733740490000112
it should be noted that one skilled in the art may arbitrarily select at least one of MSE (Mean Square Error), RMSE (Root Mean Square Error), MAE (Mean Absolute Error), MAPE (Mean Absolute Percentage Error), or SMAPE (Symmetric Mean Absolute Percentage Error) as the Error correction factor, and the SMAPE provided in this embodiment may be selected as the Error correction factor by one skilled in the artijBut is only one non-limiting preference.
Step 4, as shown in FIG. 2, training the neural network with historical data, and using Xnet_ijRepresents the input of the neural network and,
Figure BDA0002733740490000113
with Ynet_ijThe output of the neural network is represented as,
Ynet_ij＝Pf_ij
the neural network uses a BP neural network model, which is expressed in the following formula,
Figure BDA0002733740490000121
in the formula:
a1βrepresents the output of the beta-th neuron of the hidden layer,
m represents the number of hidden layer neurons,
f1(s) represents a transfer function of the transfer,
s represents the intermediate variable(s) and,
wθβrepresents the connection weight of the theta input unit at the beta neuron of the hidden layer,
xθit indicates the theta-th input unit,
b1βrepresents the bias of the beta neuron of the hidden layer;
Figure BDA0002733740490000122
a2the output of the output layer is represented,
f2(s) represents a transfer function of the transfer,
wβdenotes a1βThe connection weight of (a) is set,
b2indicating the bias of the output layer.
And a Levenberg-Marquardt optimization method is used as a BP neural network training algorithm.
It is noted that one skilled in the art can arbitrarily select the neural network model and the training algorithm, for example, but not limited to, various choices for the neural network, such as convolutional neural network, bayesian neural network, etc., and the training algorithm may also be a conjugate gradient method, newton method, gradient descent method, etc. The Levenberg-Marquardt optimized BP neural network presented in this example is only a preferred but non-limiting model.
And 5, predicting the photovoltaic power generation power through the trained neural network by using meteorological data and time data of the day of prediction. In particular, the amount of the solvent to be used,
step 5.1, inputting the rainfall and relative humidity data of the forecast day into a fuzzy controller to obtain the cloud cover coefficient of the forecast day,
Figure BDA0002733740490000131
and 5.2, if the predicted day has no error predicted by the previous day, taking the default error as the input of the neural network with a value of 0.
Step 5.3, inputting the meteorological data, the cloud cover coefficient and the error correction factor of the predicted day into the trained neural network,
Figure BDA0002733740490000132
obtaining the output Y of the neural networknet_0j，
Ynet_0j＝Pf_0j
Namely, a prediction result of the predicted solar photovoltaic generating power is obtained.
Compared with the prior art, the method has the advantages that the specific process of the method is that historical data is used, irradiance, temperature, humidity, air pressure, wind speed and wind direction are taken as one to six inputs of a neural network input layer, the seventh input is an error factor predicted in the first five minutes to be input into a correction network, a fuzzy preprocessing tool box is introduced into a neural network system to search data correlation among relative humidity, rainfall and the time of the day, and cloud cover coefficients are classified into the eighth input of the neural network. The output of the neural network is photovoltaic output power. And carrying out network training. After training is completed, the neural network can be used for more accurately predicting the photovoltaic output power.
The invention also provides a photovoltaic power generation power prediction system of the photovoltaic power generation power prediction method based on the error correction and the fuzzy logic, which comprises the following modules:
the data acquisition module is used for acquiring historical photovoltaic power generation power data and historical meteorological data of M days before the forecast day and the meteorological data of the forecast day;
the first data preprocessing module comprises a fuzzy controller unit, uses two of the time acquired by the data acquisition module and the time meteorological data as the input of the fuzzy controller, defines the output of the fuzzy controller as the cloud amount coefficient of the time,
the second data preprocessing module is used for calculating an error correction factor according to the photovoltaic power generation power predicted value and the photovoltaic power generation power true value acquired by the data acquisition module;
the photovoltaic power generation power prediction module is internally provided with a neural network unit, the neural network unit takes meteorological historical data which are not used for calculating a cloud coefficient, the cloud coefficient and an error correction factor as the input of the neural network, and takes a photovoltaic power generation power prediction value as the output to be obtained through training; the photovoltaic power generation power prediction module predicts the photovoltaic power generation power through the trained neural network unit by using meteorological data and time data of the day of prediction;
and the data output module is used for outputting and displaying the prediction result of the photovoltaic power generation power prediction module.
The beneficial effects of the invention at least comprise:
1. and calculating a prediction error based on prediction data obtained in the first five minutes according to an error calculation formula, and returning the prediction error to the input layer of the neural network to be used as the input of prediction at the next moment and used as an error correction factor for correcting the neural network. The neural network can monitor the prediction error at a moment, so that the prediction at the next moment is more accurate.
2. The cloud covering amount has great correlation with irradiance, so that the correlation between a rainfall coefficient and three data of relative temperature, rainfall and time is found by taking the fuzzy logic theory into consideration and utilizing a fuzzy preprocessing tool box carried by MATLAB, the cloud coefficient is obtained and used as the input quantity of the neural network, and the prediction of the neural network on the photovoltaic power is further accurate.
The present applicant has described and illustrated embodiments of the present invention in detail with reference to the accompanying drawings, but it should be understood by those skilled in the art that the above embodiments are merely preferred embodiments of the present invention, and the detailed description is only for the purpose of helping the reader to better understand the spirit of the present invention, and not for limiting the scope of the present invention, and on the contrary, any improvement or modification made based on the spirit of the present invention should fall within the scope of the present invention.

Claims (15)

1. A photovoltaic power generation power prediction method based on error correction and fuzzy logic is characterized by comprising the following steps:
step 1, acquiring historical photovoltaic power generation power data and historical meteorological data of a forecast day M days ago, and meteorological data of the forecast day;
step 2, using two of the time and the time meteorological data as the input of the fuzzy controller, defining the output of the fuzzy controller as the cloud cover coefficient of the time,
step 3, calculating an error correction factor according to the predicted photovoltaic power generation power value and the true photovoltaic power generation power value;
step 4, taking meteorological historical data which are not used for calculating the cloud amount coefficient, the cloud amount coefficient and the error correction factor as the input of a neural network, and taking the photovoltaic power generation power predicted value as the output to train the neural network;
and 5, predicting the photovoltaic power generation power through the neural network trained in the step 4 by using meteorological data and time data of the day of the prediction day.
2. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 1, wherein:
in step 1, predicting photovoltaic power generation power historical data and meteorological historical data M days before the day comprises: predicting photovoltaic power generation power and meteorological historical data at the j-th time of the ith day before the day, wherein i is 1, 2., M, i is 1, represents the day before the predicted day, and j is 1, 2., N and N represent the number of sampling points per day;
predicting weather data for the day includes: the meteorological data of the j-th moment of the day before the day are predicted, wherein j is 1, 2.
3. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 2, wherein:
the meteorological data includes: irradiance vector Ix＝[Ix1，Ix2，...，IxN]Temperature vector Tx＝[Tx1，Tx2，...，TxN]Vector of wind speed WSx＝[WSx1，WSx2，...，WSxN]Wind direction vector WDx＝[WDx1，WDx2，...，WDxN]Air pressure vector Ax＝[Ax1，Ax2，...，AxN]Humidity vector Hx＝[Hx1，Hx2，...，HxN]Vector of rainfall Rx＝[Rx1，Rx2，...，RxN]Relative humidity vector RHx＝[RHx1，RHx2，...，RHxN]When x is i, the day before the prediction day is represented, and when x is 0, the day before the prediction day is represented.
4. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 3, wherein:
in step 2, the rainfall R at the j time of the ith day before the day is predictedijRelative humidity RHijAnd the sum time ij is used as input and input into a fuzzy controller to predict the cloud cover coefficient C at the j time of the ith day before the dayijAs outputs, namely:
Figure FDA0002733740480000021
in the formula:
Xfc_inan input of the fuzzy controller is represented and,
Yfc_outrepresenting the output of the fuzzy controller.
5. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 4, wherein:
the step 2 specifically comprises the following steps: and calling a fuzzy processing toolbox in the MATLAB, using a three-input single-output control structure to fuzzify three inputs into { low, normal and high }, fuzzifying an output into {1, 2 and 3}, and setting a membership function.
6. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic according to any of claims 1-5, characterized by:
step 2, the fuzzy controller uses a fuzzy triangle membership function.
7. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic according to any of claims 2-5, characterized by:
the step 3 specifically comprises the following steps: calculating an error correction factor at the jth time of the ith day before the prediction day by the following formula to obtain an error correction factor vector of the ith day before the prediction day expressed by Ei,
Figure FDA0002733740480000022
in the formula:
Eijerror correction factor representing the j time of the ith day predicted, Ei＝[Ei1，Ei2，...，EiN]，
PijRepresenting the photovoltaic power generation power, P, at the jth moment of the ith day before the day to be predictedf_ijAnd the photovoltaic power generation power predicted value at the j time of the ith day before the day to be predicted is shown.
8. The method for photovoltaic power generation power prediction based on error correction and fuzzy logic according to any of claims 3-5, characterized in that:
step 4, training the neural network by using historical data and using Xnet_ijRepresents the input of the neural network and,
Figure FDA0002733740480000031
with Ynet_ijThe output of the neural network is represented as,
Ynet_ij＝Pf_ij
in the formula:
Eijerror correction factor representing the j time of the ith day predicted, Ei＝[Ei1，Ei2，...，EiN]，
FijRepresenting the photovoltaic power generation power, P, at the jth moment of the ith day before the day to be predictedf_ijAnd the photovoltaic power generation power predicted value at the j time of the ith day before the day to be predicted is shown.
9. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 8, wherein:
the neural network uses a BP neural network model, which is expressed in the following formula,
Figure FDA0002733740480000032
in the formula:
a1βrepresents the output of the beta-th neuron of the hidden layer,
m represents the number of hidden layer neurons,
f1(s) represents a transfer function of the transfer,
s represents the intermediate variable(s) and,
wθβrepresents the connection weight of the theta input unit at the beta neuron of the hidden layer,
xθit indicates the theta-th input unit,
b1βrepresents the bias of the beta neuron of the hidden layer;
Figure FDA0002733740480000041
a2representing outputs of output layersAnd then the mixture is discharged out of the furnace,
f2(s) represents a transfer function of the transfer,
wβdenotes a1βThe connection weight of (a) is set,
b2indicating the bias of the output layer.
10. The photovoltaic power generation power prediction method based on error correction and fuzzy logic according to claim 8 or 9, characterized in that:
a Levenberg-Marquardt optimization method is used as a neural network training algorithm.
11. The method of photovoltaic power generation power prediction based on error correction and fuzzy logic of claim 10, wherein:
the step 5 specifically comprises the following steps:
step 5.1, inputting the rainfall and relative humidity data of the forecast day into a fuzzy controller to obtain the cloud cover coefficient of the forecast day,
Figure FDA0002733740480000042
and 5.2, if the predicted day has no error predicted by the previous day, taking the default error as the input of the neural network with a value of 0.
Step 5.3, inputting the meteorological data, the cloud cover coefficient and the error correction factor of the predicted day into the trained neural network,
Figure FDA0002733740480000043
obtaining the output Y of the neural networknet_0j，
Ynet_0j＝Pf_0j
Namely, a prediction result of the predicted solar photovoltaic generating power is obtained.
12. A photovoltaic power generation power prediction system based on the error correction and fuzzy logic photovoltaic power generation power prediction method according to any one of claims 1 to 11, comprising the following modules:
the data acquisition module is used for acquiring historical photovoltaic power generation power data and historical meteorological data of M days before the forecast day and the meteorological data of the forecast day;
the first data preprocessing module comprises a fuzzy controller unit, uses two of the time acquired by the data acquisition module and the time meteorological data as the input of the fuzzy controller, defines the output of the fuzzy controller as the cloud amount coefficient of the time,
the second data preprocessing module is used for calculating an error correction factor according to the photovoltaic power generation power predicted value and the photovoltaic power generation power true value acquired by the data acquisition module;
the photovoltaic power generation power prediction module is internally provided with a neural network unit, the neural network unit takes meteorological historical data which are not used for calculating a cloud coefficient, the cloud coefficient and an error correction factor as the input of the neural network, and takes a photovoltaic power generation power prediction value as the output to be obtained through training; the photovoltaic power generation power prediction module predicts the photovoltaic power generation power through the trained neural network unit by using meteorological data and time data of the day of prediction;
and the data output module is used for outputting and displaying the prediction result of the photovoltaic power generation power prediction module.
13. The error correction and fuzzy logic based photovoltaic power generation power prediction system of claim 12, wherein:
the data acquisition module randomly selects 15 days in each season of the year, and the number of sampling points per day is 288.
14. The error correction and fuzzy logic based photovoltaic power generation power prediction system of claim 12 or 13, wherein:
the second data preprocessing module comprises at least one of a mean square error calculation unit, a root mean square error calculation unit, a mean absolute percentage error calculation unit or a symmetric mean absolute percentage error calculation unit.
15. The error correction and fuzzy logic based photovoltaic power generation power prediction system of claim 12 or 13, wherein:
the built-in neural network unit is at least one of a convolutional neural network unit, a Bayesian neural network unit or a BP neural network unit.
CN202011126378.8A 2020-10-20 2020-10-20 Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic Active CN112215428B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011126378.8A CN112215428B (en) 2020-10-20 2020-10-20 Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011126378.8A CN112215428B (en) 2020-10-20 2020-10-20 Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic

Publications (2)

Publication Number Publication Date
CN112215428A true CN112215428A (en) 2021-01-12
CN112215428B CN112215428B (en) 2022-08-26

Family

ID=74056119

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011126378.8A Active CN112215428B (en) 2020-10-20 2020-10-20 Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic

Country Status (1)

Country Link
CN (1) CN112215428B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112819239A (en) * 2021-02-19 2021-05-18 北京骑胜科技有限公司 Product quantity prediction method and device, electronic equipment and readable storage medium
CN113408785A (en) * 2021-05-20 2021-09-17 上海晨翘智能科技有限公司 Method, device, equipment and storage medium for predicting optical power
CN114825418A (en) * 2022-06-07 2022-07-29 北京信云筑科技有限责任公司 Multifunctional flexible complementary digital power supply management method
CN116596395A (en) * 2023-05-29 2023-08-15 深圳市中联信信息技术有限公司 Operation quality control platform for engineering project evaluation unit guidance and detection
CN116880580A (en) * 2023-08-21 2023-10-13 普宁市九鼎聚投新能源有限公司 But automatically regulated angle's photovoltaic board power generation system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6850274B1 (en) * 1997-07-15 2005-02-01 Silverbrook Research Pty Ltd Image texture mapping camera
CN103218674A (en) * 2013-04-07 2013-07-24 国家电网公司 Method for predicating output power of photovoltaic power generation system based on BP (Back Propagation) neural network model
CN104268638A (en) * 2014-09-11 2015-01-07 广州市香港科大霍英东研究院 Photovoltaic power generation system power predicting method of elman-based neural network
CN106961249A (en) * 2017-03-17 2017-07-18 广西大学 A kind of diagnosing failure of photovoltaic array and method for early warning
CN111260126A (en) * 2020-01-13 2020-06-09 燕山大学 Short-term photovoltaic power generation prediction method considering correlation degree of weather and meteorological factors

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6850274B1 (en) * 1997-07-15 2005-02-01 Silverbrook Research Pty Ltd Image texture mapping camera
CN103218674A (en) * 2013-04-07 2013-07-24 国家电网公司 Method for predicating output power of photovoltaic power generation system based on BP (Back Propagation) neural network model
CN104268638A (en) * 2014-09-11 2015-01-07 广州市香港科大霍英东研究院 Photovoltaic power generation system power predicting method of elman-based neural network
CN106961249A (en) * 2017-03-17 2017-07-18 广西大学 A kind of diagnosing failure of photovoltaic array and method for early warning
CN111260126A (en) * 2020-01-13 2020-06-09 燕山大学 Short-term photovoltaic power generation prediction method considering correlation degree of weather and meteorological factors

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112819239A (en) * 2021-02-19 2021-05-18 北京骑胜科技有限公司 Product quantity prediction method and device, electronic equipment and readable storage medium
CN112819239B (en) * 2021-02-19 2024-05-14 北京骑胜科技有限公司 Product quantity prediction method, device, electronic equipment and readable storage medium
CN113408785A (en) * 2021-05-20 2021-09-17 上海晨翘智能科技有限公司 Method, device, equipment and storage medium for predicting optical power
CN114825418A (en) * 2022-06-07 2022-07-29 北京信云筑科技有限责任公司 Multifunctional flexible complementary digital power supply management method
CN116596395A (en) * 2023-05-29 2023-08-15 深圳市中联信信息技术有限公司 Operation quality control platform for engineering project evaluation unit guidance and detection
CN116596395B (en) * 2023-05-29 2023-12-01 深圳市中联信信息技术有限公司 Operation quality control platform for engineering project evaluation unit guidance and detection
CN116880580A (en) * 2023-08-21 2023-10-13 普宁市九鼎聚投新能源有限公司 But automatically regulated angle's photovoltaic board power generation system

Also Published As

Publication number Publication date
CN112215428B (en) 2022-08-26

Similar Documents

Publication Publication Date Title
CN112215428B (en) Photovoltaic power generation power prediction method and system based on error correction and fuzzy logic
Li et al. Photovoltaic power forecasting with a hybrid deep learning approach
CN112186743B (en) Dynamic power system economic dispatching method based on deep reinforcement learning
Wu et al. Multi-timescale forecast of solar irradiance based on multi-task learning and echo state network approaches
CN107766990A (en) A kind of Forecasting Methodology of photovoltaic power station power generation power
Guo et al. Study on short-term photovoltaic power prediction model based on the Stacking ensemble learning
CN114970362B (en) Power grid load scheduling prediction method and system under multi-energy structure
Park et al. Multi-layer RNN-based short-term photovoltaic power forecasting using IoT dataset
Teferra et al. Fuzzy-based prediction of solar PV and wind power generation for microgrid modeling using particle swarm optimization
CN112116171A (en) Novel photovoltaic power generation power prediction method based on neural network
CN115115127A (en) Low-voltage transformer area intelligent regulation and control method and system based on deep learning
CN112101626A (en) Distributed photovoltaic power generation power prediction method and system
Widodo et al. Renewable energy power generation forecasting using deep learning method
CN113344279B (en) Resident load prediction method based on LSTM-SAM model and pooling
CN114444805A (en) Control method for smooth output of multi-photovoltaic power station shared energy storage system
Alharbi et al. Short-term solar irradiance forecasting model based on bidirectional long short-term memory deep learning
Souabi et al. Data-driven prediction models of photovoltaic energy for smart grid applications
CN108694475B (en) Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model
Alharbi et al. Short-term wind speed and temperature forecasting model based on gated recurrent unit neural networks
Lu et al. Photovoltaic power prediction under insufficient historical data based on dendrite network and coupled information analysis
CN113393119B (en) Stepped hydropower short-term scheduling decision method based on scene reduction-deep learning
CN115481788A (en) Load prediction method and system for phase change energy storage system
CN115392387A (en) Low-voltage distributed photovoltaic power generation output prediction method
Shi et al. Short-term photovoltaic power forecast based on long short-term memory network
CN115021331A (en) Clean energy-based ground water wind-solar integrated compensation method based on contribution coefficient method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant