CN113762646A - Photovoltaic short-term power intelligent prediction method and system - Google Patents
Photovoltaic short-term power intelligent prediction method and system Download PDFInfo
- Publication number
- CN113762646A CN113762646A CN202111205193.0A CN202111205193A CN113762646A CN 113762646 A CN113762646 A CN 113762646A CN 202111205193 A CN202111205193 A CN 202111205193A CN 113762646 A CN113762646 A CN 113762646A
- Authority
- CN
- China
- Prior art keywords
- photovoltaic
- term power
- prediction
- short
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 claims abstract description 39
- 238000010248 power generation Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims description 21
- 210000002569 neuron Anatomy 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 4
- 210000004205 output neuron Anatomy 0.000 claims description 3
- 230000006698 induction Effects 0.000 abstract description 16
- 239000013598 vector Substances 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000010606 normalization Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000001965 increasing effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000002945 steepest descent method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
-
- 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
-
- 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
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- 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
-
- 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
-
- 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)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Power Engineering (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Molecular Biology (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Public Health (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)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The present disclosure provides a photovoltaic short-term power intelligent prediction method, including: acquiring meteorological data in a photovoltaic power generation area; obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model; when the artificial neural network photovoltaic short-term power prediction model is trained, the total error curved surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected; according to the method, induction and statistics are performed on historical meteorological data, a mutual relation model between factors and load values in the meteorological data is established, the mutual relation between the factors and the load values is scientifically and reasonably summarized, a prediction induction model with higher precision is obtained, and therefore the accuracy of prediction induction is greatly improved.
Description
Technical Field
The disclosure belongs to the technical field of photovoltaic power generation, and particularly relates to a photovoltaic short-term power intelligent prediction method and system.
Background
The premise of planning and operation in the power system is that the prediction and induction of the short-term load value of the power system are related to the benefit and safety in the whole power system, and the load is influenced by a plurality of other factors in the prediction process; in the calculation process of the power load prediction analysis, the power load refers to the sum of electric power taken by electric equipment of an electric energy user to a power system at a certain moment, and the sum is a variable value which changes along with time, and on the premise of fully considering the working characteristics, the climatic characteristics, the regional development level, the supply form of electric energy and the like of the power system, the preference change of the power and the preference trend of the electric quantity can be predicted by applying a scientific and intelligent prediction means; likewise, the technology can also be used in the field of photovoltaic power generation of power grids.
The inventor of the present disclosure finds that there is an increasing relation between the prediction of the photovoltaic power load and the meteorological elements, however, the uncertainty of the meteorological elements itself also brings certain difficulties to the prediction, and affects the accuracy of the prediction and induction.
Disclosure of Invention
The method comprises the steps of carrying out induction and statistics on historical meteorological data, creating a mutual relation model between factors and load values in the meteorological data, scientifically and reasonably summarizing the mutual relation between the factors and the load values, and obtaining a high-precision prediction induction model, so that the accuracy of prediction induction is greatly improved.
In order to achieve the purpose, the invention is realized by the following technical scheme:
in a first aspect, the present disclosure provides a photovoltaic short-term power intelligent prediction method, including:
acquiring meteorological data in a photovoltaic power generation area;
obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
Further, the training of the artificial neural network photovoltaic short-term power prediction model comprises:
collecting learning samples through an artificial neural network structure, and collecting input samples of each neuron in a threshold value and weight value range of the last iteration of the network;
changing the threshold and the weight from the last layer to calculate the influence of each threshold and weight on the error and change the back propagation of the error;
the above two steps are continuously alternated.
Further, the gradient of the error surface is determined by a total error performance function output by the network, where the total error performance function is:
one to output a sample:
n output samples:
in the formula: e (k) is the total error performance function of the network output for the kth iteration, S2The neuron number of the second layer network, w is the connection weight, b is the threshold value, and t is the time.
Further, the photovoltaic short-term power prediction model of the artificial neural network takes one output neuron as a single output form of a parameter, and the prediction of each neuron by each neuron is summarized into a load value.
Further, when designing a training sample, a photovoltaic power load of 24 hours in the previous day is obtained in a one-hour period.
Further, the input and output variables of the training samples are normalized.
Further, the meteorological data includes temperature and wind speed.
In a second aspect, the present disclosure further provides a photovoltaic short-term power intelligent prediction system considering meteorological factors, which includes a data acquisition module and a prediction module;
the data acquisition module configured to: acquiring meteorological data in a photovoltaic power generation area;
the prediction module configured to: obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
In a third aspect, the present disclosure also provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the photovoltaic short-term power intelligent prediction method of the first aspect.
In a fourth aspect, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the steps of the photovoltaic short-term power intelligent prediction method according to the first aspect when executing the program.
Compared with the prior art, the beneficial effect of this disclosure is:
1. the method comprises the steps of establishing and optimizing an artificial neural network photovoltaic short-term power prediction model by taking temperature, wind speed index and weather factors as main input data, and establishing an RBP model considering real-time meteorological factors to solve the problem that the precision of a prediction induction result is not high;
2. in the method, the established model prediction result considering meteorological factors is compared with the prediction induction result not considering meteorological factors, and the prediction induction precision is higher when the meteorological factors are considered.
Drawings
The accompanying drawings, which form a part hereof, are included to provide a further understanding of the present embodiments, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the present embodiments and together with the description serve to explain the present embodiments without unduly limiting the present embodiments.
FIG. 1 is a plot of actual real photovoltaic power load and load prediction analysis versus line of the present disclosure;
FIG. 2 is a graph of normalized error between a predictive generalized photovoltaic power load value and an actual load value of the present disclosure;
FIG. 3 is a graph of a predictive inductive photovoltaic power load value versus an actual load value of the present disclosure;
fig. 4 is a graph of the error between the predictive generalized photovoltaic power load value and the actual load value of the present disclosure.
The specific implementation mode is as follows:
the present disclosure is further described with reference to the following drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
Example 1:
the embodiment provides an intelligent photovoltaic short-term power prediction method, which comprises the following steps:
acquiring meteorological data in a photovoltaic power generation area;
obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
The artificial neural network in the embodiment adopts a BP neural network, and the principle and the improvement are as follows:
a common algorithm of a modern artificial neural network is a BP neural network algorithm, and in order to realize the input and output mapping relation given by the network, a traditional BP neural network needs to learn and modify the threshold and the weight of a power grid; the established neural network is trained, and the training can be divided into two stages:
step 1: learning samples are collected through the network structure and input samples for each neuron are collected over the threshold and weight range of the last iteration of the network.
And 3, step 3: the magnitude and weight of the threshold are changed from the last layer to calculate the effect of each threshold and weight on the error, thereby changing the back-propagation of the error.
These two processes are continuously alternated until the goal is reached. The standard algorithm is the steepest descent method, in which the correction of the weighting values occurs on the slope of the error performance. The calculation principle is that iteration times are set for two layers of neural networks, and weight and threshold correction in each time is carried out according to the following formula:
x(k+1)=x(k)-ag(k) (1)
in the formula: x (k) is the connection weight and threshold vector between layers of the kth iteration; a is the learning speed, which is constant at the time of training;the components of the connection weights and thresholds between the layers for the kth iteration. The negative sign represents the opposite direction of the gradient, i.e. the steepest falling direction of the gradient. E (k) is the total error performance function of the network output for the kth iteration.
E (k) different expressions for different numbers of input samples, one output sample:
n output samples, then:
in the formula: s2The neuron number of the second layer network, w is a connection weight, b is a threshold, and t is time.
From the equation (2) or (3) and the transfer function of each layer, the total error surface gradient of the k-th iteration can be obtainedAnd respectively substituting the formula (1), and gradually correcting the weight value and the threshold value to reduce the total error until the total error is met.
The BP artificial neural network algorithm obtained through the analysis and the improvement of the method is an effective algorithm and has the advantages of simple structure, strong adaptability and strong fault-tolerant capability.
In this embodiment, the prediction model considering real-time meteorological factors:
the disclosure provides an improved RBP neural network algorithm, and the following needs to adopt a neural network to predict and summarize the annual photovoltaic power load of an embodiment area, and simultaneously input real-time meteorological load data into a prediction summarization model.
1) Model overall structure
The neural network may be divided into several inputs and outputs. Because the number of the basic structure neurons is relative, the training time of the point load value of the whole day can be realized by adopting the output form only by one network including the output source. The structure of the single output model is to construct multiple neural networks with multiple prediction and induction points per day. The novel output simulator has the characteristics of small network structure, high running speed and difficulty in accepting over-training. The model generates a neural network by generalizing the prediction of each neuron to each neuron into a load value in a single output form with one output neuron as a parameter.
2) Input/output vector design
And measuring and forecasting the power load of the previous day by taking one hour as a period, and obtaining the photovoltaic power load of 24 hours. In general, the points of the curve adjacent to the load between the continuations, i.e. the values between adjacent points, do not change abruptly, the data points of the load values then directly influencing the previous load point. Therefore, real-time data of 24 hours a day is converted into training sample data of the neural network for input.
For the target output vector, 24 load value data of the day to be predicted, i.e., a 24-dimensional output vector is used. After the input and output variables are obtained, they are normalized to manipulate the data to data between intervals [0, 1 ]. The normalization method has various forms, and the following formula is adopted here:
inputting training samples into 24 points of the previous day of the vector, taking the sensing of the highest and lowest temperatures of the data loading, and loading formula (5) to calculate the effective load data when the 24 points of the specified predicted day of the vector data are output. The measured load data enables the network to be trained effectively. If we want to further improve the accuracy of the network prediction and induction, it may be appropriate to increase the number of network training samples or increase the load points on the previous day to obtain more multidimensional input vectors. Thus, the performance of the model cannot be fully selected. Because the accuracy of the model affects too many models, the training time of the network is greatly increased, and redundant information is serious. In general, the number of samples is selected based on load characteristics and uniformity.
3) Input sample pre-operation
The neural network can only obtain data from visiting the training samples, so the quantity and quality of the training samples are important factors of the learning speed and the learning efficiency of the neural network, and the induction precision of prediction is greatly influenced. If training the network with the raw data would immediately result in neuron saturation, the data should be pre-processed before training the network to eliminate adverse effects that arise in a different manner than the raw data. The method is used for carrying out standardization operation on input data, distorting data with different orders of magnitude in the same range and accelerating constraint of a neural network. The output layer is then converted back to the original value by normalized inversion. There are two methods in total as follows:
the method comprises the following steps: normalized to [0, 1]
Normalization:
reverse normalization:
the method 2 comprises the following steps: normalized to [ -1, 1]
Normalization:
reverse normalization:
in the formula, xmax,For maximum and minimum values, x, of input variables in a training sample set1,y1The values before and after normalization of the input samples are respectively.
Example 2:
the embodiment carries out modeling analysis by using photovoltaic power load data of 7 months and 8 months in summer of a certain city; the influence of meteorological factors of the region is fully considered, so that the data of real-time temperature and wind speed of the region are used as input in the disclosure to establish an artificial neural network to predict the load of a day, and the created neural network model has only one hidden layer. We then use trial and error to select the number of neurons in the hidden layer. The result is gradually achieved by increasing the number of the neurons, so that the training is successful, and the error reaches the established requirement. Two types of dates were selected, weekday dates and weekend days of each week. The load values for 24 hours throughout the day were compared using the input model.
1. Weather factor input model:
the real-time temperature, wind speed data and photovoltaic power load value data at each moment are taken as input quantities to perform modeling prediction induction, and the input quantities are shown in the following table 1:
TABLE 1 Meteorological factor input model Table
Inputting these quantities into the model from the table above we can get the following figures: FIG. 1, a comparison line graph of actual real photovoltaic power load and load prediction analysis; FIG. 2 is a graph of normalized error between predicted and generalized photovoltaic load values and actual load values; FIG. 3 is a graph plotting predictive generalized photovoltaic power load values versus actual load values; fig. 4, a graph of a comparison of the error between the predictive generalized photovoltaic power load value and the actual load value.
From the comparison results of the above figures, it can be seen that the prediction accuracy of the correction model adopted in the present disclosure is much higher than that of the immediate input model. In summary, the present disclosure takes real-time temperature, wind speed data and photovoltaic power load value as main input data and a gradient descent method of adaptive machine learning speed as a main training means, and predicts and summarizes the photovoltaic power load value of a certain region in a certain city in 8 months and 1 day in summer. The method disclosed by the invention has the advantages that the higher prediction and induction accuracy is obtained through the data processing, the method is proved to have the calculation rapidness and accuracy in the actual operation process, the photovoltaic power load value influenced by the temperature and the wind speed in summer can be predicted, and the correction method is effective.
Example 3:
the embodiment provides a photovoltaic short-term power intelligent prediction system considering meteorological factors, which comprises a data acquisition module and a prediction module;
the data acquisition module configured to: acquiring meteorological data in a photovoltaic power generation area;
the prediction module configured to: obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
Example 4:
the present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the photovoltaic short-term power intelligent prediction method described in embodiment 1.
Example 5:
the embodiment provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the steps of the photovoltaic short-term power intelligent prediction method described in embodiment 1 are implemented.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and those skilled in the art can make various modifications and variations. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present embodiment should be included in the protection scope of the present embodiment.
Claims (10)
1. A photovoltaic short-term power intelligent prediction method is characterized by comprising the following steps:
acquiring meteorological data in a photovoltaic power generation area;
obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
2. The intelligent photovoltaic short-term power prediction method as claimed in claim 1, wherein the training of the artificial neural network photovoltaic short-term power prediction model comprises:
collecting learning samples through an artificial neural network structure, and collecting input samples of each neuron in a threshold value and weight value range of the last iteration of the network;
changing the threshold and the weight from the last layer to calculate the influence of each threshold and weight on the error and change the back propagation of the error;
the above two steps are continuously alternated.
3. The intelligent photovoltaic short-term power prediction method as claimed in claim 1, wherein the gradient of the error surface is determined by a total error performance function of the network output, and the total error performance function is:
one to output a sample:
n output samples:
in the formula: e (k) is the total error performance function of the network output for the kth iteration, S2The neuron number of the second layer network is, a is learning speed, w is connection weight, b is threshold value, and t is time.
4. The intelligent photovoltaic short-term power prediction method as claimed in claim 1, wherein the artificial neural network photovoltaic short-term power prediction model generalizes the prediction of each neuron by each neuron to a load value in a single output form with one output neuron as a parameter.
5. The intelligent photovoltaic short-term power prediction method as claimed in claim 1, wherein when designing the training samples, the photovoltaic power load of 24 hours of the previous day is obtained in a period of one hour.
6. The intelligent photovoltaic short-term power prediction method as claimed in claim 5, wherein the input and output variables of the training samples are normalized.
7. The intelligent photovoltaic short-term power prediction method as claimed in claim 1, wherein the meteorological data comprises temperature and wind speed.
8. A photovoltaic short-term power intelligent prediction system considering meteorological factors is characterized by comprising a data acquisition module and a prediction module;
the data acquisition module configured to: acquiring meteorological data in a photovoltaic power generation area;
the prediction module configured to: obtaining a photovoltaic power load prediction result according to the acquired meteorological data and a preset artificial neural network photovoltaic short-term power prediction model;
when the artificial neural network photovoltaic short-term power prediction model is trained, the total error surface gradient and the learning speed in each iteration are considered, and the magnitude of the weight and the magnitude of the threshold are corrected.
9. A computer-readable storage medium, on which a computer program is stored for fingerprint similarity calculation, characterized in that the program, when being executed by a processor, implements the steps of the photovoltaic short-term power intelligent prediction method according to any one of claims 1 to 7.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the photovoltaic short term power intelligent prediction method as claimed in any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111205193.0A CN113762646A (en) | 2021-10-15 | 2021-10-15 | Photovoltaic short-term power intelligent prediction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111205193.0A CN113762646A (en) | 2021-10-15 | 2021-10-15 | Photovoltaic short-term power intelligent prediction method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113762646A true CN113762646A (en) | 2021-12-07 |
Family
ID=78799677
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111205193.0A Pending CN113762646A (en) | 2021-10-15 | 2021-10-15 | Photovoltaic short-term power intelligent prediction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113762646A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116742622A (en) * | 2023-08-09 | 2023-09-12 | 山东理工职业学院 | Photovoltaic power generation-based power generation amount prediction method and system |
-
2021
- 2021-10-15 CN CN202111205193.0A patent/CN113762646A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116742622A (en) * | 2023-08-09 | 2023-09-12 | 山东理工职业学院 | Photovoltaic power generation-based power generation amount prediction method and system |
CN116742622B (en) * | 2023-08-09 | 2023-11-03 | 山东理工职业学院 | Photovoltaic power generation-based power generation amount prediction method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109659933B (en) | Electric energy quality prediction method for power distribution network with distributed power supply based on deep learning model | |
CN110472778A (en) | A kind of short-term load forecasting method based on Blending integrated study | |
CN110163429B (en) | Short-term load prediction method based on similarity day optimization screening | |
CN112990500B (en) | Transformer area line loss analysis method and system based on improved weighted gray correlation analysis | |
CN109376950A (en) | A kind of polynary Load Forecasting based on BP neural network | |
CN111460001B (en) | Power distribution network theoretical line loss rate evaluation method and system | |
CN113471982B (en) | Cloud edge cooperation and power grid privacy protection distributed power supply in-situ voltage control method | |
CN108596242A (en) | Power grid meteorology load forecasting method based on wavelet neural network and support vector machines | |
CN107909221A (en) | Power-system short-term load forecasting method based on combination neural net | |
CN114861533A (en) | Wind power ultra-short-term prediction method based on time convolution network | |
CN109978268A (en) | A kind of short-term load forecasting method, system and relevant apparatus | |
CN111459022A (en) | Device parameter adjustment method, device control apparatus, and computer-readable storage medium | |
CN116526473A (en) | Particle swarm optimization LSTM-based electrothermal load prediction method | |
CN114119273A (en) | Park comprehensive energy system non-invasive load decomposition method and system | |
CN116681152A (en) | Short-term load prediction method based on SOM-BP neural network improved Prophet model | |
CN113762646A (en) | Photovoltaic short-term power intelligent prediction method and system | |
CN111709585A (en) | Air conditioner load prediction method and device and storage medium | |
JPH04372046A (en) | Method and device for predicting demand amount | |
CN114358485A (en) | Source-load matching evaluation method, system, medium and electronic equipment | |
CN116722541A (en) | Power system load prediction method and device based on convolutional neural network | |
Matsui et al. | Peak load forecasting using analyzable structured neural network | |
CN116231665A (en) | Power load regulation and control method and device, nonvolatile storage medium and electronic equipment | |
CN115374998A (en) | Load prediction method based on attention mechanism and CNN-BiGRU | |
CN114759579A (en) | Power grid active power optimization control system, method and medium based on data driving | |
Zhao et al. | On predicting the pue with gated recurrent unit in data centers |
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 |