CN113497445A - Combined prediction method and system for output of regional multi-scale new energy power station - Google Patents
Combined prediction method and system for output of regional multi-scale new energy power station Download PDFInfo
- Publication number
- CN113497445A CN113497445A CN202111051151.6A CN202111051151A CN113497445A CN 113497445 A CN113497445 A CN 113497445A CN 202111051151 A CN202111051151 A CN 202111051151A CN 113497445 A CN113497445 A CN 113497445A
- Authority
- CN
- China
- Prior art keywords
- output
- data
- prediction
- layer
- station
- 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
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/06—Wind turbines or wind farms
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
-
- 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
-
- 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
- H02J2300/28—The renewable source being wind 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a combined prediction method and a system for output of a regional multi-scale new energy power station, wherein the method comprises the following steps: responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical measured data; and inputting the acquired data into a combined output prediction model constructed based on a deep learning algorithm so as to output the predicted power of each station, wherein the combined output prediction model comprises an LSTM layer, a CNN layer and a full connection layer. The collected data are input into a many-to-many mapping model constructed based on a deep learning algorithm, the prediction power can be output, and the time-space correlation between the wind and light stations and the time sequence dependency of the prediction time are fully considered, so that the prediction precision of the model is higher, the prediction power is more accurate, the power prediction results of all stations can be obtained simultaneously, and the efficiency of the model is improved.
Description
Technical Field
The invention belongs to the technical field of power prediction of new energy power stations, and particularly relates to a combined prediction method and system for output of a regional multi-scale new energy power station.
Background
Most of wind power plants in the south are located in high mountain elevation areas and are affected by various disastrous weathers such as thunderstorms, ice accretion and the like, the weather types are complex and changeable, and the traditional mesoscale numerical weather forecast precision is reduced. The topography change of the high mountain wind power field is very different, and different wind and optical power stations are distributed in a distributed mode, so that the resources and the output characteristics of different power stations are obviously different. The randomness, the volatility and the intermittence of the output enable the space scale of the prediction modeling to be larger, a severe challenge is brought to the improvement of the output prediction precision of new energy, the pressure of peak regulation and frequency regulation of a power grid is greatly increased, the influence on the safe and stable operation of a power system is larger, and the consumption capacity of the power grid on wind power and photovoltaic is limited.
At present, the following problems mainly exist in the aspect of output prediction of a new energy power station: 1) as a main error of power prediction, the prediction accuracy of numerical weather prediction in complex regions and under complex weather conditions needs to be further improved; 2) the distribution and dispersion of new energy power stations in south causes the wind and light coupling rule to be complex, so that the prediction precision of regional output is not high, and the unified optimization decision of a dispatching department is not facilitated.
Therefore, in order to solve the problems existing in the output prediction of the new energy power station, a numerical weather forecast correction method with higher precision and a regional multi-scale new energy power station output combined prediction method and system considering the wind and optical coupling rules are urgently needed.
Disclosure of Invention
The invention provides a combined prediction method for output of a regional multi-scale new energy power station, which is used for solving at least one of the technical problems.
In a first aspect, the invention provides a joint prediction method for output of a regional multi-scale new energy power station, which comprises the following steps: responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical measured data; inputting the collected data into a combined output prediction model constructed based on a deep learning algorithm, so as to output the predicted power of each station, wherein the combined output prediction model comprises an LSTM layer, a CNN layer and a full-connection layer, and the specific step of outputting the predicted power of each station comprises the following steps: taking the preprocessed predicted weather forecast data as input, and performing feature extraction on the input time sequence of each station through the LSTM layer to obtain the time features output by the LSTM layer; based on the preprocessed historical measured data as input, extracting spatial features in the preprocessed historical measured data through the CNN layer to enable the spatial features output by the CNN layer to be obtained; fusing the time characteristic and the space characteristic to extract a multi-characteristic output vector; and responding to the acquired multi-feature output vector, inputting the multi-feature output vector to the fully-connected layer, and enabling the fully-connected layer to simultaneously output the predicted power of each station.
In a second aspect, the present invention provides a combined prediction system for output of a regional multi-scale new energy power station, including: the system comprises an acquisition module, a storage module and a display module, wherein the acquisition module is configured to respond to acquired acquisition data and preprocess the acquisition data, and the acquisition data comprises predicted weather forecast data and historical measured data; the output module is configured to input the acquired data into a joint output prediction model constructed based on a deep learning algorithm so as to output the predicted power of each station, wherein the joint output prediction model comprises an LSTM layer, a CNN layer and a full connection layer, and the output module comprises: a first input unit, configured to take the preprocessed predicted weather forecast data as input, and perform feature extraction on the input time series of each station through the LSTM layer, so as to obtain the time features output by the LSTM layer; a second input unit, configured to extract spatial features in the preprocessed historical measured data through the CNN layer based on the preprocessed historical measured data as input, so as to obtain spatial features output by the CNN layer; a fusion module configured to fuse the temporal features and the spatial features to extract a multi-feature output vector; and the output unit is configured to respond to the acquired multi-feature output vector, input the multi-feature output vector to the fully-connected layer and enable the fully-connected layer to simultaneously output the predicted power of each station.
In a third aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the steps of a method for jointly predicting regional multi-scale new energy plant output according to any of the embodiments of the present invention.
In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, the computer program comprising program instructions which, when executed by a computer, cause the computer to perform the steps of a method for jointly predicting regional multiscale new energy power plant output according to any of the embodiments of the present invention.
According to the regional multi-scale new energy power station output joint prediction method and system, the collected data are input into the many-to-many mapping model constructed based on the deep learning algorithm, the prediction power can be output, the time-space correlation between the wind and light field stations and the time sequence dependency of the prediction time are fully considered, the prediction precision of the model is higher, the prediction power is more accurate, the power prediction results of all the field stations can be obtained simultaneously, and the efficiency of the model is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a flowchart of a method for jointly predicting output of a regional multi-scale new energy power station according to an embodiment of the present invention;
fig. 2 is a flowchart of another method for jointly predicting output of a regional multi-scale new energy power station according to an embodiment of the present invention;
FIG. 3 is a comparison graph of mean square and error results of a wind farm station according to an embodiment of the present invention;
FIG. 4 is a comparison graph of mean square and error results for a photovoltaic power plant according to an embodiment of the present invention;
fig. 5 is a block diagram of a structure of a combined prediction system for output of a regional multi-scale new energy power station according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flowchart of a method for jointly predicting output of a regional multi-scale new energy power station according to the present application is shown.
As shown in fig. 1, a joint prediction method for output of a regional multi-scale new energy power station includes the following steps:
step S101, responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical actual measurement data.
In this embodiment, after acquiring the collected data, the collected data is subjected to data cleaning based on a quartile method, and the cleaned collected data is subjected to normalization processing, where the formula is as follows:
in the formula (I), the compound is shown in the specification,
in order to be the normalized data, the data,
as a data set
The minimum value of (a) to (b),
as a data set
The maximum value of (a) is,
is the original data.
And S102, inputting the acquired data into a joint output prediction model constructed based on a deep learning algorithm, so that the predicted power of each station is output.
In the embodiment, because the wind power plant and the photovoltaic power station which are located in the same geographic area are influenced by similar meteorological conditions, certain correlation exists between the output of the wind power plant and the output of the photovoltaic power plant. However, the numerical weather forecast data of a single position point often lacks the space-time coupling relation of the actual wind and light fluctuation process in the large-scale wind and light station range, ignores the space-time coupling characteristics of wind speed and irradiance in the field group range, and has limited adaptability to complex weather conditions. Therefore, a joint output prediction model based on an LSTM-CNN deep learning algorithm can be constructed by utilizing the time-space dependency of meteorological resources of the regional wind and light field stations.
In summary, according to the method, the collected data are input into the many-to-many mapping model constructed based on the deep learning algorithm, the prediction power can be output, and the time-space correlation between the wind and light stations and the time sequence dependency of the prediction time are fully considered, so that the prediction precision of the model is higher, the prediction power is more accurate, the power prediction results of all stations can be obtained simultaneously, and the efficiency of the model is improved.
Referring to fig. 2, a flowchart of another method for jointly predicting regional multi-scale new energy power plant output according to the present application is shown. The flowchart is a flowchart mainly illustrating a further limiting step of further limiting the case where the collected data is input to the joint contribution prediction model constructed based on the deep learning algorithm in step S102 so that the predicted power of each station is output. The joint contribution prediction model comprises an LSTM layer, a CNN layer and a full connection layer.
As shown in fig. 2, in step S201, using the preprocessed predicted weather forecast data as input, performing feature extraction on the input time series of each station through the LSTM layer, so as to obtain the time features output by the LSTM layer;
in step S202, based on the preprocessed historical measured data as an input, extracting spatial features in the preprocessed historical measured data through the CNN layer, so as to obtain spatial features output by the CNN layer;
in step S203, the temporal features and the spatial features are fused to extract a multi-feature output vector, where the expression of the multi-feature output vector is calculated as:
in the formula (I), the compound is shown in the specification,
to input the ith characteristic vector of the fully-connected layer,
for the activation function, w is the weight of the input vector, b is the bias of the neuron,
is a multiplication number;
in step S204, in response to the obtained multi-feature output vector, the multi-feature output vector is input to the fully-connected layer, so that the fully-connected layer simultaneously outputs the predicted power of each station.
In the method of this embodiment, the preprocessed predicted weather forecast data is input into the LSTM layer to obtain the time series characteristics, the preprocessed historically measured data is input into the CNN layer to obtain the space series characteristics, the time series characteristics and the space series characteristics are fused to obtain the multi-characteristic output vector, and finally the multi-characteristic vector is input into the full-link layer to realize power prediction. The method fully considers the time-space correlation between the wind and light stations and the time sequence dependency of the prediction time, realizes the simultaneous prediction of the output of a plurality of wind and light stations, and can effectively improve the prediction precision.
In a specific embodiment, the precision of the combined prediction method for the output of the regional multi-scale new energy power station based on deep learning is analyzed from four aspects of root mean square error, monthly average power prediction accuracy, monthly power generation plan declaration transmission rate and assessment electric quantity by combining' operation management and assessment implementation rules of a unified power generation enterprise (hereinafter, referred to as rules) and common evaluation indexes of wind and light power prediction.
Root mean square error (
) The calculation formula is shown as formula (1), and the method is very sensitive to the reflection of extra-large or extra-small errors in a group of measurement data, so that the overall power prediction accuracy can be well reflected.
In the formula (I), the compound is shown in the specification,
for the actual power at the time t,
the power is predicted for the time instant t,
for the installed capacity of a wind farm or photovoltaic power plant,
is the number of samples.
In the detailed rules, each power station is required to predict 96 point power generation output curves every day and report the curve to a scheduling mechanism, and power prediction and assessment requirements are formulated from two angles of monthly average power prediction accuracy and monthly power generation plan reporting transmission rate. Wherein, the monthly average power prediction accuracy rate is more than 80%, and the electricity is tested for 0.1 ten thousand kilowatt hours when the monthly average power prediction accuracy rate is less than one percentage point; the planned reporting transmission rate of each power station can reach more than 90%.
Average power per day
The calculation method of the prediction accuracy is shown in the formula (2).
In the formula (I), the compound is shown in the specification,
is composed of
The actual average power of the time period,
is composed of
The predicted average power of the time period,
the number of daily assessment periods (96 points minus the assessment-free points),
is composed of
Wind farm startup capacity at time intervals.
The monthly average wind power prediction accuracy is the arithmetic mean of the daily average prediction accuracy.
The calculation method of the transmission rate of the power generation plan declaration of the wind power plant is shown as the formula (3).
In the formula (I), the compound is shown in the specification,
the transmission rate is declared for the wind farm power generation plan,
the number of days of successful transmission for a month,
the days of the monthly calendar.
The method is proposed according to a wind power prediction and forecast assessment method issued by the State energy agency 2012: and (4) regarding the wind power plant with the average wind power prediction accuracy rate not meeting the standard in 3 assessed months in 6 continuous months as a failing condition, and performing adjustment and modification in a limited period. Under the prediction model, the monthly average wind power prediction and forecast accuracy of each wind power plant reaches the standard. The feasibility and the effectiveness of the prediction model are tested only by taking a lake surrounding wind power plant and a near-lake photovoltaic power station as examples.
Table 1 shows the annual prediction result accuracy comparison of a wind farm in a certain area:
TABLE 1 annual power prediction accuracy comparison of wind farm in certain area
As can be seen from table 1, the deep learning prediction power accuracy is improved compared to the original system. The error reduction rates of 1 and 2 wind power plants are the maximum, and are respectively 38.74% and 37.32%, because the original system prediction errors of the two wind power plants are large, and the deep learning prediction power is greatly improved under the double effects of considering the regional output correlation and correcting the numerical weather forecast precision. The prediction errors of the deep learning of the 7 wind power plants and the 8 wind power plants are larger than those of the original system, because the original systems of the two wind power plants have high prediction precision and small lifting space.
The power prediction result pair of a certain wind power plant is shown in fig. 3, and it can be seen from fig. 3 that: 1) the power prediction error of the original system in the region is 19.97%, the deep learning prediction error is 16.85%, and the error reduction rate is 15.65%; 2) the prediction precision of deep learning in most months in a year is higher than that of the original system; 3) the power prediction error of the original system exceeds 20% in 3 months in one year, and the power prediction error of deep learning only exceeds 20% in 1 month, which shows that the deep learning can well dig the mapping relation between the wind speed data and the measured power data, improve the prediction precision and reduce the examination cost.
The photovoltaic power station adopts a deep learning method to predict the power, and compares and analyzes the power with the annual prediction result precision of the original system, and the calculation result is shown in table 2.
TABLE 2 annual power prediction accuracy comparison for photovoltaic power station in certain area
As can be seen from table 2: 1) compared with the original system, the deep learning power prediction precision is improved to a certain extent; 2) the prediction precision of deep learning in the near lake region is greater than that of the station; 3) the error reduction rate ((original system error-deep learning error)/original system error) of the No. 1 photovoltaic power station is the largest and reaches 43.39%, firstly, the accuracy is greatly improved after the numerical weather forecast is corrected, and the power prediction accuracy is further improved; and secondly, a deep learning power prediction model is adopted, so that the power prediction precision can be effectively improved. Similarly, numbers 2 and 3 are also true.
The power prediction result ratio of a certain photovoltaic power station is shown in fig. 4, and as can be seen from fig. 4: 1) the error of the original system power prediction is 13.290%, the error of the deep learning prediction is 10.401%, and the error is reduced by 21.750% as a whole; 2) in a whole, the prediction error of the regional station is larger than that of deep learning prediction, and firstly, the accuracy of numerical weather forecast is improved to a certain extent after the numerical weather forecast is corrected, so that the power prediction accuracy is improved; and secondly, the power prediction accuracy can be effectively improved due to the deep learning power prediction model.
Referring to fig. 5, a block diagram of a system for jointly predicting the output of a regional multi-scale new energy power station according to the present application is shown.
As shown in fig. 5, the regional multi-scale new energy power station output joint prediction system 300 includes an acquisition module 310 and an output module 320.
The acquisition module 310 is configured to respond to acquired acquisition data, and preprocess the acquisition data, wherein the acquisition data includes predicted weather forecast data and historical measured data; an output module 320 configured to input the collected data into a joint output prediction model constructed based on a deep learning algorithm, so as to output the predicted power of each station, where the joint output prediction model includes an LSTM layer, a CNN layer, and a fully connected layer, and the output module 320 includes: a first input unit, configured to take the preprocessed predicted weather forecast data as input, and perform feature extraction on the input time series of each station through the LSTM layer, so as to obtain the time features output by the LSTM layer; a second input unit, configured to extract spatial features in the preprocessed historical measured data through the CNN layer based on the preprocessed historical measured data as input, so as to obtain spatial features output by the CNN layer; a fusion module configured to fuse the temporal features and the spatial features to extract a multi-feature output vector; and the output unit is configured to respond to the acquired multi-feature output vector, input the multi-feature output vector to the fully-connected layer and enable the fully-connected layer to simultaneously output the predicted power of each station.
It should be understood that the modules recited in fig. 5 correspond to various steps in the method described with reference to fig. 1. Thus, the operations and features described above for the method and the corresponding technical effects are also applicable to the modules in fig. 5, and are not described again here.
In other embodiments, an embodiment of the present invention further provides a computer-readable storage medium, where computer-executable instructions are stored in the computer-readable storage medium, and the computer-executable instructions may execute the combined prediction method for output of the regional multi-scale new energy power station in any of the above method embodiments;
as one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions configured to:
responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical measured data;
inputting the collected data into a combined output prediction model constructed based on a deep learning algorithm, so as to output the predicted power of each station, wherein the combined output prediction model comprises an LSTM layer, a CNN layer and a full-connection layer, and the specific step of outputting the predicted power of each station comprises the following steps:
taking the preprocessed predicted weather forecast data as input, and performing feature extraction on the input time sequence of each station through the LSTM layer to obtain the time features output by the LSTM layer;
based on the preprocessed historical measured data as input, extracting spatial features in the preprocessed historical measured data through the CNN layer to enable the spatial features output by the CNN layer to be obtained;
fusing the time characteristic and the space characteristic to extract a multi-characteristic output vector;
and responding to the acquired multi-feature output vector, inputting the multi-feature output vector to the fully-connected layer, and enabling the fully-connected layer to simultaneously output the predicted power of each station.
The computer-readable storage medium may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area can store data created according to the use of the regional multi-scale new energy power station output combined prediction system and the like. Further, the computer-readable storage medium may include high speed random access memory, and may also include memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the computer readable storage medium optionally includes memory remotely located from the processor, and the remote memory may be connected to the regional multi-scale new energy plant contribution joint prediction system over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the electronic device includes: a processor 410 and memory 420. The electronic device may further include: an input device 430 and an output device 440. The processor 410, the memory 420, the input device 430, and the output device 440 may be connected by a bus or other means, such as the bus connection in fig. 6. The memory 420 is the computer-readable storage medium described above. The processor 410 executes various functional applications and data processing of the server by running the nonvolatile software programs, instructions and modules stored in the memory 420, that is, the method for jointly predicting the output of the regional multi-scale new energy power station in the embodiment of the method is implemented. The input device 430 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the regional multi-scale new energy power plant output joint prediction system. The output device 440 may include a display device such as a display screen.
The electronic device can execute the method provided by the embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
As an embodiment, the electronic device is applied to a combined prediction system of output of a regional multi-scale new energy power station, and is used for a client, and the method includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to:
responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical measured data;
inputting the collected data into a combined output prediction model constructed based on a deep learning algorithm, so as to output the predicted power of each station, wherein the combined output prediction model comprises an LSTM layer, a CNN layer and a full-connection layer, and the specific step of outputting the predicted power of each station comprises the following steps:
taking the preprocessed predicted weather forecast data as input, and performing feature extraction on the input time sequence of each station through the LSTM layer to obtain the time features output by the LSTM layer;
based on the preprocessed historical measured data as input, extracting spatial features in the preprocessed historical measured data through the CNN layer to enable the spatial features output by the CNN layer to be obtained;
fusing the time characteristic and the space characteristic to extract a multi-characteristic output vector;
and responding to the acquired multi-feature output vector, inputting the multi-feature output vector to the fully-connected layer, and enabling the fully-connected layer to simultaneously output the predicted power of each station.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A joint prediction method for output of a regional multi-scale new energy power station is characterized by comprising the following steps:
responding to acquired data, and preprocessing the acquired data, wherein the acquired data comprises predicted weather forecast data and historical measured data;
inputting the collected data into a combined output prediction model constructed based on a deep learning algorithm, so as to output the predicted power of each station, wherein the combined output prediction model comprises an LSTM layer, a CNN layer and a full-connection layer, and the specific step of outputting the predicted power of each station comprises the following steps:
taking the preprocessed predicted weather forecast data as input, and performing feature extraction on the input time sequence of each station through the LSTM layer to obtain the time features output by the LSTM layer;
based on the preprocessed historical measured data as input, extracting spatial features in the preprocessed historical measured data through the CNN layer to enable the spatial features output by the CNN layer to be obtained;
fusing the time characteristic and the space characteristic to extract a multi-characteristic output vector;
and responding to the acquired multi-feature output vector, inputting the multi-feature output vector to the fully-connected layer, and enabling the fully-connected layer to simultaneously output the predicted power of each station.
2. The method of claim 1, wherein the computing of the expression of the multi-feature output vector is as follows:
in the formula (I), the compound is shown in the specification,
to input the ith characteristic vector of the fully-connected layer,
for the activation function, w is the weight of the input vector, b is the bias of the neuron,
is a multiplication sign.
3. The regional multi-scale new energy power station output joint prediction method of claim 1, wherein the historical measured data comprises historical measured wind speed data and historical irradiance data.
4. The method of claim 1, wherein preprocessing the collected data comprises data cleansing of the collected data based on a quartile method.
5. The method of claim 1, wherein preprocessing the collected data further comprises normalizing the collected data, wherein the normalization is performed according to the expression:
in the formula (I), the compound is shown in the specification,
in order to be the normalized data, the data,
as a data set
The minimum value of (a) to (b),
as a data set
The maximum value of (a) is,
is the original data.
6. The utility model provides a regional multiscale new forms of energy power station output joint prediction system which characterized in that includes:
the system comprises an acquisition module, a storage module and a display module, wherein the acquisition module is configured to respond to acquired acquisition data and preprocess the acquisition data, and the acquisition data comprises predicted weather forecast data and historical measured data;
the output module is configured to input the acquired data into a joint output prediction model constructed based on a deep learning algorithm so as to output the predicted power of each station, wherein the joint output prediction model comprises an LSTM layer, a CNN layer and a full connection layer, and the output module comprises:
a first input unit, configured to take the preprocessed predicted weather forecast data as input, and perform feature extraction on the input time series of each station through the LSTM layer, so as to obtain the time features output by the LSTM layer;
a second input unit, configured to extract spatial features in the preprocessed historical measured data through the CNN layer based on the preprocessed historical measured data as input, so as to obtain spatial features output by the CNN layer;
a fusion module configured to fuse the temporal features and the spatial features to extract a multi-feature output vector;
and the output unit is configured to respond to the acquired multi-feature output vector, input the multi-feature output vector to the fully-connected layer and enable the fully-connected layer to simultaneously output the predicted power of each station.
7. An electronic device, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any of claims 1 to 5.
8. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111051151.6A CN113497445A (en) | 2021-09-08 | 2021-09-08 | Combined prediction method and system for output of regional multi-scale new energy power station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111051151.6A CN113497445A (en) | 2021-09-08 | 2021-09-08 | Combined prediction method and system for output of regional multi-scale new energy power station |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113497445A true CN113497445A (en) | 2021-10-12 |
Family
ID=77997161
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111051151.6A Pending CN113497445A (en) | 2021-09-08 | 2021-09-08 | Combined prediction method and system for output of regional multi-scale new energy power station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113497445A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114006369A (en) * | 2021-10-28 | 2022-02-01 | 华北电力大学 | Regional wind and light station power joint prediction method and device, electronic equipment and storage medium |
| WO2024046137A1 (en) * | 2022-08-31 | 2024-03-07 | 中国长江三峡集团有限公司 | Power prediction model construction method for multi-energy combined power generation system and power prediction method for multi-energy combined power generation system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105680474A (en) * | 2016-02-22 | 2016-06-15 | 中国电力科学研究院 | Control method for restraining rapid power change of photovoltaic station based on energy storage system |
| KR102021992B1 (en) * | 2018-08-21 | 2019-09-18 | 한국과학기술정보연구원 | Apparatus for controling a trafic signal, method for controling a trafic signal, and recoding medium for controling a tarfic signal |
| CN110909919A (en) * | 2019-11-07 | 2020-03-24 | 哈尔滨工程大学 | Photovoltaic power prediction method of depth neural network model with attention mechanism fused |
| CN111260154A (en) * | 2020-02-17 | 2020-06-09 | 河海大学 | Short-term solar radiation prediction method and device based on CNN-LSTM |
-
2021
- 2021-09-08 CN CN202111051151.6A patent/CN113497445A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105680474A (en) * | 2016-02-22 | 2016-06-15 | 中国电力科学研究院 | Control method for restraining rapid power change of photovoltaic station based on energy storage system |
| KR102021992B1 (en) * | 2018-08-21 | 2019-09-18 | 한국과학기술정보연구원 | Apparatus for controling a trafic signal, method for controling a trafic signal, and recoding medium for controling a tarfic signal |
| CN110909919A (en) * | 2019-11-07 | 2020-03-24 | 哈尔滨工程大学 | Photovoltaic power prediction method of depth neural network model with attention mechanism fused |
| CN111260154A (en) * | 2020-02-17 | 2020-06-09 | 河海大学 | Short-term solar radiation prediction method and device based on CNN-LSTM |
Non-Patent Citations (4)
| Title |
|---|
| ZHENGMIN KONG ET AL.: "Multimodal Feature Extraction and Fusion Deep Neural Networks for Short-Term Load Forecasting", 《IEEE ACCESS》 * |
| 刘永前等: "基于深度学习的风光场群功率预测方法研究", 《分布式能源》 * |
| 肖玉东: "基于随机森林与卷积神经网络的电力负荷预测研究", 《计算技术与自动化》 * |
| 黄树帮等: "碳中和背景下多通道特征组合超短期风电功率预测", 《发电技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114006369A (en) * | 2021-10-28 | 2022-02-01 | 华北电力大学 | Regional wind and light station power joint prediction method and device, electronic equipment and storage medium |
| WO2024046137A1 (en) * | 2022-08-31 | 2024-03-07 | 中国长江三峡集团有限公司 | Power prediction model construction method for multi-energy combined power generation system and power prediction method for multi-energy combined power generation system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11398000B2 (en) | Methods and systems for machine-learning for prediction of grid carbon emissions | |
| Liu et al. | Prediction of short-term PV power output and uncertainty analysis | |
| US9286646B1 (en) | Method for managing centralized power generation with the aid of a digital computer | |
| Kang et al. | Big data analytics in China's electric power industry: modern information, communication technologies, and millions of smart meters | |
| Marino et al. | A microgrid energy management system based on chance-constrained stochastic optimization and big data analytics | |
| US20220294218A1 (en) | Method and system for predicting regional short-term energy power by taking weather into consideration | |
| Yang et al. | Classification and summarization of solar irradiance and power forecasting methods: A thorough review | |
| Zhao et al. | Resilient unit commitment for day-ahead market considering probabilistic impacts of hurricanes | |
| Wang et al. | A comprehensive review for wind, solar, and electrical load forecasting methods | |
| Fahiman et al. | Data-driven dynamic probabilistic reserve sizing based on dynamic Bayesian belief networks | |
| CN113497445A (en) | Combined prediction method and system for output of regional multi-scale new energy power station | |
| Liu et al. | Photovoltaic generation power prediction research based on high quality context ontology and gated recurrent neural network | |
| WO2018134759A1 (en) | Quantifying a combined effect of interdependent uncertain resources in an electrical power grid | |
| KR102513348B1 (en) | A system and method for improving estimation to maximize profit of adjusted payments | |
| Qiu et al. | Ensemble learning for power systems TTC prediction with wind farms | |
| Bui et al. | Forecasting energy output of a solar power plant in curtailment condition based on LSTM using P/GHI coefficient and validation in training process, a case study in Vietnam | |
| Cai et al. | Gray wolf optimization-based wind power load mid-long term forecasting algorithm | |
| WO2022168357A1 (en) | System for managing power generation amount and method for managing power generation amount | |
| CN116502771B (en) | Power distribution method and system based on electric power material prediction | |
| CN117473384A (en) | Power grid line safety constraint identification method, device, equipment and storage medium | |
| US11158007B2 (en) | Dynamic energy consumption and harvesting with feedback | |
| Khan et al. | DSPM: Dual sequence prediction model for efficient energy management in micro-grid | |
| Qiu et al. | Operational risk assessment on power system based on weather regionalization considering power ramp of renewable energy generation | |
| Hou et al. | A novel algorithm for multi-node load forecasting based on big data of distribution network | |
| Bu et al. | Distributed unit commitment scheduling in the future smart grid with intermittent renewable energy resources and stochastic power demands |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211012 |
|
| RJ01 | Rejection of invention patent application after publication |