CN113516315B

CN113516315B - Wind power generation power interval prediction method, device and medium

Info

Publication number: CN113516315B
Application number: CN202110837335.9A
Authority: CN
Inventors: 韩华; 刘宏毅; 孙尧; 施光泽; 粟梅
Original assignee: Central South University
Current assignee: Central South University
Priority date: 2021-07-23
Filing date: 2021-07-23
Publication date: 2023-02-03
Anticipated expiration: 2041-07-23
Also published as: CN113516315A

Abstract

The embodiment of the disclosure provides a method, equipment and a medium for predicting a wind power generation power interval, which belong to the technical field of measurement and specifically comprise the following steps: obtaining an initial data set and a test data set; extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method; inputting the training data set into an attention mechanism and a gated cyclic unit neural network; training the initial prediction model based on an improved interval quality evaluation system to obtain an interval prediction model; judging whether the difference value of the evaluation indexes is smaller than a threshold value; if so, inputting the test data set into an interval prediction model to obtain a wind power prediction interval combination; if not, continuing to train the initial prediction model until the difference value is smaller than the threshold value. According to the scheme, the internal characteristics of the historical data are extracted, then the upper limit and the lower limit are further limited to form a training data set, the interval prediction model is trained, the corresponding wind power prediction interval is obtained, and the prediction efficiency, the adaptability and the prediction result accuracy are improved.

Description

Wind power generation power interval prediction method, device and medium

Technical Field

The embodiment of the disclosure relates to the technical field of measurement, in particular to a method, equipment and medium for predicting a wind power generation power interval.

Background

At present, while the industry is continuously developed, the problems of energy crisis, environmental pollution and the like continuously threaten the sustainable development of the human society. Renewable energy is receiving increasing attention from countries in the world as a viable solution to alleviate the energy crisis. Wind power generation is becoming increasingly popular in modern power systems due to its clean and pollution-free advantages. However, it has intermittency and uncertainty due to the chaotic nature of weather and regional effects, as compared to traditional energy sources. This may present some challenges to the system operator as the popularity of wind power generation continues to increase. Accurate wind power prediction has important significance on the safety, stability and economic benefit of a power grid.

The wind power prediction method can be divided into two categories, namely point prediction and interval prediction, but the point prediction can represent the uncertainty of a certain period of time but cannot represent the uncertainty of a certain specific moment. The accuracy and reliability of the point prediction result are not guaranteed. The existing interval prediction method is not good at learning the time relation in the non-stationary wind power sequence, and has large calculation amount and higher cost.

Therefore, a wind power generation power interval prediction method which is efficient and high in adaptability and prediction result accuracy is needed.

Disclosure of Invention

In view of this, embodiments of the present disclosure provide a method, an apparatus, and a medium for predicting a wind power interval, which at least partially solve the problem in the prior art that the prediction efficiency, the adaptability, and the accuracy of the prediction result are poor.

In a first aspect, an embodiment of the present disclosure provides a wind power generation power interval prediction method, including:

acquiring historical wind power data of a target area and preprocessing the historical wind power data to obtain an initial data set and a test data set;

extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set;

inputting the training data set into an attention mechanism and a gated loop unit neural network to establish an initial prediction model;

training the initial prediction model based on an improved interval quality evaluation system to obtain an interval prediction model;

judging whether the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than a threshold value;

if the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than the threshold value, inputting the test data set into the interval prediction model to obtain a wind power prediction interval combination;

and if the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is larger than or equal to the threshold value, continuing to train the initial prediction model until the difference value is smaller than the threshold value.

According to a specific implementation manner of the embodiment of the disclosure, the step of obtaining and preprocessing the historical wind power data of the target area to obtain an initial data set and a test data set includes:

extracting historical wind power data of the target area from a database;

and eliminating interference data in the historical wind power data to form the initial data set and the test data set.

According to a specific implementation manner of the embodiment of the present disclosure, the step of extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set includes:

decomposing the data in the initial data set into different modes by using the variation mode decomposition method;

sequentially dividing an operation window for each mode by using the rolling fuzzy granulation method;

all the operation windows are granulated to generate fuzzy information granules corresponding to each operation window;

and rolling and granulating the fuzzy information grains corresponding to all the operation windows to obtain the training data set.

According to a specific implementation manner of the embodiment of the present disclosure, the training data set includes hour wind power generation power data and time data, the training data set is input into an attention mechanism and a gated cyclic unit neural network, and the step of establishing an initial prediction model includes:

inputting the hourly wind power generation power data and the time data into the gate control circulation unit neural network for time sequence learning to obtain output vectors corresponding to different moments;

inputting all output vectors into an attention layer corresponding to the attention mechanism, and giving corresponding weights to different information values in the output vectors;

and establishing the initial prediction model according to all the weighted output vectors.

According to a specific implementation manner of the embodiment of the present disclosure, the step of training the initial prediction model based on the improved interval quality evaluation system to obtain the interval prediction model includes:

training the initial prediction model according to a first function, and outputting an interval quality evaluation index, wherein the interval quality evaluation index comprises a prediction interval coverage rate index and a prediction interval width index;

and obtaining the interval prediction model according to a second function and the interval quality evaluation index.

In a second aspect, an embodiment of the present disclosure further provides an electronic device, including:

at least one processor; and the number of the first and second groups,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the wind power interval prediction method of the first aspect or any implementation form of the first aspect.

In a third aspect, the disclosed embodiments also provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the wind power interval prediction method in the first aspect or any implementation manner of the first aspect.

In a fourth aspect, the disclosed embodiments also provide a computer program product, the computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions that, when executed by a computer, cause the computer to perform the wind power interval prediction method in the first aspect or any implementation manner of the first aspect.

The wind power generation power interval prediction scheme in the embodiment of the disclosure comprises the following steps: acquiring historical wind power data of a target area and preprocessing the historical wind power data to obtain an initial data set and a test data set; extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set; inputting the training data set into an attention mechanism and a gate control cycle unit neural network to establish an initial prediction model; training the initial prediction model based on an improved interval quality evaluation system to obtain an interval prediction model; judging whether the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than a threshold value; if the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than the threshold value, inputting the test data set into the interval prediction model to obtain a wind power prediction interval combination; and if the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is larger than or equal to the threshold value, continuing to train the initial prediction model until the difference value is smaller than the threshold value.

The beneficial effects of the embodiment of the disclosure are: according to the scheme, the internal features of the historical data are extracted, the upper limit and the lower limit are further limited according to the internal features to form a training data set, and the corresponding wind power prediction interval is obtained when the corresponding evaluation index is not obviously improved any more according to the training data set and the interval quality evaluation system training interval prediction model, so that the prediction efficiency, the adaptability and the prediction result accuracy are improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present disclosure, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flow chart of a wind power generation power interval prediction method according to an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of an initial prediction model related to a wind power generation power interval prediction method provided by the embodiment of the disclosure;

fig. 3 is a schematic view of an attention layer input process involved in a wind power generation power interval prediction method provided by an embodiment of the present disclosure;

fig. 4 is a schematic structural diagram of a gated cyclic unit neural network involved in a wind power generation power interval prediction method according to an embodiment of the present disclosure;

fig. 5 is an overall framework diagram related to a wind power generation power interval prediction method provided in an embodiment of the present disclosure;

FIG. 6 is a schematic diagram of historical wind power generation data related to a wind power generation power interval prediction method provided by an embodiment of the present disclosure;

fig. 7 is a schematic diagram of a decomposition result related to a wind power generation power interval prediction method according to an embodiment of the present disclosure;

fig. 8 is a schematic diagram of a granulation result related to a wind power generation power interval prediction method provided by an embodiment of the present disclosure;

fig. 9 is a schematic diagram of a decomposition result related to a wind power generation power interval prediction method according to an embodiment of the present disclosure;

fig. 10 is a schematic view of an electronic device provided in an embodiment of the disclosure.

Detailed Description

The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

The embodiments of the present disclosure are described below with specific examples, and other advantages and effects of the present disclosure will be readily apparent to those skilled in the art from the disclosure in the specification. It is to be understood that the described embodiments are merely illustrative of some, and not restrictive, of the embodiments of the disclosure. The disclosure may be embodied or carried out in various other specific embodiments, and various modifications and changes may be made in the details within the description without departing from the spirit of the disclosure. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.

It is noted that various aspects of the embodiments are described below within the scope of the appended claims. It should be apparent that the aspects described herein may be embodied in a wide variety of forms and that any specific structure and/or function described herein is merely illustrative. Based on the disclosure, one skilled in the art should appreciate that one aspect described herein may be implemented independently of any other aspects and that two or more of these aspects may be combined in various ways. For example, an apparatus may be implemented and/or a method practiced using any number of the aspects set forth herein. Additionally, such an apparatus may be implemented and/or such a method may be practiced using other structure and/or functionality in addition to one or more of the aspects set forth herein.

It should be further noted that the drawings provided in the following embodiments are only schematic illustrations of the basic concepts of the present disclosure, and the drawings only show the components related to the present disclosure rather than the numbers, shapes and dimensions of the components in actual implementation, and the types, the numbers and the proportions of the components in actual implementation may be arbitrarily changed, and the layout of the components may be more complicated.

In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects may be practiced without these specific details.

The embodiment of the disclosure provides a wind power generation power interval prediction method, which can be applied to a wind power prediction process in a wind power generation scene.

Referring to fig. 1, a flow chart of a wind power generation power interval prediction method provided in the embodiment of the present disclosure is schematically illustrated. As shown in fig. 1, the method mainly comprises the following steps:

s101, acquiring historical wind power data of a target area and preprocessing the historical wind power data to obtain an initial data set and a test data set;

for example, when the future wind power generation power of the area a needs to be predicted, the historical wind power data of the area a may be obtained first, the obtained data contains more information, and the historical wind power data may be preprocessed to obtain the initial data set and the test data set.

S102, extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set;

in specific implementation, considering that the initial data set contains different wind power change trends and detail information, the internal features of the initial data set can be extracted by using a variational modal decomposition method and the rolling fuzzy granulation method, a structural granulation interval with important information for prediction is formed, and then the structural granulation interval is used as the training data set.

S103, inputting the training data set into an attention mechanism and gate control circulation unit neural network to establish an initial prediction model;

in specific implementation, the training data set can be used for training the neural network of the gated circulation unit according to the attention mechanism, the complex time sequence relation in the wind power generation historical data is learned from the training data set, and the fluctuation range of the data is captured, so that the initial prediction model which is high in prediction accuracy, easy to implement and strong in generalization capability is obtained.

S104, training the initial prediction model based on an improved interval quality evaluation system to obtain an interval prediction model;

considering that the initial prediction model established based on the attention mechanism and the gated loop unit neural network directly predicts the error adjustment of the section width and the coverage rate which may occur, and influences the prediction section quality, the initial prediction model can be trained based on an improved section quality evaluation system, the internal weight of the initial model is optimized, and the section prediction model is obtained.

S105, judging whether the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than a threshold value;

after the interval prediction model of the current training is obtained, the evaluation index corresponding to the interval prediction model may be compared with the evaluation index obtained in the previous training to obtain the difference, and then the difference is compared with the threshold to determine the next operation flow, where the threshold may be set and adjusted according to the predicted accuracy requirement.

If the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than the threshold value, executing step S106, inputting the test data set into the interval prediction model, and obtaining a wind power prediction interval combination;

in specific implementation, when the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is smaller than the threshold value, the evaluation index can be determined to be not obviously improved any more, the training can be stopped, the test data set is input into the interval prediction model, the wind power prediction interval corresponding to each mode in the test data set is calculated, then all the wind power prediction intervals are combined, the wind power prediction interval combination with high accuracy is obtained, and overfitting can be prevented.

And if the difference value between the evaluation index corresponding to the interval prediction model and the evaluation index obtained by the last training is larger than or equal to the threshold value, executing the step S107, and continuing to train the initial prediction model until the difference value is smaller than the threshold value.

In specific implementation, when it is detected whether a difference between an evaluation index corresponding to the interval prediction model and an evaluation index obtained by the last training is greater than or equal to the threshold, the evaluation index may be considered as a suboptimal solution, and a prediction model with better prediction accuracy may be obtained through training, and the initial prediction model may be continuously trained until the difference is less than the threshold.

According to the wind power generation power interval prediction method provided by the embodiment, the internal features of the historical data are extracted, the upper limit and the lower limit are further limited according to the internal features to form the training data set, the interval prediction model is trained according to the training data set and the interval quality evaluation system, when the corresponding evaluation indexes are not obviously improved any more, the corresponding wind power prediction interval is obtained, and the prediction efficiency, the adaptability and the prediction result accuracy are improved.

On the basis of the above embodiment, the step S101 of obtaining and preprocessing the historical wind power data of the target area to obtain an initial data set and a test data set includes:

extracting historical wind power data of the target area from a database;

For example, when prediction needs to be performed on future wind power generation power of the area a, historical wind power data of the area a may be obtained from the database, the obtained data contains more information, the historical wind power data may be preprocessed, the influence of interference data on prediction is considered, the calculation amount may be increased, redundant and invalid data may be removed, and the preprocessed data may be proportionally divided into the initial data set and the test data set.

Specifically, in step S102, extracting internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set, including:

in specific implementation, the wind power data can be decomposed into a plurality of modal components through the VMD based on a VMD-RFG combined feature extraction method of Variable Mode Decomposition (VMD) and Rolling Fuzzy Granulation (RFG), then a model training data set is constructed by using the RFG, internal features of the data are extracted on the premise of not changing data quantity, and a fluctuation range of the wind power data is captured.

Specifically, a variational constraint model can be constructed by utilizing the VMD and an optimal solution is solved, so that training data and verification data are decomposed into different modes, and each mode contains trend or detail information of the wind power signal. The variational constraint model constructed is as follows:

wherein, { u _k }＝{u ₁ ,...,u _k Denotes K modes resulting from decomposition, { ω _k }＝{ω ₁ ,...,ω _k Denotes the frequency center of each mode, and f denotes the wind power generation data before decomposition.

in specific implementation, after different modalities are obtained, the operation window needs to be sequentially divided for each modality by using the rolling fuzzy granulation method, for example, the sequence of the modality 1 is the sequence x = { x = ×) ₁ ,x ₂ ,...,x _n When the operation window is divided for modality 1, the operation window can be divided into small subsequences

The operation window is i =1,2, …, n, w ≧ 2.

in specific implementation, the data in the operation window is sorted from small to large and divided into two subsets:

and

then generating fuzzy information particles containing data fluctuation ranges based on the triangular membership functions:

wherein a, b, d represent the minimum, maximum and average values of the data variation in the operation window, [ w/2 ], respectively]Represents the largest integer no greater than w/2, b =1,w being even and b =2 being odd,

represents a median number within the operating window,

and

respectively representing the upper and lower bounds of the data fluctuation range in the operation window.

In specific implementation, the operation window is subjected to rolling granulation by changing i from 1 to n until all the operation windows are granulated. Finally, a structured Interval (GI) corresponding to X and including information important for prediction can be obtained, and GI can be used as a training data set of a prediction model, thereby effectively improving the quality of the prediction Interval.

Further, the training data set includes hour wind power generation power data and time data, and the step S103 includes inputting the training data set into an attention mechanism and a gated cyclic unit neural network to establish an initial prediction model, including:

in specific implementation, as shown in fig. 2 and 3, an attentive Mechanism (Att) and an Att-GRU interval prediction model of a Gated Recurrent Unit neural network (GRU) may be based on, and the GRU learns a complex time series relationship between historical data, and improves the influence of important information on a prediction result and the quality of a prediction interval by combining the attentive Mechanism.

Specifically, as shown in fig. 4, the GRU layer is formed by connecting a plurality of GRU units in time sequence, and the GRU unit at time t has two input vectors: input x at the current time _t And the output of the last GRU unit

Producing an output value by its internal learning process

This output is not only the input to the next attention layer, but also the input to the next GRU unit. The inside calculation process of the GRU layer is as follows:

r _t ＝σ(W _r ·[h _t-1 ,x _t ])

z _t ＝σ(W _z ·[h _t-1 ,x _t ])

wherein, W _r ，W _z ，W _h Is a weight matrix corresponding to the network activation function,

is a candidate matrix containing all vectors, rt and z, that may be output as GRU layers _t Outputs of the GRU internal reset gate and the refresh gate are respectively, and sigma and phi are sigmoid and tanh activation functions respectively.

in specific implementation, the attention layer determines the importance of information contained in the prediction input from the GRU, and gives different weights to highlight the influence of important history data on the prediction result. The weight calculation process is as follows:

β _t ＝q tanh(ωh _t +b)

wherein, beta _t Is represented by the formula _t Corresponding attention probability values, q and omega are weight coefficients, b is a bias coefficient, the output A of the attention layer _t This corresponds to the prediction interval at time t.

And establishing the initial prediction model according to all the output vectors with the weights.

And after the output vector corresponding to each mode is endowed with weight, establishing the initial prediction model according to all the output vectors.

On the basis of the foregoing embodiment, in step S104, the training of the initial prediction model based on the improved interval quality evaluation system to obtain an interval prediction model includes:

In specific implementation, a sigmoid function is introduced as the first function, a more strict prediction interval coverage rate evaluation index PINCP and a prediction interval width evaluation index PINCW are defined, error calculation of the evaluation indexes is avoided, the quality of the constructed prediction interval can be comprehensively evaluated, and the performance of a prediction model is improved.

Wherein, theta _i Is a Boolean variable indicating whether the data point is covered by the prediction section, yi is historical power generation data at the time of i, U _i And L _i Respectively representing the upper and lower bounds, s, of their corresponding prediction intervals>0 is the scaling factor and ☉ is the element level multiplication.

The coverage rate index is preferentially met in the optimization process, the prediction interval is in accordance with the change trend of real data as much as possible, the coverage rate punishment item and the midpoint deviation constraint based on the maximum likelihood estimation principle are defined, an improved interval prediction model training objective function SCWC can be introduced as the second function, the problem of gradient algorithm compatibility in the classic interval prediction model training objective function is effectively solved, the training complexity is reduced, a reasonable compromise is established between the reliability (coverage probability) and the accuracy (interval width) of probability information, and then the interval prediction model is formed.

RELU＝max(0,PINCP-(1-α))

wherein-logN _θ And the coverage rate penalty item only acts when the coverage rate does not reach the expected value, MID is a midpoint deviation constraint, and lambda is a control factor for ensuring that the SCWC can adapt to the change of the data size n. And calculating PINCP, PINCW and SCWC by using the verification data, evaluating the interval of each training structure, and adjusting the interval according to the evaluation result.

In the following, the overall framework of the process of wind power generation interval prediction will be described with reference to a specific embodiment, as shown in fig. 5, the U.S. california public dataset is adopted, wind power generation data of 2017 year round per hour is taken as training data, and 8736 data points are provided, and the values of the data points are different from 0 to 4914, as shown in fig. 6. The initial data set consisted of data from the previous 10 months. The remaining 2 months were used to test the final performance of the interval prediction model.

As shown in fig. 7, the decomposition result of the VMD is mainly composed of high frequency components and low frequency components, and the high frequency components will have a major influence on the prediction. The amplitude fluctuation of the mode 1 is large, and the change trend of the data is reflected. Mode 2 has some periodicity but still fluctuates. The amplitude variation of the modes 3 and 4 is small, reflecting the data high frequency region variation. This difference helps to further process the data.

After VMD decomposition, rolling time granulation is performed for each mode to generate an initial Prediction Interval (PI). Fig. 8 is the granulation results for mode 1. The initial interval can completely cover all data, but most points are located on the boundary line of the interval, and the interval is wide, which obviously cannot meet the actual requirement. Therefore, a correction by the SCWC based training procedure is required.

PI construction results of test data using the wind power generation power prediction method of the present disclosure are shown in fig. 9 and the following table. It can be seen from fig. 9 that most of the test data is covered by the constructed prediction interval. In addition, the upper and lower limits of the prediction interval have similar trends with the test data. It is shown that the proposed prediction method can track the uncertainty of the wind power. To eliminate the accidental effect of a single experimental result, the experiment was repeated five times. In addition, a comprehensive evaluation index CWC in a Lower Upper Bound Estimation (LUBE) method is also calculated for comparison. The following table lists the PI evaluation index and its median value for each experiment.

From the table, we can conclude that the experimental results of the wind power interval prediction method provided by the present embodiment are highly consistent. The constructed PI can cover the test data with high probability. The prediction interval width conveys information about the prediction accuracy. Wider intervals represent a high degree of uncertainty associated with the prediction and therefore should be more cautious at the time of decision making. In contrast, if the interval width is as narrow as in the wind power interval prediction method provided in the present embodiment, it indicates that the constructed PI can cope well with the uncertainty of the data and give a greater confidence in the decision. The SCWC is 56.29% less than the CWC, which shows that the wind power generation interval prediction method provided by the embodiment can achieve a good balance between the coverage area and the interval width.

Referring to fig. 10, an embodiment of the present disclosure also provides an electronic device 10, including: 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 wind power interval prediction method of the above method embodiment.

The disclosed embodiments also provide a non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the wind power interval prediction method in the aforementioned method embodiments.

The disclosed embodiments also provide a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, cause the computer to perform the wind power interval prediction method in the aforementioned method embodiments.

Referring now to FIG. 10, a block diagram of an electronic device 10 suitable for use in implementing embodiments of the present disclosure is shown. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., car navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. The electronic device shown in fig. 10 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 10, the electronic device 10 may include a processing means (e.g., a central processing unit, a graphic processor, etc.) 11 that may perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 12 or a program loaded from a storage means 18 into a Random Access Memory (RAM) 13. In the RAM 13, various programs and data necessary for the operation of the electronic apparatus 10 are also stored. The processing device 11, the ROM 12, and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to the bus 14.

Generally, the following devices may be connected to the I/O interface 15: an input device 16 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, or the like; an output device 17 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage devices 18 including, for example, magnetic tape, hard disk, etc.; and a communication device 19. The communication means 19 may allow the electronic device 10 to communicate with other devices wirelessly or by wire to exchange data. While the figures illustrate an electronic device 10 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.

In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 19, or may be installed from the storage means 18, or may be installed from the ROM 12. The computer program, when executed by the processing means 11, performs the above-described functions defined in the method of an embodiment of the present disclosure.

It should be noted that the computer readable medium in the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In contrast, in the present disclosure, a computer readable signal medium may comprise a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

The computer readable medium may be embodied in the electronic device; or may be separate and not incorporated into the electronic device.

The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to perform the steps associated with the method embodiments.

Alternatively, the computer readable medium carries one or more programs which, when executed by the electronic device, enable the electronic device to perform the steps associated with the method embodiments.

Computer program code for carrying out operations for aspects of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units described in the embodiments of the present disclosure may be implemented by software or hardware.

It should be understood that portions of the present disclosure may be implemented in hardware, software, firmware, or a combination thereof.

The above description is only for the specific embodiments of the present disclosure, but the scope of the present disclosure is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present disclosure should be covered within the scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

1. A wind power generation power interval prediction method is characterized by comprising the following steps:

the step of extracting the internal features of the initial data set by using a variational modal decomposition method and a rolling fuzzy granulation method to form a training data set comprises the following steps:

rolling and granulating the fuzzy information grains corresponding to all the operation windows to obtain the training data set;

inputting the training data set into an attention mechanism and a gate control cycle unit neural network to establish an initial prediction model;

the step of training the initial prediction model based on the improved interval quality evaluation system to obtain an interval prediction model comprises the following steps:

obtaining the interval prediction model according to a second function and the interval quality evaluation index;

2. The method according to claim 1, wherein the step of obtaining historical wind power data of the target region and preprocessing the historical wind power data to obtain an initial data set and a test data set comprises:

extracting historical wind power data of the target area from a database;

3. The method of claim 1, wherein the training data set comprises hourly wind power data and time data, and the step of inputting the training data set into an attention mechanism and gated cyclic unit neural network to establish an initial predictive model comprises:

inputting the hourly wind power generation power data and the time data into the gated cyclic unit neural network for time sequence learning to obtain corresponding output vectors at different moments;

inputting all output vectors into attention layers corresponding to the attention mechanism, and giving corresponding weights to different information values in the output vectors;

4. An electronic device, characterized in that the electronic device comprises:

at least one processor; and the number of the first and second groups,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the wind power interval prediction method of any of the preceding claims 1-3.

5. A non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the wind power interval prediction method of any of the preceding claims 1-3.