CN115186884A - Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium - Google Patents
Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium Download PDFInfo
- Publication number
- CN115186884A CN115186884A CN202210749902.XA CN202210749902A CN115186884A CN 115186884 A CN115186884 A CN 115186884A CN 202210749902 A CN202210749902 A CN 202210749902A CN 115186884 A CN115186884 A CN 115186884A
- Authority
- CN
- China
- Prior art keywords
- wind power
- short
- time sequence
- wavelet
- ultra
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 29
- 230000006870 function Effects 0.000 claims description 38
- 230000015654 memory Effects 0.000 claims description 21
- 238000005457 optimization Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 4
- 238000002759 z-score normalization Methods 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 claims 3
- 230000002596 correlated effect Effects 0.000 claims 2
- 230000006403 short-term memory Effects 0.000 description 12
- 230000007787 long-term memory Effects 0.000 description 11
- 230000007246 mechanism Effects 0.000 description 10
- 230000000694 effects Effects 0.000 description 5
- 230000004913 activation Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 238000010248 power generation Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 210000004027 cell Anatomy 0.000 description 2
- 230000019771 cognition Effects 0.000 description 2
- 238000006073 displacement reaction Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000013213 extrapolation Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
- 230000004580 weight loss Effects 0.000 description 1
- 230000003936 working memory Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- 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]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Power Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method for predicting ultra-short-term wind power based on wavelet decomposition and a storage medium, wherein the method comprises the following steps: acquiring a historical wind power data time sequence corresponding to the ultra-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant; performing wavelet decomposition on the historical wind power data time sequence, and inputting each wavelet component and the meteorological data time sequence into a pre-established super-short term prediction model of the generated power of the wind power plant to obtain each wavelet component prediction value at the moment to be predicted; and performing wavelet reconstruction on each wavelet component predicted value to obtain the ultra-short-term wind power predicted value at the moment to be predicted. According to the technical scheme provided by the invention, the wind power data time sequence is subjected to wavelet decomposition, and then wavelet components are substituted into a pre-established ultra-short-term prediction model of the generated power of the wind power plant, so that an ultra-short-term wind power prediction value is obtained, and the accuracy of ultra-short-term wind power prediction is improved.
Description
Technical Field
The invention relates to the field of power prediction, in particular to an ultra-short-term wind power prediction method based on wavelet decomposition and a storage medium.
Background
While renewable energy sources such as wind power and the like provide electric energy for a power system, the inherent randomness and intermittence of the natural properties of the energy sources also bring huge influence on the safe and reliable operation of a power grid, and the effective prediction of the wind power can provide important reference for the power generation scheduling of the renewable energy sources. However, if the wind power prediction error is large, the difficulties of peak load regulation and frequency modulation of the power system are caused, which is not favorable for the stable operation of the power grid and reduces the economical efficiency of the operation of the power grid. With the rapid development of artificial intelligence technology, intelligent algorithms represented by deep learning play an important role in load prediction and renewable energy power generation prediction. The deep learning technology has a deep structure of multilayer nonlinear mapping, and can fully learn the relation and the rule between input features. Time series analysis is a typical wind power prediction method, the essential attribute is reasonable extrapolation by combining historical information, and a long-term and short-term memory network is used for wind power ultra-short-term prediction as the combination of deep learning and time series.
However, wind power is frequently fluctuated data, which usually contains a high-frequency component generated by a large amount of high-frequency random disturbance, and the high-frequency component has unpredictability, but the prior art does not consider the adverse effect of the high-frequency component on wind power prediction, so that the accuracy of ultra-short-term wind power prediction is poor.
Disclosure of Invention
The application provides an ultra-short-term wind power prediction method and a storage medium based on wavelet decomposition, so as to at least solve the technical problem of poor accuracy of ultra-short-term wind power prediction in the related technology.
The embodiment of the first aspect of the application provides an ultra-short-term wind power prediction method based on wavelet decomposition, and the method comprises the following steps:
acquiring a historical wind power data time sequence corresponding to the ultra-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant;
performing wavelet decomposition on the historical wind power data time sequence to obtain each wavelet component corresponding to the historical wind power data time sequence;
inputting the wavelet components and the historical meteorological data time sequence into a pre-established wind power plant generated power ultra-short term prediction model to obtain the wavelet component prediction values at the moment to be predicted;
and performing wavelet reconstruction on each wavelet component predicted value to obtain the ultra-short-term wind power predicted value at the moment to be predicted.
Preferably, the meteorological data is meteorological data strongly related to the wind power data.
Preferably, before the wavelet components and the historical meteorological data time series are input into a pre-established wind power plant generated power ultra-short term prediction model, the method further includes:
and carrying out standardization processing on the wavelet components and the data in the historical meteorological data time sequence.
Preferably, the establishing process of the wind farm generated power ultra-short term prediction model comprises the following steps:
acquiring a wind power data time sequence and a wind speed data time sequence of a wind power plant in a historical period;
respectively performing wavelet decomposition on the wind power data time sequence and the meteorological data time sequence to obtain a low-frequency trend component and a high-frequency component of the wind power data time sequence and a low-frequency trend component and a high-frequency component of the meteorological data time sequence;
inputting the low-frequency trend component and the high-frequency component of the wind power data time sequence and the low-frequency trend component and the high-frequency component of the meteorological data time sequence into the optimized long-short term memory network, and training by adopting a learning mode of weight sharing to obtain a trained wind power plant generated power ultra-short term prediction model.
Further, the optimization of the long-short term memory network comprises:
and performing joint optimization on the time step number, the initial learning rate, the momentum and the hidden layer unit number in the initial long-short term memory network by using a bird swarm algorithm to obtain the optimized long-short term memory network.
Further, the long-term and short-term memory network is composed of an output layer, a hidden layer and an input layer.
Further, the process of acquiring the meteorological data strongly related to the wind power data includes:
acquiring a joint probability density function between the wind power data and each meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data;
determining a mutual information value between the wind power data and each meteorological data based on a joint probability density function between the wind power data and each meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data;
and taking the meteorological data of which the mutual information value is greater than a preset mutual information threshold value as meteorological data strongly related to the wind power data.
Further, the normalizing the wavelet components and the data in the historical meteorological data time series includes:
and normalizing the data in the wavelet components and the historical meteorological data time series by adopting a z-score normalization method.
Preferably, the performing wavelet reconstruction on each wavelet component prediction value to obtain the ultra-short-term wind power prediction value at the time to be predicted includes:
and adding and inversely standardizing the wavelet component prediction values, and taking the result after inverse standardization as the ultra-short-period wind power prediction value at the moment to be predicted.
An embodiment of the second aspect of the present application proposes a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the prediction method according to the first aspect of the present application.
The technical scheme provided by the embodiment of the application at least has the following beneficial effects:
the invention provides a wavelet decomposition-based ultra-short-term wind power prediction method and a storage medium, wherein the method comprises the following steps: acquiring a historical wind power data time sequence corresponding to the ultra-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant; performing wavelet decomposition on the historical wind power data time sequence to obtain each wavelet component corresponding to the wind power data time sequence; inputting the wavelet components and the historical meteorological data time sequence into a pre-established wind power plant generated power ultra-short term prediction model to obtain the wavelet component prediction values at the moment to be predicted; and performing wavelet reconstruction on each wavelet component predicted value to obtain the ultra-short-term wind power predicted value at the moment to be predicted. According to the technical scheme provided by the invention, wavelet decomposition is carried out on the historical wind power data time sequence, and then wavelet components are substituted into a pre-established ultra-short-term prediction model of the generated power of the wind power plant, so that an ultra-short-term wind power prediction value is obtained, and the accuracy of ultra-short-term wind power prediction is improved.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flowchart of a method for ultra-short term wind power prediction based on wavelet decomposition according to an embodiment of the present application;
FIG. 2 is a waveform diagram of a wavelet function and a scale function provided according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a cell structure of a long short term memory network according to an embodiment of the present application;
FIG. 4 is a diagram of a hard sharing mechanism multitasking learning mechanism provided according to one embodiment of the present application;
fig. 5 is a detailed flowchart of a method for predicting ultra-short-term wind power based on wavelet decomposition according to an embodiment of the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
The application provides an ultra-short-term wind power prediction method and a storage medium based on wavelet decomposition, wherein the method comprises the following steps: acquiring a historical wind power data time sequence corresponding to a super-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant; performing wavelet decomposition on the historical wind power data time sequence to obtain each wavelet component corresponding to the historical wind power data time sequence; inputting the wavelet components and the historical meteorological data time sequence into a pre-established wind power plant generated power ultra-short term prediction model to obtain the wavelet component prediction values at the moment to be predicted; and performing wavelet reconstruction on each wavelet component predicted value to obtain the ultra-short-term wind power predicted value at the moment to be predicted. According to the technical scheme provided by the invention, wavelet decomposition is carried out on the historical wind power data time sequence, and then wavelet components are substituted into a pre-established ultra-short-term prediction model of the generated power of the wind power plant, so that an ultra-short-term wind power prediction value is obtained, and the accuracy of ultra-short-term wind power prediction is improved.
An ultra-short-term wind power prediction method and a storage medium based on wavelet decomposition according to embodiments of the present application are described below with reference to the accompanying drawings.
Example 1
Fig. 1 is a flowchart of a method for predicting ultra-short-term wind power based on wavelet decomposition according to an embodiment of the present disclosure, as shown in fig. 1, the method includes:
step 1: acquiring a historical wind power data time sequence corresponding to the ultra-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant; and the historical wind power data time sequence and the historical meteorological data time sequence corresponding to the wind power plant are N moments before the moment to be predicted.
It should be noted that the meteorological data is meteorological data strongly related to the wind power data.
Wherein, the acquisition process of the meteorological data strongly related to the wind power data comprises the following steps:
acquiring a joint probability density function between the wind power data and each meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data;
determining mutual information values between the wind power data and the meteorological data based on a joint probability density function between the wind power data and the meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data; wherein use is made of the formula
Determining a mutual information value between the wind power data and each meteorological data, f X (x)，f Y (y) is the edge probability density function, f X,Y (X, Y) is a joint probability density function of two input features, and MI (X; Y) is a mutual information value between X and Y; the higher the mutual information value is, the more closely the corresponding meteorological factors are related to the wind power.
And taking the meteorological data of which the mutual information value is greater than a preset mutual information threshold value as meteorological data strongly related to the wind power data.
Step 2: performing wavelet decomposition on the historical wind power data time sequence to obtain each wavelet component corresponding to the historical wind power data time sequence;
it should be noted that the time series is decomposed into a low-frequency trend component and a high-frequency component by using a Mallet algorithm.
Further, the time series is subjected to Daubechies (db) 4 wavelet decomposition to obtain a low-frequency trend component and a high-frequency component corresponding to the time series.
The db4 wavelet has no definite formula, the wavelet function is determined by the scale function and the filter coefficient, and the db4 wavelet function ψ (t) and the scale function phi (t) have waveforms as shown in fig. 2.
Assuming that X (t) is the square integrable signal, the continuous wavelet transform formula for X (t) is as follows:
in the formula (I), the compound is shown in the specification,
called as a base wavelet, alpha and beta are respectively a scale parameter and a displacement parameter of wavelet transformation. Because the problem of information redundancy exists in continuous wavelet decomposition, the calculated amount is large, and the requirement of rapidity of wind power prediction is not met, discrete wavelet transformation is carried out on X (t) in the method, so that
The available formula is as follows:
in the formula, WT X For discrete wavelet transformed signals, X (t) is the input signal, v j,k (t) is a discrete wavelet, j and k are discrete minimums respectivelyScale parameters and displacement parameters of the wave transform.
In this disclosure, before step 3, the method further includes:
and standardizing the wavelet components and the data in the historical meteorological data time sequence.
And normalizing the wavelet components and the data in the historical meteorological data time series by adopting a z-score normalization method.
Illustratively, using formulas
Normalizing, wherein X is the actual value in the time sequence, X' is the value of the normalized actual value in the time sequence,
is the average value of each data in time series, X i Is the actual value of the ith time in the time sequence, and N is the total time in the time sequence.
And step 3: inputting the wavelet components and the historical meteorological data time sequence into a pre-established wind power plant generated power ultra-short term prediction model to obtain the wavelet component prediction values at the moment to be predicted;
it should be noted that the process for establishing the ultra-short term prediction model of the generated power of the wind power plant includes:
acquiring a wind power data time sequence and a wind speed data time sequence of a wind power plant in a historical period;
respectively performing wavelet decomposition on the wind power data time sequence and the meteorological data time sequence to obtain a low-frequency trend component and a high-frequency component of the wind power data time sequence and a low-frequency trend component and a high-frequency component of the meteorological data time sequence;
inputting the low-frequency trend component and the high-frequency component of the wind power data time sequence and the low-frequency trend component and the high-frequency component of the meteorological data time sequence into the optimized long-short term memory network, and training by adopting a learning mode of weight sharing to obtain a trained wind power plant generated power ultra-short term prediction model.
Further, the optimization of the long-short term memory network comprises:
1) And performing joint optimization on the time step number, the initial learning rate, the momentum and the hidden layer unit number in the initial long-short term memory network by using a bird swarm algorithm to obtain the optimized long-short term memory network.
The long-term and short-term memory network is composed of an output layer, a hidden layer and an input layer.
By way of example, the construction of the model specifically includes:
constructing a long and short term memory network, and carrying out combined optimization on four important hyper-parameters of the long and short term memory network through a bird swarm algorithm to achieve the optimal prediction effect;
the long-term and short-term memory network can be regarded as an improved model of a common Recurrent Neural Network (RNN), the problems of gradient explosion and gradient disappearance of the common RNN are solved, long-term learning can be carried out, and historical information is fully utilized. The long-short term memory network is composed of an output layer, a hidden layer and an input layer. Unlike the RNN of a single hidden layer, the LSTM stores information in a control unit outside the normal information flow direction of the RNN, and introduces a new state unit C, the memory of the previous sequence being controlled by a forgetting gate f. The working memory is used for output, and the output gate controls the part to be written in the current memory. The input gate i is responsible for controlling the current state information h to be written to the memory cell t-1 And current input X t Part (c) of (a). Current state information h t-1 And current input X t The linear combination is jointly determined by a nonlinear activation function. The unit structure of the long-short term memory network is shown in FIG. 3; the variables in the graph may be calculated as: f. of t ＝σ(W f ·[h t-1 ,x t ]+b f )、i t ＝σ(W i ·[h t-1 ,x t ]+b i )、
h t ＝σ(W o [h t-1 ,x t ]+b 0 )*tanh(C t )、
In the formula: w f For the weight matrix used to control forgetting gate behavior, W C Output vector, W, for connecting neuron activation functions i Weight matrix connected for input gates, W o To connect with X t A weight matrix of (a); h is t-1 Is the output signal at the time of t-1 of the hidden layer, h t Outputting a signal for the hidden layer at the moment t; x is the number of t Is the output of the input layer at time t, b f ，b i ，b c And b 0 Is a bias vector; σ, tanh are activation functions; c t ，C * t ，C t-1 In a temporary state, jointly determining whether to forget the previous information; f. of t For the current memory module state value, i t The effect on the memory module state for the current data input.
Further, the optimization of the hyper-parameters by the bird swarm algorithm specifically includes:
because the four parameters of the time step number, the initial learning rate, the momentum and the hidden layer unit number in the long-short term memory network jointly influence the prediction effect, the bird swarm algorithm is adopted to carry out joint optimization on the four parameters.
The bird swarm algorithm is a swarm intelligent optimization algorithm provided by using bird foraging, warning and flying behaviors for reference, and has the advantages of fast convergence, good robustness and the like compared with traditional optimization algorithms such as a genetic algorithm and the like. Taking the root mean square error of the training data prediction result as a fitness function of the algorithm, and setting parameters of the bird swarm algorithm as follows: dimension 4, maximum iteration number 60, population number 30, individual cognition coefficient 1.6, population cognition coefficient 1.6 and migration period 3.
The bird swarm algorithm is used as a search algorithm, the four optimized hyper-parameter values have the phenomenon of small fluctuation, and in order to suppress uncertainty to the maximum extent, the average value of 5 suboptimal results is used as the parameter selection basis of the long-term and short-term memory network. And after the bird swarm optimization, assigning the parameter values to the long-term and short-term memory network. By testing the change of the iteration times from 50 to 300, after a plurality of times of trial, when the iteration times are set to 200, the training effect of the method of the embodiment is optimal.
2) Training the optimized long-term and short-term memory network by adopting a learning mode of weight sharing comprises the following steps:
and performing multi-task parallel learning by using a multi-task learning mechanism to obtain a wind power plant generated power ultra-short term prediction model.
Illustratively, the decomposed high and low frequency components D are respectively used as the input of each long and short term memory network LSTM, a learning mode of weight sharing is adopted, each component is input into the corresponding long and short term memory network for prediction, then the predicted values of each component are summed and denormalized to obtain an actual predicted value, the obtained predicted values are subjected to error analysis until the error is smaller than a preset error threshold value, and the training is stopped.
Further, a hard sharing mechanism in the multitask learning mechanism has the characteristic that multiple sub-tasks share the same feature sharing layer and feature parameters are completely the same, the problem that overfitting is not easily generated under the condition that a wind power plant generated power ultra-short term prediction model to be constructed has multiple parameters and a complex structure is solved, the generalization capability of the model is stronger, and the hard sharing mechanism multitask learning machine drawing is shown in fig. 4.
The multi-task learning mechanism comprises a set of tasks y t Where t ∈ M and the data set
Wherein N is the number of tasks, and the number of data samples in the data set defines a prediction function f t (x,Θ s ，Θ t ):x-→y t It should be noted that, for the equal weight loss function of the multi-input multi-output network model, the generalization capability of the main task is improved by adopting an auxiliary task manner, and the overall optimization loss function of the multi-task learning mechanism may be defined as follows:
in the formula: theta s ，Θ t Respectively, parameters shared in common between different tasks and parameters related to the task. MI j i For the ith task sequence of the jth component, w is the mutual information value of the corresponding component of the main task sequence j i Is the ith task weight coefficient of the j component, L j i Is the ith task loss function of the jth component, L j total The overall loss function of the multi-task learning mechanism for the j component of the present application.
In the process of processing a problem, the multi-task learning mechanism can learn and obtain auxiliary coupling information provided by other related subtasks by using the sharing layer, so that the aims of improving the output precision of the model and enhancing the generalization ability are fulfilled.
According to the method, the historical wind power time sequence is used as a main task, and the wind speed, the wind direction and other meteorological information are converted into auxiliary tasks to improve the prediction accuracy of wind power prediction.
And 4, step 4: and performing wavelet reconstruction on each wavelet component predicted value to obtain the ultra-short-term wind power predicted value at the moment to be predicted.
In an embodiment of the present disclosure, the step 4 includes:
and adding and inversely standardizing the wavelet component prediction values, and taking the result after inverse standardization as the ultra-short-period wind power prediction value at the moment to be predicted.
In the embodiment of the present disclosure, a detailed flowchart of the method is shown in fig. 5.
To sum up, according to the ultra-short-term wind power prediction method based on wavelet decomposition provided by the embodiment of the disclosure, wavelet decomposition is performed on a historical wind power data time sequence, and then wavelet components are substituted into a pre-established ultra-short-term wind power generation power prediction model of a wind power plant, so that an ultra-short-term wind power prediction value is obtained, and the accuracy of ultra-short-term wind power prediction is improved.
Example 2
In order to implement the above embodiments, the present disclosure also proposes a computer-readable storage medium.
The present embodiment provides a computer device having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method of embodiment 1.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Moreover, various embodiments or examples and features of various embodiments or examples described in this specification can be combined and combined by one skilled in the art without being mutually inconsistent.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.
Claims (10)
1. A method for predicting ultra-short-term wind power based on wavelet decomposition is characterized by comprising the following steps:
acquiring a historical wind power data time sequence corresponding to the ultra-short-term wind power to-be-predicted moment of a wind power plant and a historical meteorological data time sequence corresponding to the wind power plant;
performing wavelet decomposition on the historical wind power data time sequence to obtain each wavelet component corresponding to the historical wind power data time sequence;
inputting the wavelet components and the historical meteorological data time sequence into a pre-established wind power plant generated power ultra-short term prediction model to obtain the wavelet component prediction values at the moment to be predicted;
and performing wavelet reconstruction on the wavelet component predicted values to obtain the ultra-short-term wind power predicted value at the moment to be predicted.
2. The method of claim 1, wherein the meteorological data is meteorological data strongly correlated with the wind power data.
3. The method of claim 1, wherein before inputting the wavelet components and the historical meteorological data time series into the pre-established wind farm generated power ultra-short term prediction model, further comprising:
and standardizing the wavelet components and the data in the historical meteorological data time sequence.
4. The method of claim 1, wherein the establishing process of the wind farm generated power ultra-short term prediction model comprises:
acquiring a wind power data time sequence and a wind speed data time sequence of a wind power plant in a historical period;
respectively performing wavelet decomposition on the wind power data time sequence and the meteorological data time sequence to obtain a low-frequency trend component and a high-frequency component of the wind power data time sequence and a low-frequency trend component and a high-frequency component of the meteorological data time sequence;
inputting the low-frequency trend component and the high-frequency component of the wind power data time sequence and the low-frequency trend component and the high-frequency component of the meteorological data time sequence into the optimized long-short term memory network, and training by adopting a learning mode of weight sharing to obtain a trained wind power plant generated power ultra-short term prediction model.
5. The method of claim 4, wherein the optimization of the long-short term memory network comprises:
and performing joint optimization on the time step number, the initial learning rate, the momentum and the hidden layer unit number in the initial long-short term memory network by using a bird swarm algorithm to obtain the optimized long-short term memory network.
6. The method of claim 4, wherein the long-short term memory network is comprised of an output layer, a hidden layer, and an input layer.
7. The method of claim 2, wherein the obtaining of meteorological data strongly correlated with the historical wind power data comprises:
acquiring a joint probability density function between the wind power data and each meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data;
determining a mutual information value between the wind power data and each meteorological data based on a joint probability density function between the wind power data and each meteorological data, an edge probability density function of the wind power data and an edge probability density function of the meteorological data;
and taking the meteorological data of which the mutual information value is greater than a preset mutual information threshold value as meteorological data strongly related to the wind power data.
8. The method of claim 3, wherein normalizing the wavelet components and the data in the time series of historical meteorological data comprises:
and normalizing the data in the wavelet components and the historical meteorological data time series by adopting a z-score normalization method.
9. The method of claim 1, wherein the performing wavelet reconstruction on each wavelet component prediction value to obtain the ultra-short-term wind power prediction value at the time to be predicted comprises:
and adding and inversely standardizing the wavelet component prediction values, and taking the result after inverse standardization as the ultra-short-period wind power prediction value at the moment to be predicted.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210749902.XA CN115186884A (en) | 2022-06-29 | 2022-06-29 | Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210749902.XA CN115186884A (en) | 2022-06-29 | 2022-06-29 | Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115186884A true CN115186884A (en) | 2022-10-14 |
Family
ID=83515972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210749902.XA Pending CN115186884A (en) | 2022-06-29 | 2022-06-29 | Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115186884A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117056735A (en) * | 2023-10-13 | 2023-11-14 | 云南电投绿能科技有限公司 | Short-term wind speed prediction method, device and equipment for wind farm and storage medium |
-
2022
- 2022-06-29 CN CN202210749902.XA patent/CN115186884A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117056735A (en) * | 2023-10-13 | 2023-11-14 | 云南电投绿能科技有限公司 | Short-term wind speed prediction method, device and equipment for wind farm and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | A novel hybrid system based on multi-objective optimization for wind speed forecasting | |
| Wang et al. | A novel combined model based on hybrid optimization algorithm for electrical load forecasting | |
| CN113363998B (en) | Power distribution network voltage control method based on multi-agent deep reinforcement learning | |
| CN109255726A (en) | A kind of ultra-short term wind power prediction method of Hybrid Intelligent Technology | |
| CN116562908A (en) | Electric price prediction method based on double-layer VMD decomposition and SSA-LSTM | |
| CN115619025A (en) | Power load triangular fuzzy prediction method based on MEEMD and optimal combination integration | |
| CN115186884A (en) | Ultra-short-term wind power prediction method based on wavelet decomposition and storage medium | |
| Liu et al. | Research on wind speed prediction of wind power system based on GRU deep learning | |
| Bose | Artificial intelligence applications in renewable energy systems and smart grid–Some novel applications | |
| Niu et al. | Short term load forecasting model using support vector machine based on artificial neural network | |
| CN117132132A (en) | Photovoltaic power generation power prediction method based on meteorological data | |
| Na et al. | A novel heuristic artificial neural network model for urban computing | |
| Sen | Time series prediction based on improved deep learning | |
| Masero et al. | A fast implementation of coalitional model predictive controllers based on machine learning: Application to solar power plants | |
| CN113872213B (en) | Autonomous optimization control method and device for power distribution network voltage | |
| Whig et al. | Renewable energy optimization system using fuzzy logic | |
| Wang et al. | Transient stability preventive control based on graph convolution neural network and transfer deep reinforcement learning | |
| Zeng et al. | Simulation of fuzzy neural network algorithm in dynamic nonlinear system | |
| CN112330044A (en) | Support vector regression model based on iterative aggregation grid search algorithm | |
| Cheng et al. | Research on prediction method based on ARIMA-BP combination model | |
| CN116865248A (en) | Ultra-short-term wind power prediction method and electronic equipment | |
| Wang et al. | Transfer-Reinforcement-Learning-Based rescheduling of differential power grids considering security constraints | |
| Dragomir et al. | MLP neural network as load forecasting tool on short-term horizon | |
| Rasmani et al. | Weighted linguistic modelling based on fuzzy subsethood values | |
| Zhang et al. | A Novel Combined Model Based on Hybrid Data Decomposition, MSWOA and ENN for Short-Term Wind Speed Forecasting |
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 |