CN112686464A - Short-term wind power prediction method and device - Google Patents
Short-term wind power prediction method and device Download PDFInfo
- Publication number
- CN112686464A CN112686464A CN202110018074.8A CN202110018074A CN112686464A CN 112686464 A CN112686464 A CN 112686464A CN 202110018074 A CN202110018074 A CN 202110018074A CN 112686464 A CN112686464 A CN 112686464A
- Authority
- CN
- China
- Prior art keywords
- wind power
- short
- term
- prediction model
- value
- 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 43
- 239000002245 particle Substances 0.000 claims abstract description 105
- 230000015654 memory Effects 0.000 claims abstract description 57
- 238000012360 testing method Methods 0.000 claims abstract description 50
- 230000007787 long-term memory Effects 0.000 claims abstract description 41
- 230000006403 short-term memory Effects 0.000 claims abstract description 41
- 238000010606 normalization Methods 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims abstract description 36
- 238000005457 optimization Methods 0.000 claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 15
- 230000001133 acceleration Effects 0.000 claims description 52
- 210000002569 neuron Anatomy 0.000 claims description 28
- 238000011156 evaluation Methods 0.000 claims description 25
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000007637 random forest analysis Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 1
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000029087 digestion Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Images
Classifications
-
- 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
Abstract
The application discloses a short-term wind power prediction method and a short-term wind power prediction device, wherein the method comprises the following steps: acquiring historical wind power data; carrying out normalization processing on historical wind power data to obtain a historical wind power data set; dividing a historical wind power data set into a training set data set and a test set data set; constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer; performing particle swarm optimization on the long and short term memory network prediction model by utilizing a training set data set to determine the optimal parameters of the long and short term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters; and taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result. By adopting the scheme, the optimal parameters of the model can be quickly searched and determined, the prediction precision is improved, and the wind power can be timely and accurately predicted, so that the wind power plant can be accurately scheduled and operated.
Description
Technical Field
The application relates to the technical field of wind power prediction, in particular to a short-term wind power prediction method and device.
Background
Wind power prediction and application thereof in energy scheduling operation are the basis for promoting new energy consumption, and the research on the wind power prediction technology is helpful for weakening adverse effects on an electric power system when wind power is connected to the grid, reducing the operation cost of the electric network, improving the operation reliability of the electric power system and effectively ensuring the safety of the electric network. The short-term wind power prediction is used for economic dispatching by predicting the wind power plant power within 24-72 hours of a time scale, so that the digestion capacity can be enhanced, and the safety and stability of a power system are improved.
With the development of the neural network technology, the neural network technology has been successfully applied to the field of wind power prediction, and the neural network technology solves the problem of a static model by digging the implicit relation between input and output so as to realize wind power prediction. However, the wind power belongs to a non-stationary time sequence, and the change rule of the wind power is not only related to the current state, but also influenced by the change process of historical data. With the development of artificial intelligence, an artificial intelligence prediction algorithm has been successfully applied to the field of short-term wind power prediction, and the artificial intelligence prediction algorithm is to train and analyze monitored data through a contemporary computer technology to further establish a prediction model, which is commonly a Random Forest (RF), a Back Propagation Neural Network (BPNN), a Support Vector Machine (SVM), and the like. However, most of the current prediction models are the problems of insufficient model fitting capability and unsatisfactory training effect caused by selecting key parameters according to experience, and further the accuracy of the prediction result is influenced.
Disclosure of Invention
The application provides a short-term wind power prediction method and device, and aims to solve the problems that most of current prediction models select key parameters according to experience, so that the model fitting capability is not enough, the training effect is not ideal, and the accuracy of prediction results is affected.
In a first aspect, an embodiment of the present application provides a short-term wind power prediction method, including:
obtaining historical wind power data, wherein the historical wind power data comprises historical average wind power data;
carrying out normalization processing on the historical wind power data to obtain a historical wind power data set;
dividing the historical wind power data set into a training set data set and a test set data set;
constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer;
performing particle swarm optimization on the long-short term memory network prediction model by using the training set data set to determine the optimal parameters of the long-short term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters;
and taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
With reference to the first aspect, in an implementation manner, the historical wind power data is normalized by using the following formula:
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminRepresenting the minimum value of a sequence before normalization.
With reference to the first aspect, in one implementation manner, the long-term and short-term memory network prediction model is constructed by the following method:
the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller;
predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain a plurality of preliminary prediction results;
the output layer adopts a plurality of preliminary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain final prediction results, and finally carries out evaluation analysis on the prediction results by combining with experimental evaluation indexes.
With reference to the first aspect, in an implementation manner, the training set dataset is used to perform particle swarm optimization on the long-short term memory network prediction model, and optimal parameters of the long-short term memory network prediction model are determined; and establishing a short-term wind power prediction model according to the optimal parameters, wherein the short-term wind power prediction model comprises the following steps:
1) initializing parameters of the long-term and short-term memory network prediction model, setting respective value ranges and search ranges of the number of neurons and the learning rate, and determining the maximum iteration times, the population number, the maximum inertia weight value, the minimum inertia weight value, the initial value of a first acceleration factor, the final value of the first acceleration factor, the initial value of a second acceleration factor and the final value of the second acceleration factor;
2) according to the initialized values of the number of the neurons and the learning rate, a long-short term memory prediction model is used for carrying out model prediction on a test set data set, the average absolute percentage error of the obtained prediction result is used as the initial fitness value of the particles, and the fitness function is obtained by adopting the following formula:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) The predicted value of the wind power at the moment i is obtained;
3) taking two parameters of the number of neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formula, obtaining the optimal values of the two parameters by optimizing through a particle swarm optimization, and reestablishing an optimized long-short term memory network prediction model on the basis of the optimal values;
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax)；
wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vki id represents the velocity of the particle; x k id represents the position of the particle; p k id represents an individual local optimal solution of the particle; p k gd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factor c1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2A final value of;
4) continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
With reference to the first aspect, in an implementation manner, after obtaining the short-term wind power prediction result, the method further includes: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) And (i is 1,2, … n) is a predicted value of the wind power at the ith moment.
In a second aspect, an embodiment of the present application provides a short-term wind power prediction apparatus, including:
the historical wind power data acquisition module is used for acquiring historical wind power data, and the historical wind power data comprises historical average wind power data;
the normalization module is used for performing normalization processing on the historical wind power data to obtain a historical wind power data set;
the dividing module is used for dividing the historical wind power data set into a training set data set and a test set data set;
the long-short term memory network prediction model construction module is used for constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer;
the short-term wind power prediction model construction module is used for performing particle swarm optimization on the long and short-term memory network prediction model by utilizing the training set data set to determine the optimal parameters of the long and short-term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters;
and the prediction module is used for taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
With reference to the second aspect, in an implementation manner, the historical wind power data is normalized by using the following formula:
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminRepresenting the minimum value of a sequence before normalization.
With reference to the second aspect, in one implementation manner, the long-short term memory network prediction model is constructed by adopting the following method:
the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller;
predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain a plurality of preliminary prediction results;
the output layer adopts a plurality of preliminary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain final prediction results, and finally carries out evaluation analysis on the prediction results by combining with experimental evaluation indexes.
With reference to the second aspect, in one implementation manner, the short-term wind power prediction model building module is configured to perform the following operations:
1) initializing parameters of the long-term and short-term memory network prediction model, setting respective value ranges and search ranges of the number of neurons and the learning rate, and determining the maximum iteration times, the population number, the maximum inertia weight value, the minimum inertia weight value, the initial value of a first acceleration factor, the final value of the first acceleration factor, the initial value of a second acceleration factor and the final value of the second acceleration factor;
2) according to the initialized values of the number of the neurons and the learning rate, a long-short term memory prediction model is used for carrying out model prediction on a test set data set, the average absolute percentage error of the obtained prediction result is used as the initial fitness value of the particles, and the fitness function is obtained by adopting the following formula:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) The predicted value of the wind power at the moment i is obtained;
3) taking two parameters of the number of neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formula, obtaining the optimal values of the two parameters by optimizing through a particle swarm optimization, and reestablishing an optimized long-short term memory network prediction model on the basis of the optimal values;
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax)；
wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vki id represents the velocity of the particle; x k id represents the position of the particle(ii) a P k id represents an individual local optimal solution of the particle; p k gd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factor c1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2A final value of;
4) continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
With reference to the second aspect, in an implementation manner, after obtaining the short-term wind power prediction result, the method further includes: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) (i-1, 2, … n) is the fourthAnd (5) predicting the wind power at the moment i.
The application discloses a short-term wind power prediction method and a short-term wind power prediction device, wherein the method comprises the following steps: obtaining historical wind power data, wherein the historical wind power data comprises historical average wind power data; carrying out normalization processing on the historical wind power data to obtain a historical wind power data set; dividing the historical wind power data set into a training set data set and a test set data set; constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer; performing particle swarm optimization on the long-short term memory network prediction model by using the training set data set to determine the optimal parameters of the long-short term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters; and taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result. By adopting the scheme, the problems of insufficient model fitting capability and low prediction precision caused by parameter selection according to experience are solved, the optimal parameters of the long-term and short-term memory network model can be quickly searched and determined by using the prediction method, the training efficiency is high, the prediction precision is further improved, the wind power can be timely and accurately predicted, and the accurate scheduling operation of the wind power plant is realized.
Drawings
In order to more clearly explain the technical solution of the present application, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious to 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 short-term wind power prediction method provided in an embodiment of the present application;
fig. 2 is a schematic structural diagram of a short-term wind power prediction device provided in an embodiment of the present application;
FIG. 3 is a diagram illustrating the variation of particle fitness value with the number of iterations in an embodiment of the present application;
FIG. 4 is a diagram illustrating the number of neurons varying with the number of iterations in an embodiment of the present application;
FIG. 5 is a diagram illustrating a variation of a learning rate with a variation of an iteration number according to an embodiment of the present application;
FIG. 6 is a diagram illustrating a comparison between a predicted curve and an actual curve of each model in the embodiment of the present application.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, the present application is described in further detail with reference to the accompanying drawings and the detailed description.
As can be known from the description of the background art, short-term wind power prediction in the prior art mainly depends on experience to select parameters, which results in poor fitting degree of a prediction model and further reduces prediction precision, and therefore, to solve the above problems, the embodiment of the present application discloses a short-term wind power prediction method and apparatus based on Particle Swarm Optimization (PSO), which can quickly search and determine optimal parameters of a long-term and short-term memory network model, have high training efficiency, further improve prediction precision, and predict wind power timely and accurately, thereby realizing accurate scheduling operation of a wind farm.
The embodiment of the application discloses a short-term wind power prediction method, and with reference to fig. 1, the method comprises the following steps:
and S11, acquiring historical wind power data, wherein the historical wind power data comprises historical average wind power data.
The method mainly comprises the steps of collecting historical wind power data, preprocessing the historical wind power data after collection, for example, removing obviously abnormal data, and calculating an average value of the historical wind power data.
And S12, performing normalization processing on the historical wind power data to obtain a historical wind power data set.
In this step, the historical wind power data obtained in step S11 is normalized, that is, the sample data is mapped between [0, 1], and the normalization formula is shown in formula (1):
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminRepresenting the minimum value of a sequence before normalization.
And S13, dividing the historical wind power data set into a training set data set and a test set data set.
In this step, the historical wind power data set after normalization processing in step S12 is randomly divided into a training set data set and a test set data set, and the division ratio may be determined according to actual needs, for example: and randomly taking 90% of historical wind power data in the historical wind power data set as a training set data set, and taking the remaining 10% as a test set data set.
And S14, constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer.
The long-short term memory (LSTM) network prediction model is a time cycle network, and is relatively fitted with the characteristic that wind power changes according to a time sequence, so that the long-short term memory network prediction model is selected in the embodiment.
Optionally, the long-term and short-term memory network prediction model is constructed by adopting the following method:
and the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller.
Predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain n primary prediction results P1、P2、…Pn。
The output layer adopts n primary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain a final prediction result, and finally carries out evaluation analysis on the prediction result by combining with experimental evaluation indexes.
S15, performing particle swarm optimization on the long-short term memory network prediction model by using the training set data set to determine the optimal parameters of the long-short term memory network prediction model; and establishing a short-term wind power prediction model according to the optimal parameters.
The short-term wind power prediction model is obtained by optimizing a long-term and short-term memory network prediction model by using a particle swarm optimization, and therefore, the short-term wind power prediction model can also be called as a PSO-LSTM model.
Optionally, the step may be specifically implemented by the following method:
1) initializing parameters of the long-short term memory network prediction model, setting respective value ranges and search ranges of the number m of neurons and the learning rate lr, and determining the maximum iteration number TmaxWith population quantity pop and maximum value of inertial weight ωmaxMinimum value of inertia weight ωminA first acceleration factor c1Initial value c1,iniA first acceleration factor c1Final value c1,finA second acceleration factor c2Initial value c2,iniA second acceleration factor c2Final value c2,fin。
2) According to the initialized values of the number of the neurons and the learning rate, performing model prediction on a test set data set by using a long-short term memory prediction model, taking the average absolute percentage error of the obtained prediction result as the initial fitness value of the particles, and using a fitness function fitiThe following formula is used to obtain:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) And the predicted value is the wind power at the moment i.
3) Taking two parameters of the number of the neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formulas (3) to (7), obtaining the optimal value of the two parameters by the particle swarm optimization, and then reestablishing the optimized long-short term memory network prediction model on the basis of the optimal value.
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2 (3)；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax) (4)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax) (5)；
Wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vki id represents the velocity of the particle; x k id represents the position of the particle; p k id represents an individual local optimal solution of the particle; p k gd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factor c1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2The final value of (c).
4) Continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
And S16, taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
In this step, the test set data set is used as the input variable of the short-term wind power prediction model obtained in step S15, and the output parameter is the short-term wind power prediction result.
Optionally, after the short-term wind power prediction result is obtained, in order to verify the accuracy of the result, the method further includes: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) And (i is 1,2, … n) is a predicted value of the wind power at the ith moment.
Wherein y isMAPEAnd yRMSEThe smaller the value, the greater the goodness of fit, and the more accurate the model prediction results.
According to the scheme, historical wind power data of the wind power plant are collected firstly; preprocessing and normalizing all collected data, and then dividing a training data set and a testing data set; then, constructing a long-term and short-term memory network prediction model, optimizing the long-term and short-term memory network prediction model through a particle swarm algorithm to obtain two optimal prediction model parameters, and reestablishing a short-term wind power prediction model according to the obtained optimal prediction model parameters; and finally, the test data set is used as input, the short-term wind power to be predicted is used as output, and the prediction of the short-term wind power is realized. The method solves the problems of insufficient model fitting capability and low prediction precision caused by parameter selection according to experience, can quickly search and determine the optimal parameters of the long-term and short-term memory network model, is high in training efficiency, further improves the prediction precision, and can timely and accurately predict the wind power, so that the wind power plant can be accurately scheduled and operated.
The following are embodiments of the apparatus of the present invention that may be used to perform embodiments of the method of the present invention. For details which are not disclosed in the embodiments of the apparatus of the present invention, reference is made to the embodiments of the method of the present invention.
The embodiment of the application also discloses a short-term wind power prediction device, referring to fig. 2, the device includes:
the historical wind power data acquisition module 10 is used for acquiring historical wind power data, wherein the historical wind power data comprises historical average wind power data;
the normalization module 20 is configured to perform normalization processing on the historical wind power data to obtain a historical wind power data set;
the dividing module 30 is configured to divide the historical wind power data set into a training set data set and a test set data set;
the long-short term memory network prediction model building module 40 is used for building a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer;
the short-term wind power prediction model building module 50 is used for performing particle swarm optimization on the long and short-term memory network prediction model by utilizing the training set data set to determine the optimal parameters of the long and short-term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters;
and the prediction module 60 is configured to use the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
Optionally, the historical wind power data is normalized by using the following formula:
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminRepresenting the minimum value of a sequence before normalization.
Optionally, the long-term and short-term memory network prediction model is constructed by adopting the following method:
the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller;
predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain a plurality of preliminary prediction results;
the output layer adopts a plurality of preliminary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain final prediction results, and finally carries out evaluation analysis on the prediction results by combining with experimental evaluation indexes.
Optionally, the short-term wind power prediction model building module is configured to perform the following operations:
1) initializing parameters of the long-term and short-term memory network prediction model, setting respective value ranges and search ranges of the number of neurons and the learning rate, and determining the maximum iteration times, the population number, the maximum inertia weight value, the minimum inertia weight value, the initial value of a first acceleration factor, the final value of the first acceleration factor, the initial value of a second acceleration factor and the final value of the second acceleration factor;
2) according to the initialized values of the number of the neurons and the learning rate, a long-short term memory prediction model is used for carrying out model prediction on a test set data set, the average absolute percentage error of the obtained prediction result is used as the initial fitness value of the particles, and the fitness function is obtained by adopting the following formula:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) The predicted value of the wind power at the moment i is obtained;
3) taking two parameters of the number of neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formula, obtaining the optimal values of the two parameters by optimizing through a particle swarm optimization, and reestablishing an optimized long-short term memory network prediction model on the basis of the optimal values;
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax)；
wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vki id represents the velocity of the particle; x k id represents the position of the particle; p k id represents an individual local optimal solution of the particle; p k gd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factorc1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2A final value of;
4) continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
Optionally, after obtaining the short-term wind power prediction result, the method further includes: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) And (i is 1,2, … n) is a predicted value of the wind power at the ith moment.
Examples
In order to verify the effectiveness of the short-term wind power prediction method based on the PSO-LSTM combination, the method further discloses a specific embodiment, historical average wind power data of a certain wind power plant generator set from 09/15/2020/15 to 10/15/2020/15 (30 days in total) are selected as input data of the PSO-LSTM model, a wind power predicted value is used as output, modeling analysis is carried out, and the data sampling time interval is 15 min. 2688 data in the first 28 days are selected as a training set for training the PSO-LSTM model, 192 data in the last two days are selected as a test data set for the model to perform tests, and the wind power of 48h in the future is predicted.
Aiming at short-term wind power prediction, the method comprises the steps of firstly collecting historical wind power data of a wind power plant; preprocessing and normalizing all collected data, and dividing a training set and a test set; then constructing a long-term and short-term memory network prediction model, optimizing the long-term and short-term memory network prediction model through a particle swarm algorithm, obtaining two optimal prediction model parameters, namely a learning rate lr and the number m of neurons, and reestablishing the short-term wind power prediction model according to the obtained optimal prediction model parameters; and finally, the test data set is used as input, the wind power to be predicted is used as output, and the short-term wind power is predicted.
And the sample data is used for prediction under different models, and the effectiveness of the short-term wind power prediction method based on the PSO-LSTM is verified. In the process of optimizing the long-term and short-term memory network prediction model by using the particle swarm optimization, the setting of model parameters is shown in table 1, the comparison of different prediction models is carried out, and the setting of model parameters of the LSTM model and the RNN model before optimization is shown in table 2.
TABLE 1 particle swarm optimization for optimizing parameters of long and short term memory network model
| Parameter(s) | Value taking | Parameter(s) | Value taking |
| Number of |
200 | Maximum value of inertial weight ωmax | 0.9 |
| |
100 | Minimum value of inertial weight ωmin | 0.4 |
| Value range of m | [1，9] | Initial value c of acceleration factor1,ini | 2 |
| m search ranges | [-1，1] | Final value c of acceleration factor1,fin | 0.5 |
| Range of lr values | [0.01，0.01] | Initial value c of acceleration factor2,ini | 0.5 |
| lr search scope | [-0.01，0.01] | Final value c of |
2 |
TABLE 2 LSTM, RNN, SVM and RF model parameters
Fig. 3 shows a rule that a particle fitness value of a long and short term memory network prediction model optimized by using an improved particle swarm optimization changes with the change of the iteration times, fig. 4 shows a rule that the number of neurons of a long and short term memory network prediction model optimized by using the improved particle swarm optimization changes with the change of the iteration times, and fig. 5 shows a rule that a learning rate of a long and short term memory network prediction model optimized by using the improved particle swarm optimization changes with the change of the iteration times, so that the particle fitness value is finally stabilized at 0.17, the number of neurons is stabilized at 7, and the learning rate is finally stabilized at 0.0014 with the change of the iteration times.
FIG. 6 is a line graph comparing the predicted values and the true values of the wind power of different prediction models by using the 192 sets of test data sets, wherein 3 models such as a PSO-LSTM model, a long-short term memory network model LSTM and a recurrent neural network model RNN which are optimized by the particle swarm optimization have better prediction performance.
The results of the long-short term memory network prediction model PSO-LSTM and the long-short term memory network model LSTM which predict the wind power and are optimized based on the particle swarm optimization are shown in table 3 in combination with evaluation indexes, and the long-short term memory network prediction model IPSO-LSTM and the long-short term memory network model LSTM which are optimized based on the particle swarm optimization compare the prediction effect of the long-short term memory network model LSTM, wherein yMAPEThe indexes are reduced by 36.40 percent, 67.83 percent and yRMSEThe index is reduced by 75.69%, which shows that the prediction accuracy of the long-short term memory network prediction model PSO-LSTM optimized by the particle swarm algorithm is improved, and the necessity of parameter optimization of the long-short term memory network prediction model LSTM is reflected.
TABLE 3 comparison of prediction results for different prediction models
| Model (model) | Average relative error (y)MAPE/％) | Root mean square error (y)RMSE/％) |
| IPSO-LSTM | 0.170 | 0.432 |
| LSTM | 0.873 | 0.759 |
Therefore, according to the method and the device for predicting the short-term wind power of the long-term and short-term memory network optimized by the particle swarm algorithm, the optimal parameters of the long-term and short-term memory network model can be quickly searched and determined, the training efficiency is high, the problems of insufficient model fitting capability and low prediction precision caused by parameter selection according to experience are solved, the prediction precision is further improved, a basis is provided for the operation condition of a wind power plant, and a reference is provided for a dispatcher to provide an operation plan.
The same and similar parts in the various embodiments in this specification may be referred to each other. In particular, as for the apparatus embodiment, since it is substantially similar to the method embodiment, the description is simple, and the relevant points can be referred to the description in the method embodiment.
The present application has been described in detail with reference to specific embodiments and illustrative examples, but the description is not intended to limit the application. Those skilled in the art will appreciate that various equivalent substitutions, modifications or improvements may be made to the presently disclosed embodiments and implementations thereof without departing from the spirit and scope of the present disclosure, and these fall within the scope of the present disclosure. The protection scope of this application is subject to the appended claims.
Claims (10)
1. A short-term wind power prediction method is characterized by comprising the following steps:
obtaining historical wind power data, wherein the historical wind power data comprises historical average wind power data;
carrying out normalization processing on the historical wind power data to obtain a historical wind power data set;
dividing the historical wind power data set into a training set data set and a test set data set;
constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer;
performing particle swarm optimization on the long-short term memory network prediction model by using the training set data set to determine the optimal parameters of the long-short term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters;
and taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
2. The short-term wind power prediction method according to claim 1, characterized in that the historical wind power data is normalized using the following formula:
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminExpress a certainMinimum before sequence normalization.
3. The short-term wind power prediction method according to claim 1, characterized in that the long-term and short-term memory network prediction model is constructed by adopting the following method:
the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller;
predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain a plurality of preliminary prediction results;
the output layer adopts a plurality of preliminary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain final prediction results, and finally carries out evaluation analysis on the prediction results by combining with experimental evaluation indexes.
4. The short-term wind power prediction method according to claim 1, characterized in that the training set data set is used for performing particle swarm optimization on the long-term and short-term memory network prediction model to determine optimal parameters of the long-term and short-term memory network prediction model; and establishing a short-term wind power prediction model according to the optimal parameters, wherein the short-term wind power prediction model comprises the following steps:
1) initializing parameters of the long-term and short-term memory network prediction model, setting respective value ranges and search ranges of the number of neurons and the learning rate, and determining the maximum iteration times, the population number, the maximum inertia weight value, the minimum inertia weight value, the initial value of a first acceleration factor, the final value of the first acceleration factor, the initial value of a second acceleration factor and the final value of the second acceleration factor;
2) according to the initialized values of the number of the neurons and the learning rate, a long-short term memory prediction model is used for carrying out model prediction on a test set data set, the average absolute percentage error of the obtained prediction result is used as the initial fitness value of the particles, and the fitness function is obtained by adopting the following formula:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) The predicted value of the wind power at the moment i is obtained;
3) taking two parameters of the number of neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formula, obtaining the optimal values of the two parameters by optimizing through a particle swarm optimization, and reestablishing an optimized long-short term memory network prediction model on the basis of the optimal values;
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax)；
wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vk id represents the velocity of the particle; xk id represents the position of the particle; pkid represents the individual local optimal solution of the particle; pkgd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factor c1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2A final value of;
4) continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
5. The short-term wind power prediction method according to claim 1, further comprising, after obtaining the short-term wind power prediction result: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) And (i is 1,2, … n) is a predicted value of the wind power at the ith moment.
6. A short-term wind power prediction device, comprising:
the historical wind power data acquisition module is used for acquiring historical wind power data, and the historical wind power data comprises historical average wind power data;
the normalization module is used for performing normalization processing on the historical wind power data to obtain a historical wind power data set;
the dividing module is used for dividing the historical wind power data set into a training set data set and a test set data set;
the long-short term memory network prediction model construction module is used for constructing a long-short term memory network prediction model comprising an input layer, a hidden layer and an output layer;
the short-term wind power prediction model construction module is used for performing particle swarm optimization on the long and short-term memory network prediction model by utilizing the training set data set to determine the optimal parameters of the long and short-term memory network prediction model; establishing a short-term wind power prediction model according to the optimal parameters;
and the prediction module is used for taking the test set data set as an input variable of the optimized short-term wind power prediction model to obtain a short-term wind power prediction result.
7. The short-term wind power prediction device of claim 6, wherein the historical wind power data is normalized using the following formula:
wherein y represents data before normalization; y is*Represents the normalized data; y ismaxRepresents the maximum value of a sequence before normalization; y isminRepresenting the minimum value of a sequence before normalization.
8. The short-term wind power prediction device according to claim 6, wherein the long-term and short-term memory network prediction model is constructed by adopting the following method:
the hidden layer trains the long-short term memory network prediction model by adopting the training set data set, and determines the parameters of the long-short term memory network prediction model under the condition that the evaluation standard error of the prediction result is smaller;
predicting the test set data by adopting the long-term and short-term memory network prediction model to obtain a plurality of preliminary prediction results;
the output layer adopts a plurality of preliminary prediction results obtained by calculation in a mode of average absolute percentage error, carries out inverse normalization processing to obtain final prediction results, and finally carries out evaluation analysis on the prediction results by combining with experimental evaluation indexes.
9. The short-term wind power prediction device according to any of claims 6-8, characterized by a short-term wind power prediction model construction module for performing the following operations:
1) initializing parameters of the long-term and short-term memory network prediction model, setting respective value ranges and search ranges of the number of neurons and the learning rate, and determining the maximum iteration times, the population number, the maximum inertia weight value, the minimum inertia weight value, the initial value of a first acceleration factor, the final value of the first acceleration factor, the initial value of a second acceleration factor and the final value of the second acceleration factor;
2) according to the initialized values of the number of the neurons and the learning rate, a long-short term memory prediction model is used for carrying out model prediction on a test set data set, the average absolute percentage error of the obtained prediction result is used as the initial fitness value of the particles, and the fitness function is obtained by adopting the following formula:
wherein n represents the sample capacity of the test set data set; xact(i) The real value of the wind power at the moment i; xpred(i) The predicted value of the wind power at the moment i is obtained;
3) taking two parameters of the number of neurons and the learning rate as particles, taking the initial fitness value as the initial particle fitness value, iteratively updating the speed and the position of the two particles by adopting the following formula, obtaining the optimal values of the two parameters by optimizing through a particle swarm optimization, and reestablishing an optimized long-short term memory network prediction model on the basis of the optimal values;
ω＝ωmax-(ωmax-ωmin)(k/Tmax)2；
c1＝c1,ini-(c1,ini-c1,fin)(k/Tmax)；
c2＝c2,ini+(c2,fin-c2,ini)(k/Tmax)；
wherein k represents the current iteration number; t ismaxRepresenting the maximum iteration number; vki id represents the velocity of the particle; x k id represents the position of the particle; p k id represents an individual local optimal solution of the particle; pkgd represents a global optimal solution for the particle; r is1And r2Is [0, 1]]A random number in between; omega is an inertia factor, omegamaxRepresenting the maximum value of the iterative inertia weight; omegaminRepresenting an iterative inertia weight minimum; c. C1,iniRepresents a first acceleration factor c1An initial value of (1); c. C1,finRepresents a first acceleration factor c1A final value of; c. C2,iniRepresenting a second acceleration factor c2An initial value of (1); c. C2,finRepresenting a second acceleration factor c2A final value of;
4) continuously iteratively updating the speed and the position of the particles, calculating a corresponding particle fitness value, and comparing the local optimal solution with the global optimal solution; when the particle fitness value tends to be stable or the iteration times reach the maximum, the termination condition is met, the optimal neuron number and the optimal learning rate are obtained, and a short-term wind power prediction model optimized by a particle swarm algorithm is obtained; otherwise, returning to execute the step 3) and continuing to carry out iterative updating.
10. The short-term wind power prediction device of claim 6, further comprising, after obtaining the short-term wind power prediction result: evaluating the short-term wind power prediction result by using the average relative percentage error and root mean square error evaluation indexes;
the average relative percentage error is obtained by adopting the following formula:
the root mean square error is obtained by adopting the following formula:
wherein, yMAPERepresents the homogeneous relative percentage error; y isRMSERepresenting root mean square error; n represents the sample capacity of the test data set; xact(i) (i is 1,2, … n) is the true value of the wind power at the ith moment; xpred(i) And (i is 1,2, … n) is a predicted value of the wind power at the ith moment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110018074.8A CN112686464A (en) | 2021-01-07 | 2021-01-07 | Short-term wind power prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110018074.8A CN112686464A (en) | 2021-01-07 | 2021-01-07 | Short-term wind power prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112686464A true CN112686464A (en) | 2021-04-20 |
Family
ID=75456236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110018074.8A Pending CN112686464A (en) | 2021-01-07 | 2021-01-07 | Short-term wind power prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686464A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113111592A (en) * | 2021-05-06 | 2021-07-13 | 云南电力技术有限责任公司 | Short-term wind power prediction method based on EMD-LSTM |
| CN113112091A (en) * | 2021-05-06 | 2021-07-13 | 云南电力技术有限责任公司 | Nuclear power unit power prediction method based on PCA and LSTM |
| CN113268913A (en) * | 2021-06-24 | 2021-08-17 | 广州鼎泰智慧能源科技有限公司 | Intelligent building air conditioner cooling machine system operation optimization method based on PSO-ELM algorithm |
| CN113468817A (en) * | 2021-07-13 | 2021-10-01 | 淮阴工学院 | Ultra-short-term wind power prediction method based on IGOA (optimized El-electric field model) |
| CN113468810A (en) * | 2021-07-01 | 2021-10-01 | 天行智控(成都)科技有限公司 | Intelligent floor sensing indoor tumble prediction model and establishment method thereof |
| CN113515852A (en) * | 2021-06-07 | 2021-10-19 | 河北工业大学 | Sustainable learning fan output prediction method, system, equipment and storage medium |
| CN113570120A (en) * | 2021-07-07 | 2021-10-29 | 上海电机学院 | Wind power prediction method and device based on improved particle swarm optimization |
| CN113673775A (en) * | 2021-08-25 | 2021-11-19 | 武汉科技大学 | Time-space combination prediction method based on CNN-LSTM and deep learning |
| CN113687433A (en) * | 2021-09-15 | 2021-11-23 | 湖南师范大学 | Bi-LSTM-based magnetotelluric signal denoising method and system |
| CN113705922A (en) * | 2021-09-06 | 2021-11-26 | 内蒙古科技大学 | Improved ultra-short-term wind power prediction algorithm and model establishment method |
| CN113779875A (en) * | 2021-08-31 | 2021-12-10 | 润联软件系统(深圳)有限公司 | Wind power generation capacity prediction method based on graph neural model and related equipment thereof |
| CN113887804A (en) * | 2021-09-30 | 2022-01-04 | 沈阳工程学院 | Short-term wind power prediction method based on improved local abnormal factor algorithm |
| CN113988398A (en) * | 2021-10-22 | 2022-01-28 | 华北电力大学 | Wind turbine generator power prediction method and device, electronic equipment and storage medium |
| CN116960960A (en) * | 2023-07-24 | 2023-10-27 | 青岛哈尔滨工程大学创新发展中心 | Short-term wind power prediction method for wind turbine |
| CN117034785A (en) * | 2023-10-09 | 2023-11-10 | 中国华能集团清洁能源技术研究院有限公司 | Wind turbine generator power prediction method and device and electronic equipment |
| CN117313043A (en) * | 2023-10-25 | 2023-12-29 | 四川大学 | Wind power generation power prediction method |
| CN117313043B (en) * | 2023-10-25 | 2024-04-30 | 四川大学 | Wind power generation power prediction method |
-
2021
- 2021-01-07 CN CN202110018074.8A patent/CN112686464A/en active Pending
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113111592A (en) * | 2021-05-06 | 2021-07-13 | 云南电力技术有限责任公司 | Short-term wind power prediction method based on EMD-LSTM |
| CN113112091A (en) * | 2021-05-06 | 2021-07-13 | 云南电力技术有限责任公司 | Nuclear power unit power prediction method based on PCA and LSTM |
| CN113515852A (en) * | 2021-06-07 | 2021-10-19 | 河北工业大学 | Sustainable learning fan output prediction method, system, equipment and storage medium |
| CN113268913A (en) * | 2021-06-24 | 2021-08-17 | 广州鼎泰智慧能源科技有限公司 | Intelligent building air conditioner cooling machine system operation optimization method based on PSO-ELM algorithm |
| CN113468810A (en) * | 2021-07-01 | 2021-10-01 | 天行智控(成都)科技有限公司 | Intelligent floor sensing indoor tumble prediction model and establishment method thereof |
| CN113570120A (en) * | 2021-07-07 | 2021-10-29 | 上海电机学院 | Wind power prediction method and device based on improved particle swarm optimization |
| CN113468817A (en) * | 2021-07-13 | 2021-10-01 | 淮阴工学院 | Ultra-short-term wind power prediction method based on IGOA (optimized El-electric field model) |
| CN113468817B (en) * | 2021-07-13 | 2023-08-22 | 淮阴工学院 | Ultra-short-term wind power prediction method based on IGOA (insulated gate bipolar transistor) optimized ELM (ELM) |
| CN113673775A (en) * | 2021-08-25 | 2021-11-19 | 武汉科技大学 | Time-space combination prediction method based on CNN-LSTM and deep learning |
| CN113779875A (en) * | 2021-08-31 | 2021-12-10 | 润联软件系统(深圳)有限公司 | Wind power generation capacity prediction method based on graph neural model and related equipment thereof |
| CN113705922A (en) * | 2021-09-06 | 2021-11-26 | 内蒙古科技大学 | Improved ultra-short-term wind power prediction algorithm and model establishment method |
| CN113705922B (en) * | 2021-09-06 | 2023-09-12 | 内蒙古科技大学 | Improved ultra-short-term wind power prediction algorithm and model building method |
| CN113687433A (en) * | 2021-09-15 | 2021-11-23 | 湖南师范大学 | Bi-LSTM-based magnetotelluric signal denoising method and system |
| CN113887804A (en) * | 2021-09-30 | 2022-01-04 | 沈阳工程学院 | Short-term wind power prediction method based on improved local abnormal factor algorithm |
| CN113988398A (en) * | 2021-10-22 | 2022-01-28 | 华北电力大学 | Wind turbine generator power prediction method and device, electronic equipment and storage medium |
| CN116960960A (en) * | 2023-07-24 | 2023-10-27 | 青岛哈尔滨工程大学创新发展中心 | Short-term wind power prediction method for wind turbine |
| CN116960960B (en) * | 2023-07-24 | 2024-02-09 | 青岛哈尔滨工程大学创新发展中心 | Short-term wind power prediction method for wind turbine |
| CN117034785A (en) * | 2023-10-09 | 2023-11-10 | 中国华能集团清洁能源技术研究院有限公司 | Wind turbine generator power prediction method and device and electronic equipment |
| CN117313043A (en) * | 2023-10-25 | 2023-12-29 | 四川大学 | Wind power generation power prediction method |
| CN117313043B (en) * | 2023-10-25 | 2024-04-30 | 四川大学 | Wind power generation power prediction method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112686464A (en) | Short-term wind power prediction method and device | |
| Hong et al. | A hybrid deep learning-based neural network for 24-h ahead wind power forecasting | |
| Tian | Short-term wind speed prediction based on LMD and improved FA optimized combined kernel function LSSVM | |
| CN109902801B (en) | Flood collective forecasting method based on variational reasoning Bayesian neural network | |
| CN106446540B (en) | A kind of Wind turbines health status real time evaluating method | |
| CN111860982A (en) | Wind power plant short-term wind power prediction method based on VMD-FCM-GRU | |
| Wang et al. | Wind speed forecasting based on multi-objective grey wolf optimisation algorithm, weighted information criterion, and wind energy conversion system: A case study in Eastern China | |
| Zhou et al. | Short-term wind power prediction optimized by multi-objective dragonfly algorithm based on variational mode decomposition | |
| CN111967696B (en) | Neural network-based electric vehicle charging demand prediction method, system and device | |
| CN111241755A (en) | Power load prediction method | |
| Peng et al. | Wind power prediction for wind farm clusters based on the multi-feature similarity matching method | |
| Kaplan et al. | A novel method based on Weibull distribution for short-term wind speed prediction | |
| Xiyun et al. | Wind power probability interval prediction based on bootstrap quantile regression method | |
| CN113983646A (en) | Air conditioner interaction end energy consumption prediction method based on generation countermeasure network and air conditioner | |
| Wibawa et al. | Long Short-Term Memory to Predict Unique Visitors of an Electronic Journal | |
| CN115099511A (en) | Photovoltaic power probability estimation method and system based on optimized copula | |
| CN115239105A (en) | Method and device for evaluating wind resources of in-service wind power plant | |
| Danish | A framework for modeling and optimization of data-driven energy systems using machine learning | |
| CN116307139A (en) | Wind power ultra-short-term prediction method for optimizing and improving extreme learning machine | |
| CN115616333A (en) | Power distribution network line loss prediction method and system | |
| CN112581311B (en) | Method and system for predicting long-term output fluctuation characteristics of aggregated multiple wind power plants | |
| CN113762591A (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM counterstudy | |
| Shen et al. | An interval analysis scheme based on empirical error and mcmc to quantify uncertainty of wind speed | |
| CN106097141A (en) | The ternary interval number appraisal procedure of photovoltaic generating system confidence capacity | |
| Indralaksono et al. | Hierarchical Clustering and Deep Learning for Short-Term Load Forecasting with Influenced Factors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210420 |
|
| RJ01 | Rejection of invention patent application after publication |