CN111191814A - Electricity price prediction method, system and computer readable storage medium - Google Patents
Electricity price prediction method, system and computer readable storage medium Download PDFInfo
- Publication number
- CN111191814A CN111191814A CN201911157776.3A CN201911157776A CN111191814A CN 111191814 A CN111191814 A CN 111191814A CN 201911157776 A CN201911157776 A CN 201911157776A CN 111191814 A CN111191814 A CN 111191814A
- Authority
- CN
- China
- Prior art keywords
- electricity price
- prediction
- mode
- daily fluctuation
- day
- 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.)
- Withdrawn
Links
- 230000005611 electricity Effects 0.000 title claims abstract description 197
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 34
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000013528 artificial neural network Methods 0.000 claims description 14
- 238000003909 pattern recognition Methods 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000004931 aggregating effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0206—Price or cost determination based on market factors
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0283—Price estimation or determination
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a power price prediction method, a system and a computer readable storage medium, wherein the method comprises the following steps: gathering the historical electricity price data into m types through a K-means algorithm, and obtaining the historical electricity price data with the mode label; training and constructing a power price mode recognition model according to historical power price data with mode labels; establishing n independent day-ahead electricity price prediction models; respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models; inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes; calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode; the method and the device can improve the accuracy of the daily fluctuation mode prediction of the electricity price and further improve the accuracy of the day-ahead electricity price prediction.
Description
Technical Field
The present invention relates to the field of power, and in particular, to a method, a system, and a computer-readable storage medium for predicting electricity prices.
Background
The electricity price prediction has important significance for making a reasonable quotation strategy, maintaining the safety and stability of the electric power market and improving the economical efficiency of system operation. The day-ahead electricity price prediction is an important component of electricity price prediction of an electric power market and is mainly used for forecasting the electricity price trend within 24h in the future. With the continuous promotion of the market reformation of electric power in China. Higher requirements are put on the accuracy of the electricity price prediction. Therefore, the prediction of day-ahead electricity prices has become a focus of research in the power field.
In recent years, different theories and methods have been proposed for the price prediction of electricity. The prediction method comprises a traditional time series prediction method (such as accumulative autoregressive moving average, generalized autoregressive conditional variance and the like), an emerging machine learning algorithm (such as an artificial neural network, a support vector machine and the like) and a combined prediction method (such as multi-model combined prediction based on evidence theory, combined prediction based on a neural network and an adaptive neural fuzzy processing system and the like). Although the above electricity price prediction methods greatly improve the original model through data preprocessing or parameter optimization and the like, and simulation verification is performed in respective set scenes or power markets, the unified modeling method based on all historical data cannot distinguish the difference of the day fluctuation modes of the electricity price sequence, so that electricity price data of different day fluctuation modes are influenced mutually in the establishment process of the prediction model, the fitting degree of the model to input and output is poor, and a more ideal prediction result is difficult to obtain.
Disclosure of Invention
In order to solve at least one of the above technical problems, the present invention provides a power rate prediction method, system and computer-readable storage medium.
In order to achieve the above object, a first aspect of the present invention provides an electricity price prediction method, including:
gathering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with the mode labels;
training and constructing a power price mode recognition model according to historical power price data with mode labels;
selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models;
inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
Further, calculating a final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, specifically comprising:
respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
and respectively accumulating the weights voted by different day-ahead power price prediction models for each power price daily fluctuation mode, and obtaining the final voting score of each power price daily fluctuation mode.
Further, the expression of the weighted voting algorithm is as follows:
wherein: wX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
Further, the method for aggregating the historical electricity price data into m types in an unsupervised mode through a K-means algorithm specifically comprises the following steps:
randomly selecting m samples from the historical electricity price data as initial clustering centers cj(j＝1，2，...，m)；
For each sample x in the historical electricity price dataiCalculating Euclidean distances between the samples and m cluster centers, and classifying the samples to the cluster center with the minimum distanceClass (c);
aiming at a new class formed after the sample is divided, according to a calculation formula:
recalculating cluster centers, where | cj| represents class cjThe number of samples in (1);
and repeating the two steps until the position of the clustering center is not changed.
Further, training and constructing a power price pattern recognition model according to the historical power price data with the pattern label specifically comprises the following steps:
establishing an electricity price mode identification model;
extracting the characteristics of the historical electricity price data with the mode label to obtain a characteristic vector;
and training a power price mode recognition model based on the characteristic vector and the mode label by adopting a machine learning method so as to optimize the recognition accuracy of the power price mode recognition model.
Further, after the power rate pattern recognition model is trained and constructed according to the historical power rate data with the pattern labels, the method further comprises the following steps:
establishing a day-ahead electricity price prediction model based on an RBF neural network according to corresponding historical electricity price data;
and obtaining a power price prediction sequence of a prediction day based on a day-ahead power price prediction model of the RBF neural network.
Further, the algorithm in the RBF neural network-based day-ahead electricity price prediction model is as follows:
in the formula: p1, 2, P, and P is the number of samples; w is aqrRepresenting the connection weight between the hidden layer and the output layer; h is the number of nodes of the hidden layer; y isrRepresenting the output result of the input sample corresponding to the r node of the network; II xp-cqII is the Euclidean norm; c. CqAnd σ represents the center and variance of the gaussian function, respectively; xp＝x1p，x2p，...，xkpRepresenting the P input vector; k is the number of nodes of the input layer.
The second aspect of the present invention also provides an electricity price prediction system, including: a memory and a processor, the memory including a power rate prediction method program, the power rate prediction method program when executed by the processor implementing the steps of:
gathering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with the mode labels;
training and constructing a power price mode recognition model according to historical power price data with mode labels;
selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models;
inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
Further, calculating a final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, specifically comprising:
respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
respectively accumulating the weights voted by different day-ahead electricity price prediction models for each electricity price daily fluctuation mode, and obtaining the final voting score of each electricity price daily fluctuation mode;
the expression of the weighted voting algorithm is as follows:
in the formula, WX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
The third aspect of the present invention also provides a computer-readable storage medium, which includes a power rate prediction method program, and when the power rate prediction method program is executed by a processor, the method realizes the steps of the power rate prediction method.
On the basis of clustering historical electricity price data, an electricity price daily fluctuation mode identification model is established for predicting an electricity price daily fluctuation mode; taking the day-ahead power price prediction results of the n basic prediction models as the input of the fluctuation mode identification model to obtain the corresponding n power price day fluctuation mode prediction results; and establishing a day-ahead power rate daily fluctuation mode prediction model based on the credibility weighted combination, distributing different weights to the mode prediction results of the n models, and determining a final mode prediction result according to the voting values of different modes, so that the precision of the power rate daily fluctuation mode prediction is improved, and the accuracy of the day-ahead power rate prediction is further improved.
Additional aspects and advantages of the invention 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 invention.
Drawings
FIG. 1 illustrates a flow chart of a method of electricity price prediction of the present invention;
FIG. 2 is a flow chart illustrating a method for calculating a final voting score for each power rate daily fluctuation mode according to the present invention;
fig. 3 shows a block diagram of an electricity rate prediction system of the present invention.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited by the specific embodiments disclosed below.
Fig. 1 shows a flowchart of a power rate prediction method of the present invention.
As shown in fig. 1, a first aspect of the present invention provides a method for predicting electricity prices, including:
s102, clustering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with mode labels;
s104, training and constructing a power price mode recognition model according to the historical power price data with the mode labels;
s106, selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
s108, obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models respectively;
s110, inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and S112, calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
It should be noted that accurate power rate daily fluctuation mode prediction can provide a clear direction for data selection of a day-ahead power rate prediction model, and the prediction precision of the model is improved from the perspective of improving a modeling data environment.
Fig. 2 is a flowchart illustrating a method for calculating a final voting score for each power rate daily fluctuation pattern according to the present invention.
As shown in fig. 2, calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes specifically includes:
s202, respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
and S204, respectively accumulating the weights voted by different day-ahead power rate prediction models for each power rate daily fluctuation mode, and obtaining the final voting score of each power rate daily fluctuation mode.
Further, the expression of the weighted voting algorithm is as follows:
wherein: wX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
It should be noted that, the credibility weighted combination means that different weights are assigned to each model according to the statistical result of the predictive performance of each day-ahead electricity price prediction model to different electricity price daily fluctuation modes in multiple independent predictions. The principle of the confidence-weighted combination is illustrated by the following specific example. It is assumed that 2 power rate daily fluctuation patterns, i.e., pattern 1 and pattern 2, are obtained by clustering in step S102; three different methods of A, B and B are selected to establish a conventional day-ahead electricity price prediction model, namely m is 2, and n is 3. And setting a table 1 as a mode prediction result obtained after the day-ahead electricity price prediction model passes through the electricity price daily fluctuation mode identification model.
Table 1:
| day-ahead electricity price prediction model | Mode prediction results |
| A | Mode 1 |
| B | Mode 2 |
| C | Mode 1 |
According to table 1, the voting scores of the final daily fluctuation pattern l and the final daily fluctuation pattern 2 are respectively:
S1＝WA，1+WC，1；
S2＝WB，2；
judgment S1And S2If S is1＞S2If the final electricity price mode prediction result is the mode 1; otherwise, it is mode 2.
It can be understood that in order to reflect the performance difference of different methods in the aspect of power rate daily fluctuation mode prediction, different weights are distributed to the mode prediction results obtained by the n prediction methods by adopting credibility weighted combination, so that the accuracy of power rate daily fluctuation mode prediction is improved.
According to the embodiment of the invention, the historical electricity price data is gathered into m types in an unsupervised mode through a K-means algorithm, and the method specifically comprises the following steps:
randomly selecting m samples from the historical electricity price data as initial clustering centers cj(j＝1，2，...，m)；
For each sample x in the historical electricity price dataiCalculate it and m gathersThe Euclidean distance of the class center, and classifying the sample into the class corresponding to the clustering center with the minimum distance;
aiming at a new class formed after the sample is divided, according to a calculation formula:
recalculating cluster centers, where | cj| represents class cjThe number of samples in (1);
and repeating the two steps until the position of the clustering center is not changed.
According to the embodiment of the invention, the power price pattern recognition model is trained and constructed according to the historical power price data with the pattern label, and the method specifically comprises the following steps:
establishing an electricity price mode identification model;
extracting the characteristics of the historical electricity price data with the mode label to obtain a characteristic vector;
and training a power price mode recognition model based on the characteristic vector and the mode label by adopting a machine learning method so as to optimize the recognition accuracy of the power price mode recognition model.
According to an embodiment of the present invention, after training and constructing the power rate pattern recognition model according to the historical power rate data with the pattern labels, the method further comprises:
establishing a day-ahead electricity price prediction model based on an RBF neural network according to corresponding historical electricity price data;
and obtaining a power price prediction sequence of a prediction day based on a day-ahead power price prediction model of the RBF neural network.
Further, the algorithm in the RBF neural network-based day-ahead electricity price prediction model is as follows:
in the formula: p1, 2, P, and P is the number of samples; w is aqrRepresenting the connection weight between the hidden layer and the output layer; h is the number of nodes of the hidden layer; y isrOutput node representing the input sample corresponding to the r-th node of the networkFruit; II xp-cqII is the Euclidean norm; c. CqAnd σ represents the center and variance of the gaussian function, respectively; x is the number ofp＝x1p，x2p，...，xkpRepresenting the P input vector; k is the number of nodes of the input layer.
Fig. 3 shows a block diagram of an electricity rate prediction system of the present invention.
As shown in fig. 3, the second aspect of the present invention also provides an electricity price prediction system 3, where the electricity price prediction system 3 includes: a memory 31 and a processor 32, wherein the memory includes a power rate prediction method program, and the power rate prediction method program when executed by the processor implements the following steps:
gathering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with the mode labels;
training and constructing a power price mode recognition model according to historical power price data with mode labels;
selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models;
inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
It should be noted that the system of the present invention can be operated in a terminal device such as a PC, a mobile phone, a PAD, etc.
It should be noted that the Processor may be a Central Processing Unit (CPU), other general-purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, a discrete hardware component, and so on. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Further, calculating a final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, specifically comprising:
respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
and respectively accumulating the weights voted by different day-ahead power price prediction models for each power price daily fluctuation mode, and obtaining the final voting score of each power price daily fluctuation mode.
Further, the expression of the weighted voting algorithm is as follows:
wherein: wX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
Further, the method for aggregating the historical electricity price data into m types in an unsupervised mode through a K-means algorithm specifically comprises the following steps:
randomly selecting m samples from the historical electricity price data as initial clustering centers cj(j＝1，2，...，m)；
For each sample x in the historical electricity price dataiCalculating Euclidean distances between the samples and m clustering centers, and classifying the samples into a class corresponding to the clustering center with the minimum distance;
for the new class formed after the sample division, according to calculationFormula (II):
recalculating cluster centers, where | cj| represents class cjThe number of samples in (1);
and repeating the two steps until the position of the clustering center is not changed.
Further, training and constructing a power price pattern recognition model according to the historical power price data with the pattern label specifically comprises the following steps:
establishing an electricity price mode identification model;
extracting the characteristics of the historical electricity price data with the mode label to obtain a characteristic vector;
and training a power price mode recognition model based on the characteristic vector and the mode label by adopting a machine learning method so as to optimize the recognition accuracy of the power price mode recognition model.
The electricity price prediction method program further realizes the following steps when executed by the processor:
establishing a day-ahead electricity price prediction model based on an RBF neural network according to corresponding historical electricity price data;
and obtaining a power price prediction sequence of a prediction day based on a day-ahead power price prediction model of the RBF neural network.
Further, the algorithm in the RBF neural network-based day-ahead electricity price prediction model is as follows:
in the formula: p1, 2, P, and P is the number of samples; w is aqrRepresenting the connection weight between the hidden layer and the output layer; h is the number of nodes of the hidden layer; y isrRepresenting the output result of the input sample corresponding to the r node of the network; II xp-cqII is the Euclidean norm; c. CqAnd σ represents the center and variance of the gaussian function, respectively; x is the number ofp＝x1p，x2p，...，xkpRepresenting the P input vector; k is the number of nodes of the input layer.
The third aspect of the present invention also provides a computer-readable storage medium, which includes a power rate prediction method program, and when the power rate prediction method program is executed by a processor, the method realizes the steps of the power rate prediction method.
On the basis of clustering historical electricity price data, an electricity price daily fluctuation mode identification model is established for predicting an electricity price daily fluctuation mode; taking the day-ahead power price prediction results of the n basic prediction models as the input of the fluctuation mode identification model to obtain the corresponding n power price day fluctuation mode prediction results; and establishing a day-ahead power rate daily fluctuation mode prediction model based on the credibility weighted combination, distributing different weights to the mode prediction results of the n models, and determining a final mode prediction result according to the voting values of different modes, so that the precision of the power rate daily fluctuation mode prediction is improved, and the accuracy of the day-ahead power rate prediction is further improved.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described device embodiments are merely illustrative, for example, the division of the unit is only a logical functional division, and there may be other division ways in actual implementation, such as: multiple units or components may be combined, or may be integrated into another system, or some features may be omitted, or not implemented. In addition, the coupling, direct coupling or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection between the devices or units may be electrical, mechanical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units; can be located in one place or distributed on a plurality of network units; some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, all the functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately regarded as one unit, or two or more units may be integrated into one unit; the integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
Alternatively, the integrated unit of the present invention may be stored in a computer-readable storage medium if it is implemented in the form of a software functional module and sold or used as a separate product. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or a part contributing to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, a ROM, a RAM, a magnetic or optical disk, or various other media that can store program code.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (10)
1. A method for predicting electricity prices, the method comprising:
gathering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with the mode labels;
training and constructing a power price mode recognition model according to historical power price data with mode labels;
selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models;
inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
2. The method for predicting electricity prices according to claim 1, wherein calculating the final voting score of each electricity price daily fluctuation mode according to the prediction results of the n electricity price daily fluctuation modes specifically comprises:
respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
and respectively accumulating the weights voted by different day-ahead power price prediction models for each power price daily fluctuation mode, and obtaining the final voting score of each power price daily fluctuation mode.
3. The electricity price prediction method according to claim 2,
the expression of the weighted voting algorithm is as follows:
wherein: wX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
4. The electricity price prediction method according to claim 1, wherein the clustering of the historical electricity price data into m classes in a way without supervision of a mentor through a K-means algorithm specifically comprises:
randomly selecting m samples from the historical electricity price data as initial clustering centers cj(j1，2，...，m)；
For each sample x in the historical electricity price dataiCalculating Euclidean distances between the samples and m clustering centers, and classifying the samples into a class corresponding to the clustering center with the minimum distance;
aiming at a new class formed after the sample is divided, according to a calculation formula:
recalculating cluster centers, where | cj| represents class cjThe number of samples in (1);
and repeating the two steps until the position of the clustering center is not changed.
5. The electricity price prediction method according to claim 1, wherein training and constructing an electricity price pattern recognition model according to historical electricity price data with pattern labels specifically comprises:
establishing an electricity price mode identification model;
extracting the characteristics of the historical electricity price data with the mode label to obtain a characteristic vector;
and training a power price mode recognition model based on the characteristic vector and the mode label by adopting a machine learning method so as to optimize the recognition accuracy of the power price mode recognition model.
6. The electricity price prediction method according to claim 1, wherein after training and constructing the electricity price pattern recognition model according to the historical electricity price data with the pattern labels, the method further comprises:
establishing a day-ahead electricity price prediction model based on an RBF neural network according to corresponding historical electricity price data;
and obtaining a power price prediction sequence of a prediction day based on a day-ahead power price prediction model of the RBF neural network.
7. The electricity price prediction method according to claim 6, wherein the algorithm in the RBF neural network-based day-ahead electricity price prediction model is as follows:
in the formula: p1, 2, P, and P is the number of samples; w is aqrRepresenting the connection weight between the hidden layer and the output layer; h is the number of nodes of the hidden layer; y isrRepresenting the output result of the input sample corresponding to the r node of the network; II xp-cqII is the Euclidean norm; c. CqAnd σ represents the center and variance of the gaussian function, respectively; x is the number ofp＝x1p，x2p，...，xkpRepresenting the P input vector; k is the number of nodes of the input layer.
8. An electricity rate prediction system, characterized in that the electricity rate prediction system comprises: a memory and a processor, the memory including a power rate prediction method program, the power rate prediction method program when executed by the processor implementing the steps of:
gathering the historical electricity price data into m types in a mode without supervision of a guide through a K-means algorithm, respectively representing the m types by using labels l, 2, … and m, and obtaining the historical electricity price data with the mode labels;
training and constructing a power price mode recognition model according to historical power price data with mode labels;
selecting n electricity price prediction methods and establishing n independent day-ahead electricity price prediction models based on the same historical electricity price data;
respectively obtaining n electricity price prediction sequences of a prediction day by n independent day-ahead electricity price prediction models;
inputting the n electricity price prediction sequences into the electricity price mode recognition model, and outputting the prediction results of the n electricity price daily fluctuation modes;
and calculating the final voting score of each power rate daily fluctuation mode according to the prediction results of the n power rate daily fluctuation modes, and selecting the power rate daily fluctuation mode with the highest score as the final predicted power rate daily fluctuation mode.
9. The method for predicting electricity prices according to claim 1, wherein calculating the final voting score of each electricity price daily fluctuation mode according to the prediction results of the n electricity price daily fluctuation modes specifically comprises:
respectively calculating the voting weight of different day-ahead electricity price prediction models to each electricity price daily fluctuation mode according to a weighted voting algorithm;
respectively accumulating the weights voted by different day-ahead electricity price prediction models for each electricity price daily fluctuation mode, and obtaining the final voting score of each electricity price daily fluctuation mode;
the expression of the weighted voting algorithm is as follows:
in the formula, WX，YRepresenting the voting weight of the day-ahead electricity price prediction model Y to the electricity price daily fluctuation mode X; d represents the number of times of independent operation of the algorithm; cX，Y，dThe number of samples for which the day-ahead electricity price prediction model Y predicts the electricity price daily fluctuation mode X correctly in the d-th operation is represented; t isX，dAnd (4) representing that all day-ahead power rate prediction models predict the correct number of samples for the power rate daily fluctuation mode X in the d-th operation.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium includes a power rate prediction method program which, when executed by a processor, implements the steps of a power rate prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911157776.3A CN111191814A (en) | 2019-11-22 | 2019-11-22 | Electricity price prediction method, system and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911157776.3A CN111191814A (en) | 2019-11-22 | 2019-11-22 | Electricity price prediction method, system and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111191814A true CN111191814A (en) | 2020-05-22 |
Family
ID=70709156
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911157776.3A Withdrawn CN111191814A (en) | 2019-11-22 | 2019-11-22 | Electricity price prediction method, system and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111191814A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111695742A (en) * | 2020-06-22 | 2020-09-22 | 南方电网科学研究院有限责任公司 | Day-ahead scheduling interval optimization method, system and equipment for comprehensive energy system |
| CN112232886A (en) * | 2020-10-30 | 2021-01-15 | 南方电网能源发展研究院有限责任公司 | Electricity price probability prediction method, system, computer equipment and storage medium |
| CN112308335A (en) * | 2020-11-12 | 2021-02-02 | 南方电网能源发展研究院有限责任公司 | Short-term electricity price prediction method and device based on xgboost algorithm |
| CN112348597A (en) * | 2020-11-27 | 2021-02-09 | 新智数字科技有限公司 | Power price prediction method and device based on least square support vector machine model |
| CN115249166A (en) * | 2021-12-20 | 2022-10-28 | 国家电投集团电站运营技术（北京）有限公司 | Method and device for forecasting discharged electricity price, computer equipment and storage medium |
| CN116028838A (en) * | 2023-01-09 | 2023-04-28 | 广东电网有限责任公司 | Clustering algorithm-based energy data processing method and device and terminal equipment |
-
2019
- 2019-11-22 CN CN201911157776.3A patent/CN111191814A/en not_active Withdrawn
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111695742A (en) * | 2020-06-22 | 2020-09-22 | 南方电网科学研究院有限责任公司 | Day-ahead scheduling interval optimization method, system and equipment for comprehensive energy system |
| CN112232886A (en) * | 2020-10-30 | 2021-01-15 | 南方电网能源发展研究院有限责任公司 | Electricity price probability prediction method, system, computer equipment and storage medium |
| CN112232886B (en) * | 2020-10-30 | 2021-06-04 | 南方电网能源发展研究院有限责任公司 | Electricity price probability prediction method, system, computer equipment and storage medium |
| CN112308335A (en) * | 2020-11-12 | 2021-02-02 | 南方电网能源发展研究院有限责任公司 | Short-term electricity price prediction method and device based on xgboost algorithm |
| CN112348597A (en) * | 2020-11-27 | 2021-02-09 | 新智数字科技有限公司 | Power price prediction method and device based on least square support vector machine model |
| CN115249166A (en) * | 2021-12-20 | 2022-10-28 | 国家电投集团电站运营技术（北京）有限公司 | Method and device for forecasting discharged electricity price, computer equipment and storage medium |
| CN116028838A (en) * | 2023-01-09 | 2023-04-28 | 广东电网有限责任公司 | Clustering algorithm-based energy data processing method and device and terminal equipment |
| CN116028838B (en) * | 2023-01-09 | 2023-09-19 | 广东电网有限责任公司 | Clustering algorithm-based energy data processing method and device and terminal equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111191814A (en) | Electricity price prediction method, system and computer readable storage medium | |
| Abedinia et al. | A new feature selection technique for load and price forecast of electrical power systems | |
| Kwon et al. | A hybrid neurogenetic approach for stock forecasting | |
| CN112085565B (en) | Deep learning-based information recommendation method, device, equipment and storage medium | |
| CN110046634B (en) | Interpretation method and device of clustering result | |
| CN116523277B (en) | Intelligent energy management method and system based on demand response | |
| CN116028838B (en) | Clustering algorithm-based energy data processing method and device and terminal equipment | |
| Hong et al. | Locational marginal price forecasting using deep learning network optimized by mapping-based genetic algorithm | |
| CN109902823A (en) | A kind of model training method and equipment based on generation confrontation network | |
| Bernardo et al. | A genetic type-2 fuzzy logic based system for financial applications modelling and prediction | |
| CN113706291A (en) | Fraud risk prediction method, device, equipment and storage medium | |
| CN108921693A (en) | A kind of deriving method of data, device and equipment | |
| CN115718826A (en) | Method, system, device and medium for classifying target nodes in graph structure data | |
| Zhang et al. | Locational marginal price forecasting using convolutional long-short term memory-based generative adversarial network | |
| CN113886697A (en) | Clustering algorithm based activity recommendation method, device, equipment and storage medium | |
| CN114037182A (en) | Building load prediction model training method and device and nonvolatile storage medium | |
| CN110232397A (en) | A kind of multi-tag classification method of combination supporting vector machine and projection matrix | |
| CN111932302A (en) | Method, device, equipment and system for determining number of service sites in area | |
| CN117669834A (en) | Traffic flow prediction method, device and storage medium | |
| CN117592595A (en) | Method and device for building and predicting load prediction model of power distribution network | |
| CN115982646B (en) | Management method and system for multisource test data based on cloud platform | |
| CN110309857A (en) | Book classification device, method, equipment and storage medium based on artificial intelligence | |
| Sinaga et al. | Stock trend prediction using SV-kNNC and SOM | |
| CN113920374A (en) | Random forest based user classification method, device, equipment and storage medium | |
| Zhang et al. | Load prediction based on depthwise separable convolution model |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200522 |
|
| WW01 | Invention patent application withdrawn after publication |