CN115034485A - Wind power interval prediction method and device based on data space - Google Patents
Wind power interval prediction method and device based on data space Download PDFInfo
- Publication number
- CN115034485A CN115034485A CN202210686617.8A CN202210686617A CN115034485A CN 115034485 A CN115034485 A CN 115034485A CN 202210686617 A CN202210686617 A CN 202210686617A CN 115034485 A CN115034485 A CN 115034485A
- Authority
- CN
- China
- Prior art keywords
- wind power
- prediction
- data
- meteorological
- moment
- 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 78
- 238000013135 deep learning Methods 0.000 claims abstract description 30
- 238000012795 verification Methods 0.000 claims description 69
- 238000012549 training Methods 0.000 claims description 31
- 238000010200 validation analysis Methods 0.000 claims description 18
- 238000003860 storage Methods 0.000 claims description 16
- 230000002159 abnormal effect Effects 0.000 claims description 14
- 238000007781 pre-processing Methods 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 9
- 238000004140 cleaning Methods 0.000 claims description 8
- 125000004432 carbon atom Chemical group C* 0.000 claims description 4
- 238000000354 decomposition reaction Methods 0.000 description 19
- 238000004364 calculation method Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 11
- 230000015654 memory Effects 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 9
- 230000000694 effects Effects 0.000 description 9
- 238000011156 evaluation Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 8
- 108090000672 Annexin A5 Proteins 0.000 description 6
- 241000282461 Canis lupus Species 0.000 description 6
- 238000009826 distribution Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000005457 optimization Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 230000006872 improvement Effects 0.000 description 4
- 238000003062 neural network model Methods 0.000 description 4
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 3
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 210000004027 cell Anatomy 0.000 description 3
- 238000013500 data storage Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000002203 pretreatment Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000006403 short-term memory Effects 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 238000010220 Pearson correlation analysis Methods 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 2
- 230000001364 causal effect Effects 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 101100542977 Arabidopsis thaliana PIPC gene Proteins 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- IVBHGBMCVLDMKU-GXNBUGAJSA-N piperacillin Chemical compound O=C1C(=O)N(CC)CCN1C(=O)N[C@H](C=1C=CC=CC=1)C(=O)N[C@@H]1C(=O)N2[C@@H](C(O)=O)C(C)(C)S[C@@H]21 IVBHGBMCVLDMKU-GXNBUGAJSA-N 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 238000012360 testing method 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Primary Health Care (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Power Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of wind power prediction, and particularly provides a wind power interval prediction method and device based on a data space, wherein the method comprises the following steps: acquiring meteorological feature data related to wind power at a prediction moment; inputting the meteorological feature data related to the wind power at the prediction moment into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment; and determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment. The invention provides a scientific, reasonable, practical, efficient, accurate and reliable method for predicting the short-term wind power point and the interval of the wind power plant, which can realize the short-term power point prediction of the wind power plant for 24h in the future and the interval prediction under different confidence degrees and provide decision support for the dispatching operation of a power system and the report participation of the wind power plant in the power market.
Description
Technical Field
The invention relates to the technical field of wind power prediction, in particular to a wind power interval prediction method and device based on a data space.
Background
Because wind power has randomness, volatility and intermittence and has obvious peak-reverse regulation characteristics, the rapid development of the wind power brings serious challenges to the planning, dispatching and safe operation of a power system and the economic and reasonable consumption of the wind power. After a large-scale wind power plant is connected to a power grid, the intermittency and fluctuation of wind power bring great pressure to the power and electric quantity balance and frequency stability regulation of the power grid. In order to deal with the uncertainty of large-scale intermittent wind power access to the power grid, accurate wind power prediction is increasingly important. The short-term wind power point and interval prediction results are main reference data for making a power generation plan and adjusting an operation strategy by a power grid dispatching management department, and are also important bases for wind power enterprises to report and quote to participate in the electric power spot market. In addition, accurate and reliable short-term wind power point and interval prediction is also beneficial to reducing technical and economic risks brought by uncertainty of wind power and faced by power market participants, the income of wind power enterprises participating in power market transactions is improved, and new energy development and clean low-carbon transformation of a power system are further promoted.
With the arrival of the 5G era, new generation information technologies such as big data, cloud service, artificial intelligence and the like are widely applied, and reliable data quality and strong computing power support can be provided for short-term wind power prediction. Massive multi-source, multi-dimensional And multi-mode Data resources are collected And stored by platforms such as an SCADA (supervisory Control And Data acquisition), a Numerical Weather Prediction (NWP) System And a Geographic Information System (GIS) through Data collection And monitoring Control, And a basis is provided for construction of a Data space of a wind power plant. The data space inherits a new data management concept different from the traditional big data management and is changed from service-oriented to object-oriented. The data space is a bottom-layer framework of distributed multi-element label data storage of a full-object-oriented full life cycle, and is a technical system for safe and efficient connection of data. The method conforms to the trend of transformation construction of the digital wind power plant, and the construction of the data space of the wind power plant is an important means for improving the digital operation and maintenance management level of the wind power plant. Through a data storage technology and related services provided by a data space, a wind power plant can systematically manage a plurality of data sources in a comprehensive, safe and efficient mode, and can selectively extract useful information, so that the utilization degree of the data is effectively improved, and decision support is provided for wind power prediction and wind power plant operation and maintenance management.
Based on different modeling ideas, wind power prediction methods can be roughly divided into physical methods, statistical methods, artificial intelligence and machine learning methods and combined optimization methods. However, the physical method has high requirements on the accuracy and the integrity of the NWP data, and requires explicit mathematical equation description on atmospheric physical characteristics, wind turbine characteristics and the like, and is generally not suitable for short-term power prediction. Statistical models, artificial intelligence and machine learning models, especially mixed multi-stage models, have wide application in the field of short-term prediction of wind power. However, with the construction of information platforms and the improvement of big data technologies, the data scale of wind farms is continuously enlarged and the data types are more diversified. And the traditional statistics and machine learning models are difficult to describe the mapping relation between complex wind power fluctuation and multi-dimensional meteorological data, so that the improvement of point prediction precision is limited. In addition, in practical engineering application, due to risks caused by strong fluctuation of wind power, wind power interest correlators put forward new requirements for power probability interval prediction.
In conclusion, certain differences exist in the aspects of the existing technologies of cleaning and feature mining of multidimensional data space data of a wind power plant, intelligent decomposition and noise reduction of wind power time series data, and comprehensive application of a point prediction method and an interval prediction method, and the advanced deep learning algorithm is not applied in engineering practice, so that the further improvement of the wind power prediction precision is limited to a certain extent.
Disclosure of Invention
In order to overcome the defects, the invention provides a wind power interval prediction method and device based on a data space.
In a first aspect, a wind power interval prediction method based on a data space is provided, and the wind power interval prediction method based on the data space includes:
acquiring meteorological characteristic data related to wind power at a prediction moment;
inputting the meteorological feature data related to the wind power at the prediction moment to a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
and determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
Preferably, before the obtaining of the meteorological feature data related to the wind power at the predicted time, the method includes:
acquiring historical power data and meteorological characteristic data of a wind power plant;
preprocessing the historical power data and the meteorological feature data of the wind power plant;
and calculating the correlation between the historical power data and various meteorological feature data, and screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data.
Further, the pre-treatment comprises at least one of the following: abnormal value identification, data cleaning and wind speed data preprocessing.
Further, the calculating the correlation between the historical power data and the various meteorological feature data, and the screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data comprises:
calculating a Pearson correlation coefficient among various meteorological characteristic data;
if the Pearson correlation coefficient among various meteorological characteristic data exceeds a first threshold value, one meteorological characteristic data is removed;
calculating a Pearson correlation coefficient and a gray correlation degree between historical power data and various meteorological characteristic data;
and if the Pearson correlation coefficient between the meteorological feature data and the historical power data is significant under the preset significance level, and the gray correlation degree between the meteorological feature data and the historical power data exceeds a second threshold value, the meteorological feature corresponding to the meteorological feature data is the meteorological feature related to the wind power.
Preferably, the training process of the pre-constructed deep learning wind power prediction model includes:
constructing a training set and a verification set by using an inherent modal component and a residual component obtained after the historical power data of the wind power plant are decomposed and meteorological feature data related to the wind power;
and training an initial deep learning wind power prediction model by utilizing the training set and the verification set to obtain the pre-constructed deep learning wind power prediction model.
Further, the determining the wind power prediction interval at the prediction time based on the wind power prediction value at the prediction time includes:
inputting meteorological feature data related to the wind power in the verification set into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value of each sample point of the verification set;
taking the difference between the wind power predicted value of each sample point in the verification set and the historical wind power of each sample point in the verification set as a wind power prediction error of each sample point in the verification set;
determining a relative error sample sequence of the verification set based on the wind power prediction value of each sample point of the verification set and the wind power prediction error of each sample point of the verification set;
obtaining a confidence upper limit and a confidence lower limit corresponding to the relative error sample sequence of the verification set based on a nuclear density estimation method;
and determining the wind power prediction interval at the prediction moment based on the confidence upper limit and the confidence lower limit corresponding to the relative error sample sequence of the verification set.
Further, the calculation formula of the relative error sample sequence of the verification set is as follows:
e p ＝e r /p v
in the above formula, e p For the relative error sample sequence of the validation set, e r Predicting an error sequence, p, for the wind power at each sample point of the validation set v And the wind power prediction value of each sample point of the verification set is obtained.
Further, the calculation formula of the wind power prediction interval at the prediction time is as follows:
in the above formula, the first and second carbon atoms are,
in order to predict the wind power prediction interval at the moment,
is the upper limit of the wind power predicted value at the predicted moment under the confidence level mu,
and the lower limit of the wind power predicted value at the predicted moment under the confidence level mu is set.
Further, the calculation formula of the upper limit of the wind power predicted value at the predicted time under the confidence level μ is as follows:
the lower limit of the wind power predicted value at the predicted time under the confidence level mu is calculated according to the following formula:
in the above formula, p i The wind power predicted value at the predicted time is the wind power predicted value,
an upper limit for the confidence level mu corresponding to the sequence of relative error samples of the validation set,
the lower limit of the confidence level mu corresponding to the relative error sample sequence of the validation set.
In a second aspect, a data space-based wind power interval prediction device is provided, and the data space-based wind power interval prediction device includes:
the acquisition module is used for acquiring meteorological feature data related to wind power at a prediction moment from a wind power plant data space;
the first determination module is used for inputting the meteorological feature data related to the wind power at the prediction moment into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
and the second determination module is used for determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
In a third aspect, a computer device is provided, comprising: one or more processors;
the processor to store one or more programs;
the one or more programs, when executed by the one or more processors, implement the data space-based wind power interval prediction method.
In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed, implements the data space-based wind power interval prediction method.
One or more technical schemes of the invention at least have one or more of the following beneficial effects:
the invention provides a wind power interval prediction method and device based on a data space, which comprises the following steps: acquiring meteorological feature data related to wind power at a prediction moment; inputting the meteorological feature data related to the wind power at the prediction moment to a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment; and determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment. The invention provides a scientific, reasonable, practical and efficient accurate point prediction and interval prediction method for short-term wind power of a wind power plant, which can realize the short-term power point prediction of the wind power plant for 24h in the future and the interval prediction under different confidence degrees and provide decision support for the dispatching operation of a power system and the report participation of the wind power plant in a power market.
Drawings
FIG. 1 is a flow chart illustrating main steps of a wind power interval prediction method based on a data space according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an LSTM neural network model according to an embodiment of the present invention;
FIG. 3 is a detailed step flow diagram of a wind power interval prediction method based on data space according to an embodiment of the present invention;
FIG. 4 is a graph of the results of a training set wind power data sequence after metamorphic mode decomposition according to an embodiment of the present invention;
FIG. 5 is a prediction curve and a relative error diagram of a wind power point prediction method based on BilSTM-Attention for predicting collected wind power according to an embodiment of the present invention;
FIG. 6 is a prediction interval diagram of the prediction collection wind power under different confidence levels calculated by the wind power interval prediction method based on the data space and based on the nuclear density estimation in the embodiment of the invention;
fig. 7 is a main structural block diagram of a wind power interval prediction apparatus based on a data space according to an embodiment of the present invention.
Detailed Description
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Aiming at the defects of the existing wind power prediction technology, the invention aims to provide a scientific, reasonable, practical, efficient, accurate and reliable wind power plant short-term wind power interval prediction method by combining the data storage management technology of a data space and the strong nonlinear data fitting capacity of a deep learning algorithm, can realize the short-term power point prediction (the time resolution is within 15 min) of the wind power plant for 24h in the future and the interval prediction under different confidence degrees, and provides decision support for the scheduling operation of a power system and the participation of the wind power plant in the power market.
Due to redundant characteristic dimensions, a prediction model is too complex, and the improvement of the prediction efficiency and the prediction precision of the model is limited. Based on key feature screening of Pearson correlation analysis and grey correlation analysis, the causal association relationship between input features and wind power can be preliminarily mined from multi-source and multi-dimensional wind power influence factor data. By screening key characteristics influencing wind power, effective information resources in data can be fully mined, prior knowledge is added for model training, and therefore model prediction accuracy and goodness of fit are improved.
Because the original wind power sequence has strong randomness and volatility, the accuracy and the stability of a prediction model are adversely affected. When the wind power sequence signal is decomposed by using the variational modal decomposition, the value of the modal decomposition total number and the secondary punishment coefficient directly influences the final effect of the original signal decomposition. The grey wolf algorithm optimized variation modal decomposition method can achieve optimized values of two parameters of modal decomposition total number and secondary punishment coefficient, and further decomposes the wind power signal with strong randomness and volatility into stable components with different frequencies, so that the prediction stability and accuracy of the model are improved.
Because the precision of the current wind power point prediction method needs to be further improved, the method constructs a BilSTM-Attention-based deep learning model to realize short-term wind power point prediction. By means of the sensitivity of a Long short-term memory neural network (LSTM) model to time sequence data, a Bidirectional Long short-term memory network (BilTM) can give consideration to information of time moments before and after the time sequence data, can capture information ignored by the unidirectional LSTM, more fully excavates potential information in the data sequence, and enhances the capability of the network to process complex situations. The BiLSTM model is used for predicting each component of the wind power, the time sequence characteristics of data are strengthened, the causal association relation between the input characteristics and the wind power is deeply mined, and the model prediction precision can be further improved. Attention (Attention) mechanisms can use limited computing resources to process more important information. The invention introduces a soft attention mechanism into the BilSTM, improves the defect that the BilSTM network loses important information due to overlong time sequence data by using a probability distribution weight mode, and highlights the influence of the important information in key characteristics, thereby improving the prediction precision and the prediction efficiency of the model.
Most of traditional wind power prediction research focuses on improving point prediction accuracy, and additional information in prediction errors is ignored. The interval prediction can provide more information for a power system decision maker in practical engineering application by describing the upper and lower boundaries of the wind power prediction curve, and is beneficial to improving the stability of a power grid and reducing the operation cost. In order to improve the decision reference value of the wind power prediction result, valuable information needs to be extracted from massive multi-source high-dimensional data, the non-stationarity of a wind power sequence is reduced, the quality of a model training set is improved, and the accuracy of wind power point prediction and interval prediction is further improved.
Example 1
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating main steps of a wind power interval prediction method based on a data space according to an embodiment of the present invention. As shown in fig. 1, the wind power interval prediction method based on the data space in the embodiment of the present invention mainly includes the following steps:
step S101: acquiring meteorological feature data related to wind power at a prediction moment;
step S102: inputting the meteorological feature data related to the wind power at the prediction moment to a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
step S103: and determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
In this embodiment, in step S101, information such as historical power data and meteorological features (including wind speed, wind direction, temperature, humidity, air pressure, and the like) of the wind farm may be acquired through an SCADA, NWP, and other information systems, a meteorological feature data value of a day in the future may be acquired through an NWP system, and then data acquired from different platforms (data with consistent time resolution) are aggregated, integrated, and stored in a wind farm data space according to a unified data standard, specification, and protocol for unified management, so as to provide a safe, stable, and efficient data support service for wind power prediction.
In this embodiment, before the step S101, the method includes:
acquiring historical power data and meteorological characteristic data of a wind power plant;
preprocessing the historical power data and the meteorological feature data of the wind power plant;
and calculating the correlation between the historical power data and various meteorological feature data, and screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data.
Wherein the pre-treatment comprises at least one of: abnormal value identification, data cleaning and wind speed data preprocessing.
In one embodiment, to avoid that the data outliers and missing values in the historical data set affect the accuracy of the data-driven-based model, the wind power and its influencing factor (feature) data set in the wind farm data space need to be preprocessed, specifically:
step 1: and (4) abnormal value identification. The situation that an abnormal value exists in an original wind power plant data set can be mainly divided into the following categories: the wind speed is greater than the cut-in wind speed, but the power value is 0; the wind power is greater than 0, but the wind speed is recorded as 0; sampling faults of the data sensor, abnormal data acquisition or data recording loss and the like. For the missing data record of the whole piece, delete it from the sample, carry on further processing to missing value and abnormal value.
Step 2: and (6) data cleaning. The data mean value of adjacent sampling time intervals before and after the missing data value is taken as a supplement value of the missing value or as a substitute value of an abnormal value in consideration of certain continuity of meteorological and power data change.
And step 3: and (4) preprocessing wind speed data. The range of the wind direction is 0-360 degrees, in physical sense, 0 degrees, 360 degrees, 1 degrees and 359 degrees are basically equivalent to the output of the fan, but the difference of the input numerical values is very large for the deep learning method based on data driving, therefore, trigonometric functional processing is adopted for the wind direction, namely, the sin value of the wind direction is taken to replace the original wind direction data.
And 4, step 4: and obtaining a data sample after abnormal value identification, data cleaning and wind speed data preprocessing, and dividing a training set, a verification set and a verification set.
Further, the calculating the correlation between the historical power data and the various meteorological feature data, and the screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data comprises:
calculating a Pearson correlation coefficient among various meteorological characteristic data;
if the Pearson correlation coefficient among various meteorological characteristic data exceeds a first threshold value, one meteorological characteristic data is removed;
calculating a Pearson correlation coefficient and a gray correlation degree between historical power data and various meteorological characteristic data;
and if the Pearson correlation coefficient between the meteorological feature data and the historical power data is significant under the preset significance level, and the gray correlation degree between the meteorological feature data and the historical power data exceeds a second threshold value, the meteorological feature corresponding to the meteorological feature data is the meteorological feature related to the wind power.
In one embodiment, wind power influence factors are analyzed by using the Pearson correlation coefficient and the grey correlation degree, key features are screened, and data redundancy is reduced; in the embodiment provided by the invention, the first threshold is set to 0.95, the preset significance level is 0.01, and the second threshold is set to 0.80.
In this embodiment, the training process of the pre-constructed deep learning wind power prediction model includes:
constructing a training set and a verification set by using an inherent modal component and a residual component obtained after the historical power data of the wind power plant are decomposed and meteorological feature data related to the wind power;
and training an initial deep learning wind power prediction model by utilizing the training set and the verification set to obtain the pre-constructed deep learning wind power prediction model.
Specifically, in the specific process of wind power sequence Decomposition, a Grey Wolf Optimization algorithm (GWO) optimized Variational Modal Decomposition (VMD) method is used to decompose a historical wind power sequence, so as to obtain Intrinsic Modal Functions (IMFs) and a Residual component (RES) of a plurality of different frequencies.
By utilizing a gray wolf optimization algorithm, taking the local envelope entropy minimum value of each modal component as a fitness function, and carrying out modal decomposition on the wind power sequence to obtain a total K and a secondary punishment coefficient alpha 0 And optimizing. The envelope entropy calculation formula of the modal component x (j), j ═ 1,2, …, n is as follows.
Wherein a (j) is a value obtained by a metamorphic modal decomposition 0 The modal component x (j) is Hilbert demodulatedThe resulting envelope signal, p j Is a (j) normalized form.
The optimal modal decomposition total K and the secondary penalty coefficient alpha are obtained by optimizing a gray wolf optimization algorithm 0 And finally, calculating a plurality of natural modal components (IMFs) and a residual component (RES) after the wind power sequence is decomposed as parameters of the variation modal decomposition.
In one embodiment, in the training process of the pre-constructed deep learning wind power prediction model, a wind farm historical data set is divided into a training set and a verification set, the deep learning prediction model based on BilSTM-Attention (bidirectional long short term memory neural network) is trained by the training set, the hyper-parameters of the deep learning model are determined by the verification set, and the prediction error of the verification set is recorded. And (3) normalizing all input data by adopting a min-max (maximum-minimum) normalization method in the prediction model, and outputting a model prediction result after performing inverse normalization. And inputting the key characteristic data of the wind power plant to be predicted in the future day into the BilSTM-Attention model trained by the history data set to obtain a point prediction result of the wind power.
Specifically, an LSTM neural network model is established, as shown in fig. 2.
i t ＝σ(W i x t +U i h t-1 +b i )
f t ＝σ(W f x t +U f h t-1 +b f )
o t ＝σ(W o x t +U o h t-1 +b o )
h t ＝o t ⊙tanh(c t )
Where σ denotes a Sigmoid activation function, W i ，W f And W and o respectively indicating weight vectors from an input layer to an input gate, a forgetting gate, an output gate and a cell state; u shape i ，U f ，U o Respectively indicating weight vectors from the hidden layer to the input gate, the forgetting gate, the output gate and the cell state; b i ，b f The reference numerals refer to the input gate, the forgetting gate, the output gate, and the bias of the cell state, respectively.
On the basis, a BilSTM neural network model is established. the specific calculation formula of the total output value h' of the BilSTM calculation unit at the time t is as follows:
in the formula (I), the compound is shown in the specification,
and
respectively representing the output values of the forward LSTM unit and the backward LSTM unit,
is a vector stitching operation.
A soft Attention mechanism is introduced into the BilSTM, the defect that the BiLSTM network loses important information due to overlong time sequence data is overcome by replacing a random weight distribution mode with a probability weight distribution mode, and therefore the prediction precision and the prediction efficiency of the BilSTM-Attention model are improved.
Further, inputting the screened key characteristic sequence and the decomposed wind power sequence data into a deep learning prediction model based on the BilSTM-Attention; training the prediction model by using a training set, determining a hyper-parameter of the model based on a verification set, and recording an error of the verification set; and then inputting the key characteristic data of the prediction set into a BilSTM-Attention model trained by the history data set to respectively obtain the predicted values of all components of the power sequence, and summing all the components to obtain the point prediction result of the wind power.
In one embodiment, the invention further performs wind power point prediction result evaluation, specifically as follows:
selecting average Absolute Error (MAE), average Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE) and R 2 (R-squared) as an evaluation index of the model point prediction result.
Further, the smaller the average absolute error, the average absolute percentage error and the root mean square error are, the higher the model prediction accuracy is, and the better the point prediction effect is; r 2 The closer to 1, the better the fitting degree of the predicted power curve and the actual power curve, and the better the point prediction effect.
In this embodiment, calculating confidence upper limits and confidence lower limits of verification set relative error sample sequences under different confidence degrees by using a nuclear density estimation method, and calculating confidence upper limits and confidence lower limits of wind power predicted values under different confidence degrees by combining point prediction results at a time to be predicted, so as to obtain a wind power prediction interval, where determining the wind power prediction interval at the prediction time based on the wind power predicted value at the prediction time includes:
inputting meteorological feature data related to the wind power in the verification set into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value of each sample point of the verification set;
taking the difference between the wind power predicted value of each sample point in the verification set and the historical wind power of each sample point in the verification set as a wind power prediction error of each sample point in the verification set;
determining a relative error sample sequence of the verification set based on the wind power prediction value of each sample point of the verification set and the wind power prediction error of each sample point of the verification set;
obtaining a confidence upper limit and a confidence lower limit corresponding to the relative error sample sequence of the verification set based on a nuclear density estimation method;
and determining the wind power prediction interval at the prediction moment based on the confidence upper limit and the confidence lower limit corresponding to the relative error sample sequence of the verification set.
In one embodiment, the relative error sample sequence of the validation set is calculated as follows:
e p ＝e r /p v
in the above formula, e p For the relative error sample sequence of the validation set, e r Predicting an error sequence, p, for the wind power at each sample point of the validation set v And the wind power prediction value of each sample point of the verification set is obtained.
In one embodiment, the calculation formula of the wind power prediction section at the prediction time is as follows:
in the above formula, the first and second carbon atoms are,
in order to predict the wind power prediction interval at the moment,
is the upper limit of the wind power predicted value at the predicted moment under the confidence level mu,
and the lower limit of the wind power predicted value at the predicted moment under the confidence level mu is set.
In one embodiment, the upper limit of the wind power predicted value at the predicted time under the confidence level μ is calculated as follows:
the lower limit of the wind power predicted value at the predicted time under the confidence level mu is calculated according to the following formula:
in the above formula, p i The wind power predicted value at the predicted time is the wind power predicted value,
an upper limit for the confidence level mu corresponding to the sequence of relative error samples of the validation set,
the lower limit of the confidence level mu corresponding to the relative error sample sequence of the validation set.
In one embodiment, for the wind power Interval Prediction effect evaluation based on the data space, a Prediction Interval Coverage (PICP), a Prediction Interval normalized average bandwidth (PINAW), and a comprehensive reliability index (CWC) are selected for comparative analysis.
The prediction interval coverage rate index reflects the probability that the actual wind power falls in the prediction interval under the specified confidence coefficient, and the PICP is smaller than the confidence coefficient mu, which indicates that the prediction is invalid; otherwise, the prediction is valid. The greater the PICP, the greater the probability that the actual power falls between the upper and lower predicted limits, which is expressed as follows.
Wherein N is the number of predicted samples, c i Values for boolean variables are shown below.
The average bandwidth index of the prediction interval can reflect the average value of the width between the upper limit and the lower limit of the prediction interval, and when the PICPs of prediction results are the same, a smaller PINAW corresponds to a better prediction effect. Specifically, the method can be represented as follows:
wherein Z is a variation interval of the wind power value.
Since the PICP and the PINAW reflect only one-sided evaluation criteria, the comprehensive expression of the prediction result cannot be embodied. Therefore, a comprehensive reliability evaluation index (CWC) is introduced, and the calculation formula is as follows:
CWC＝PINAW[1+γ PIPC e -η(PIPC-μ) ],η>0
in the formula, eta is a parameter larger than 0, and 1 is selected in the invention; μ is a given confidence level.
Further, the present invention provides a specific embodiment, as shown in fig. 3, specifically comprising the following implementation steps:
step 1: and acquiring and storing wind power plant data. Taking a wind power plant with 150MW installed capacity in northern China as an example, wind power data and meteorological characteristic data in an SCADA platform of the wind power plant are stored in a data space of the wind power plant. The wind power plant data set comprises wind power, 10m wind speed, 10m wind direction, 30m wind speed, 30m wind direction, 50m wind speed, 50m wind direction, 70m wind speed, hub wind direction, hub wind speed, pressure, temperature, humidity and other meteorological characteristics, and the data sampling time interval is 5 min. And selecting actual measurement historical data and power data of the wind power plant from 2013, 5, 15 days to 2018, 5, 31 days to train and test the method provided by the invention.
And 2, step: and (4) preprocessing data. And carrying out abnormal value identification, data cleaning and wind speed data preprocessing on the acquired data set.
Step 2.1: for the absence of the entire data record, it is removed from the sample.
Step 2.2: for individual missing data values, taking the data mean value of the adjacent moments before and after the missing data value as a complementary value of the missing value; for individual abnormal data values, the data mean value of the adjacent time before and after the abnormal data value is taken as the substitute value of the abnormal value.
Step 2.3: and (3) performing trigonometric functional processing on the wind direction, namely, taking the sin value of the wind direction to replace the original wind direction data.
Step 2.4: after data preprocessing, a total of 4420 data samples were obtained, as shown in table 1 below. Taking the first 3844 data as training samples for training a prediction model; then, taking the next 288 points as a verification set for debugging the hyper-parameters of the model and recording a prediction error; 288 points in the last day are used as verification sets for calculating the evaluation indexes of the prediction method, and the effects of the point prediction and interval prediction method provided by the invention are verified.
TABLE 1
And step 3: and screening key characteristics of wind power. And analyzing wind power influence factors by using Pearson correlation analysis and grey correlation analysis, screening key features, and reducing data redundancy. And (3) selecting 10m wind speed, 10m wind direction, 30m wind speed, 30m wind direction, 50m wind speed, 50m wind direction, wind speed at the hub, wind direction at the hub, temperature and humidity as input characteristics of the prediction model by integrating the calculation results of the Pearson correlation coefficient and the grey correlation degree.
Step 3.1: calculating a Pearson correlation coefficient among different features and a Pearson correlation coefficient among different features, wherein a preset coefficient threshold value is 0.95; and the Pearson correlation coefficient between different characteristics and the wind power, and the preset significance level is 0.01. Through calculation, the correlation between the wind speed at 70m and the wind speed at the hub is 0.993, the correlation between the wind direction at 70m and the wind direction at the hub is 0.998, and the correlation is greater than a preset threshold value of 0.95, so that the wind speed and the wind direction at 70m are deleted; and the correlation between the remaining features was less than 0.95. And after the wind speed and the wind direction at 70m are deleted, the Pearson correlation coefficient between all the remaining characteristics and the wind power is significant under the significance level of 0.01.
Step 3.2: and calculating grey correlation degrees between the different characteristics and the wind power, wherein a preset correlation degree threshold value is 0.80, and the result is shown in the following table 2. The grey correlation degree between the air pressure and the wind power is lower than a preset 0.80 threshold value, so that the grey correlation degree is eliminated.
TABLE 2
And 4, step 4: and decomposing the wind power sequence. Firstly, inputting an original wind power sequence into a variational modal decomposition model optimized by a gray wolf algorithm. The maximum iteration times of the wolf optimization algorithm are set to be 30, the initial population number is set to be 20, the parameters to be optimized have two modal decomposition total numbers and two secondary punishment coefficients, and therefore the variable dimension is set to be 2-dimensional; according to experience, the value range of the modal decomposition total number is set as [3,10], and the value range of the secondary penalty coefficient is set as [200,2000 ]. The results of the grey wolf optimization algorithm for optimizing the variational modal decomposition parameters are as follows: the modal decomposition total is 6 and the secondary penalty factor is 178.02. The result of the variational modal decomposition of the training set wind power data sequence is shown in fig. 4.
And 5: and predicting the short-term wind power point. The LSTM neural network model structure is shown in figure 2, and the hyper-parameters of the BiLSTM-Attention prediction model are set through an experimental method by utilizing data of a wind power plant training set and a verification set. The BilSTM network comprises 2 hidden layers, the number of hidden units is respectively 50 and 50, BatchSize is 64, Timestap is 36, the learning rate is 0.001, Optimizer is adam', and the iteration number is Epochs is 200; the Attention size is 64. In order to prove the effectiveness of the technical scheme provided by the invention, a BilSTM model, an LSTM model and a BP Neural Network (BPNN) model are selected to respectively predict the wind power of the prediction set, and the prediction result is compared with the wind power prediction result based on the BilSTM-Attention model. Wherein, the parameter setting of the LSTM is the same as that of the BilSTM network. The BilSTM-Attention, BilSTM, LSTM model was implemented using Python 3.8.8 and Tensorflow 2.4.1. The number of the BPNN input layer neurons is 10 of the number of the influencing factors, the number of the hidden layer neurons is 20, the number of the output layer neurons is 1, the training times is 5000, and the precision target is set to be 0.001. And respectively training different models according to the parameter settings, and finally inputting the key characteristic data of the prediction set into the trained prediction models to obtain the wind power point prediction results of the different models.
The prediction curve and relative error of the wind power point prediction method based on the BilSTM-Attention provided by the invention for predicting the wind collection power are shown in the attached figure 5.
Step 6: and evaluating the wind power point prediction result. According to the actual power values of the prediction set and the predicted values of different models, the average absolute error (MAE), the average absolute percentage error (MAPE), the Root Mean Square Error (RMSE) and the R-squared of the prediction results of different models are calculated, and the results are shown in the following table 3. According to the error evaluation index calculation result, the prediction error of the short-term wind power point prediction method based on the BilSTM-Attention model is small, the goodness of fit is high, and the point prediction effect is good.
TABLE 3
| Prediction model | MAE(MW) | MAPE | MSE(MW) | RMSE(MW) | R 2 |
| BiLSTM-Attention | 0.4788 | 3.06％ | 0.5665 | 0.7527 | 0.9995 |
| BiLSTM | 0.9556 | 4.32％ | 2.1214 | 1.4565 | 0.9983 |
| LSTM | 1.2017 | 5.23％ | 5.6712 | 2.3814 | 0.9955 |
| BPNN | 2.2655 | 12.61％ | 12.2017 | 3.4931 | 0.9902 |
And 7: and predicting the wind power interval based on the data space. Wind power actual value is collected by verification and is based on BiLSTM-Attention modelThe predicted value of the wind power of the verification set obtained by type calculation can be calculated to obtain a verification set relative error sample sequence e p ＝e r /p v ＝[e p1 ,e p2 ,…,e pn ]. Then, calculating the confidence upper limit of the verification set relative error sample sequence under different confidence levels based on the Gaussian kernel function by using a kernel density estimation method
And lower confidence limits
Finally, the wind power value p of the prediction set output according to the prediction model i Calculating the predicted power p of the point i Upper limit at different confidence levels μ
And lower limit
And then the upper and lower sections of the prediction curve are obtained.
The prediction intervals of the wind power interval prediction method based on the nuclear density estimation and based on the data space, which are provided by the invention, for the wind power collection prediction under different confidence levels are shown in the attached figure 6.
And 8: and evaluating the wind power interval prediction effect based on kernel density estimation. In order to verify the effectiveness of the wind power interval prediction method based on nuclear density estimation, a parameter estimation method assuming that error distribution obeys normal distribution is selected, and prediction intervals of the predicted wind power collection power under different confidence levels are calculated. According to the results of different prediction interval estimation methods, interval prediction effect evaluation indexes are respectively calculated, and the results are shown in the following table 4. As can be seen from table 4, the interval coverage PICP greater than the preset confidence level can be obtained by the gaussian kernel function-based kernel density estimation method at different confidence levels, and is relatively stable. Under the same confidence level, the comprehensive reliability index CWC of the kernel density estimation method based on the Gaussian kernel function is superior to the interval estimation method based on normal distribution. The method for predicting the wind power interval based on the kernel density estimation can closely follow the variation trend of the wind power sequence, and can cover more wind power actual values with narrower prediction interval width under the same confidence level.
TABLE 4
Example 2
Based on the same inventive concept, the present invention further provides a wind power interval prediction apparatus based on a data space, as shown in fig. 7, the wind power interval prediction apparatus based on a data space includes:
the acquisition module is used for acquiring meteorological feature data related to wind power at a prediction moment from a wind power plant data space;
the first determining module is used for inputting the meteorological characteristic data related to the wind power at the prediction moment into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
and the second determination module is used for determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
Preferably, before the obtaining of the meteorological feature data related to the wind power at the predicted time, the method includes:
collecting historical power data and meteorological characteristic data of a wind power plant;
preprocessing the historical power data and the meteorological feature data of the wind power plant;
and calculating the correlation between the historical power data and various meteorological characteristic data, and screening the meteorological characteristics related to the wind power based on the correlation between the historical power data and the various meteorological characteristic data.
Further, the pre-treatment comprises at least one of the following: abnormal value identification, data cleaning and wind speed data preprocessing.
Further, the calculating the correlation between the historical power data and the various meteorological feature data, and the screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data comprises:
calculating Pearson correlation coefficients among various meteorological characteristic data;
if the Pearson correlation coefficient among various meteorological characteristic data exceeds a first threshold value, one meteorological characteristic data is removed;
calculating a Pearson correlation coefficient and a gray correlation degree between historical power data and various meteorological characteristic data;
and if the Pearson correlation coefficient between the meteorological characteristic data and the historical power data is significant under a preset significance level, and the grey correlation degree between the meteorological characteristic data and the historical power data exceeds a second threshold value, the meteorological characteristic corresponding to the meteorological characteristic data is the meteorological characteristic related to the wind power.
Preferably, the training process of the pre-constructed deep learning wind power prediction model includes:
constructing a training set and a verification set by using an inherent modal component and a residual component obtained after the historical power data of the wind power plant are decomposed and meteorological feature data related to the wind power;
and training an initial deep learning wind power prediction model by utilizing the training set and the verification set to obtain the pre-constructed deep learning wind power prediction model.
Further, the determining the wind power prediction interval at the prediction time based on the wind power prediction value at the prediction time includes:
inputting meteorological feature data related to the wind power in the verification set into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value of each sample point of the verification set;
taking the difference between the wind power predicted value of each sample point in the verification set and the historical wind power of each sample point in the verification set as a wind power prediction error of each sample point in the verification set;
determining a relative error sample sequence of the verification set based on the wind power prediction value of each sample point of the verification set and the wind power prediction error of each sample point of the verification set;
obtaining a confidence upper limit and a confidence lower limit corresponding to the relative error sample sequence of the verification set based on a nuclear density estimation method;
and determining the wind power prediction interval at the prediction moment based on the confidence upper limit and the confidence lower limit corresponding to the relative error sample sequence of the verification set.
Further, the calculation formula of the relative error sample sequence of the verification set is as follows:
e p ＝e r /p v
in the above formula, e p For the relative error sample sequence of the validation set, e r Predicting an error sequence, p, for the wind power of each sample point of the validation set v And the wind power prediction value of each sample point of the verification set is obtained.
Further, the calculation formula of the wind power prediction interval at the prediction time is as follows:
in the above formula, the first and second carbon atoms are,
in order to predict the wind power prediction interval at the moment,
is the upper limit of the wind power predicted value at the predicted moment under the confidence level mu,
and the lower limit of the wind power predicted value at the predicted moment under the confidence level mu is set.
Further, the calculation formula of the upper limit of the wind power predicted value at the predicted time under the confidence level μ is as follows:
the lower limit of the wind power predicted value at the predicted time under the confidence level mu is calculated according to the following formula:
in the above formula, p i The wind power predicted value at the predicted time is the wind power predicted value,
an upper limit for the confidence level mu corresponding to the sequence of relative error samples of the validation set,
the lower limit of the confidence level mu corresponding to the relative error sample sequence of the validation set.
Example 3
Based on the same inventive concept, the present invention also provides a computer apparatus comprising a processor and a memory, the memory being configured to store a computer program comprising program instructions, the processor being configured to execute the program instructions stored by the computer storage medium. The Processor may be a Central Processing Unit (CPU), or may be 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, etc., which is a computing core and a control core of the terminal, and is specifically adapted to implement one or more instructions, and specifically adapted to load and execute one or more instructions in a computer storage medium so as to implement a corresponding method flow or a corresponding function, so as to implement the steps of the wind power interval prediction method based on the data space in the foregoing embodiments.
Example 4
Based on the same inventive concept, the present invention further provides a storage medium, in particular, a computer-readable storage medium (Memory), which is a Memory device in a computer device and is used for storing programs and data. It is understood that the computer readable storage medium herein can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer readable storage medium may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory. One or more instructions stored in the computer-readable storage medium may be loaded and executed by the processor to implement the steps of the wind power interval prediction method based on the data space in the foregoing embodiments.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (12)
1. A wind power interval prediction method based on a data space is characterized by comprising the following steps:
acquiring meteorological feature data related to wind power at a prediction moment;
inputting the meteorological feature data related to the wind power at the prediction moment into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
and determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
2. The method of claim 1, wherein said obtaining meteorological feature data relating wind power at a predicted time comprises, prior to:
acquiring historical power data and meteorological characteristic data of a wind power plant;
preprocessing the historical power data and the meteorological feature data of the wind power plant;
and calculating the correlation between the historical power data and various meteorological feature data, and screening the meteorological features related to the wind power based on the correlation between the historical power data and the various meteorological feature data.
3. The method of claim 2, wherein the pre-processing comprises at least one of: abnormal value identification, data cleaning and wind speed data preprocessing.
4. The method of claim 2, wherein calculating correlations between the historical power data and various meteorological signature data, and screening meteorological signatures related to wind power based on the correlations between the historical power data and the various meteorological signature data comprises:
calculating a Pearson correlation coefficient among various meteorological characteristic data;
if the Pearson correlation coefficient among various meteorological characteristic data exceeds a first threshold value, one meteorological characteristic data is removed;
calculating a Pearson correlation coefficient and a gray correlation degree between historical power data and various meteorological characteristic data;
and if the Pearson correlation coefficient between the meteorological feature data and the historical power data is significant under the preset significance level, and the gray correlation degree between the meteorological feature data and the historical power data exceeds a second threshold value, the meteorological feature corresponding to the meteorological feature data is the meteorological feature related to the wind power.
5. The method of claim 1, wherein the training process of the pre-built deep learning wind power prediction model comprises:
constructing a training set and a verification set by using an inherent modal component and a residual component obtained after the historical power data of the wind power plant are decomposed and meteorological feature data related to the wind power;
and training an initial deep learning wind power prediction model by utilizing the training set and the verification set to obtain the pre-constructed deep learning wind power prediction model.
6. The method of claim 5, wherein the determining the wind power prediction interval for the prediction time based on the wind power prediction value for the prediction time comprises:
inputting meteorological feature data related to the wind power in the verification set into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value of each sample point of the verification set;
taking the difference between the wind power predicted value of each sample point in the verification set and the historical wind power of each sample point in the verification set as a wind power prediction error of each sample point in the verification set;
determining a relative error sample sequence of the verification set based on the wind power prediction value of each sample point of the verification set and the wind power prediction error of each sample point of the verification set;
obtaining a confidence upper limit and a confidence lower limit corresponding to the relative error sample sequence of the verification set based on a nuclear density estimation method;
and determining the wind power prediction interval at the prediction moment based on the confidence upper limit and the confidence lower limit corresponding to the relative error sample sequence of the verification set.
7. The method of claim 6, wherein the validation set is computed as a sequence of relative error samples as follows:
e p ＝e r p v
in the above formula, e p A sequence of relative error samples for the validation set, e r Predicting an error sequence, p, for the wind power at each sample point of the validation set v And the wind power predicted value of each sample point of the verification set is obtained.
8. The method of claim 6, wherein the wind power prediction interval at the prediction time is calculated as follows:
in the above formula, the first and second carbon atoms are,
in order to predict the wind power prediction interval at the moment,
is the upper limit of the wind power predicted value at the predicted moment under the confidence level mu,
and the lower limit of the wind power predicted value at the predicted moment under the confidence level mu is set.
9. The method according to claim 8, characterized in that the upper limit of the wind power prediction value at the prediction moment at the confidence level μ is calculated as follows:
the lower limit of the wind power predicted value at the predicted time under the confidence level mu is calculated according to the following formula:
in the above formula, p i The wind power predicted value at the predicted time is the wind power predicted value,
an upper limit for the confidence level mu corresponding to the sequence of relative error samples of the validation set,
the lower limit of the confidence level mu corresponding to the relative error sample sequence of the validation set.
10. A wind power interval prediction device based on data space is characterized by comprising:
the acquisition module is used for acquiring meteorological characteristic data related to wind power at a prediction moment from a wind power plant data space;
the first determination module is used for inputting the meteorological feature data related to the wind power at the prediction moment into a pre-constructed deep learning wind power prediction model to obtain a wind power prediction value at the prediction moment;
and the second determination module is used for determining the wind power prediction interval at the prediction moment based on the wind power prediction value at the prediction moment.
11. A computer device, comprising: one or more processors;
the processor to store one or more programs;
the one or more programs, when executed by the one or more processors, implement the data space-based wind power interval prediction method of any of claims 1 to 9.
12. A computer-readable storage medium, on which a computer program is stored which, when executed, implements a data space-based wind power interval prediction method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210686617.8A CN115034485A (en) | 2022-06-16 | 2022-06-16 | Wind power interval prediction method and device based on data space |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210686617.8A CN115034485A (en) | 2022-06-16 | 2022-06-16 | Wind power interval prediction method and device based on data space |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115034485A true CN115034485A (en) | 2022-09-09 |
Family
ID=83125121
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210686617.8A Pending CN115034485A (en) | 2022-06-16 | 2022-06-16 | Wind power interval prediction method and device based on data space |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115034485A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115347571A (en) * | 2022-10-17 | 2022-11-15 | 国网江西省电力有限公司电力科学研究院 | Photovoltaic power generation power short-term prediction method and device based on transfer learning |
| CN115358495A (en) * | 2022-10-20 | 2022-11-18 | 中国华能集团清洁能源技术研究院有限公司 | Calculation method for wind power prediction comprehensive deviation rate |
| CN117269751A (en) * | 2023-11-22 | 2023-12-22 | 国网江西省电力有限公司电力科学研究院 | GIS isolating switch switching position confirmation method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921339A (en) * | 2018-06-22 | 2018-11-30 | 南京工程学院 | Genetic Support Vector Machine photovoltaic power interval prediction method based on quantile estimate |
| CN112818601A (en) * | 2021-02-05 | 2021-05-18 | 河海大学 | Hydroelectric generating set health assessment method based on GA-BP neural network and error statistical analysis |
| CN112949945A (en) * | 2021-04-15 | 2021-06-11 | 河海大学 | Wind power ultra-short-term prediction method for improving bidirectional long-short term memory network |
| CN113063872A (en) * | 2021-03-26 | 2021-07-02 | 广西中医药大学 | Fingerprint spectrum and quality analysis method of ethyl acetate part of panax notoginseng ginger |
-
2022
- 2022-06-16 CN CN202210686617.8A patent/CN115034485A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921339A (en) * | 2018-06-22 | 2018-11-30 | 南京工程学院 | Genetic Support Vector Machine photovoltaic power interval prediction method based on quantile estimate |
| CN112818601A (en) * | 2021-02-05 | 2021-05-18 | 河海大学 | Hydroelectric generating set health assessment method based on GA-BP neural network and error statistical analysis |
| CN113063872A (en) * | 2021-03-26 | 2021-07-02 | 广西中医药大学 | Fingerprint spectrum and quality analysis method of ethyl acetate part of panax notoginseng ginger |
| CN112949945A (en) * | 2021-04-15 | 2021-06-11 | 河海大学 | Wind power ultra-short-term prediction method for improving bidirectional long-short term memory network |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115347571A (en) * | 2022-10-17 | 2022-11-15 | 国网江西省电力有限公司电力科学研究院 | Photovoltaic power generation power short-term prediction method and device based on transfer learning |
| CN115358495A (en) * | 2022-10-20 | 2022-11-18 | 中国华能集团清洁能源技术研究院有限公司 | Calculation method for wind power prediction comprehensive deviation rate |
| CN115358495B (en) * | 2022-10-20 | 2023-02-07 | 中国华能集团清洁能源技术研究院有限公司 | Calculation method for wind power prediction comprehensive deviation rate |
| CN117269751A (en) * | 2023-11-22 | 2023-12-22 | 国网江西省电力有限公司电力科学研究院 | GIS isolating switch switching position confirmation method |
| CN117269751B (en) * | 2023-11-22 | 2024-04-02 | 国网江西省电力有限公司电力科学研究院 | GIS isolating switch switching position confirmation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hu et al. | Wind speed forecasting based on variational mode decomposition and improved echo state network | |
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| Chen et al. | Short-term wind speed forecasting based on long short-term memory and improved BP neural network | |
| CN112949945B (en) | Wind power ultra-short-term prediction method for improving bidirectional long-term and short-term memory network | |
| CN115034485A (en) | Wind power interval prediction method and device based on data space | |
| CN108388962B (en) | Wind power prediction system and method | |
| CN112818604A (en) | Wind turbine generator risk degree assessment method based on wind power prediction | |
| CN111027775A (en) | Step hydropower station generating capacity prediction method based on long-term and short-term memory network | |
| CN111241755A (en) | Power load prediction method | |
| CN113449919B (en) | Power consumption prediction method and system based on feature and trend perception | |
| Zhang et al. | Interval prediction of ultra-short-term photovoltaic power based on a hybrid model | |
| CN117013527A (en) | Distributed photovoltaic power generation power prediction method | |
| Li et al. | GMM-HMM-based medium-and long-term multi-wind farm correlated power output time series generation method | |
| CN117408394B (en) | Carbon emission factor prediction method and device for electric power system and electronic equipment | |
| Dumas et al. | Deep learning-based multi-output quantile forecasting of PV generation | |
| Liu et al. | A novel interval forecasting system for uncertainty modeling based on multi-input multi-output theory: A case study on modern wind stations | |
| Li et al. | Short-term probabilistic load forecasting method based on uncertainty estimation and deep learning model considering meteorological factors | |
| Qu et al. | Short-term wind farm cluster power prediction based on dual feature extraction and quadratic decomposition aggregation | |
| CN111697560B (en) | Method and system for predicting load of power system based on LSTM | |
| CN116739172A (en) | Method and device for ultra-short-term prediction of offshore wind power based on climbing identification | |
| CN114676931B (en) | Electric quantity prediction system based on data center technology | |
| CN110717623A (en) | Photovoltaic power generation power prediction method, device and equipment integrating multiple weather conditions | |
| Sun et al. | Ultra-short-term wind power interval prediction based on fluctuating process partitioning and quantile regression forest | |
| CN116187506A (en) | Short-term wind power combination probability prediction method and system considering meteorological classification | |
| CN115759343A (en) | E-LSTM-based user electric quantity prediction method and device |
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 |