CN117638926B - New energy power prediction method and device based on icing and power coupling modeling - Google Patents
New energy power prediction method and device based on icing and power coupling modeling Download PDFInfo
- Publication number
- CN117638926B CN117638926B CN202410100966.6A CN202410100966A CN117638926B CN 117638926 B CN117638926 B CN 117638926B CN 202410100966 A CN202410100966 A CN 202410100966A CN 117638926 B CN117638926 B CN 117638926B
- Authority
- CN
- China
- Prior art keywords
- icing
- prediction
- power
- model
- iing
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 230000008878 coupling Effects 0.000 title claims abstract description 16
- 238000010168 coupling process Methods 0.000 title claims abstract description 16
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 16
- 230000005684 electric field Effects 0.000 claims abstract description 14
- 238000005259 measurement Methods 0.000 claims abstract description 14
- 238000001556 precipitation Methods 0.000 claims abstract description 8
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 60
- 238000000576 coating method Methods 0.000 claims description 34
- 239000011248 coating agent Substances 0.000 claims description 32
- 238000002844 melting Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000013179 statistical model Methods 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 239000007788 liquid Substances 0.000 claims description 5
- 238000010801 machine learning Methods 0.000 claims description 5
- 230000008018 melting Effects 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 4
- 230000002194 synthesizing effect Effects 0.000 claims description 2
- 238000011160 research Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 3
- 230000008485 antagonism Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001010 compromised effect Effects 0.000 description 1
- 238000009833 condensation Methods 0.000 description 1
- 230000005494 condensation Effects 0.000 description 1
- 238000013434 data augmentation Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0475—Generative networks
-
- 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
- G06N3/094—Adversarial learning
-
- 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
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Power Engineering (AREA)
- Economics (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Wind Motors (AREA)
Abstract
The invention provides a new energy power prediction method and device based on Icing and power coupling modeling, which comprehensively obtain an Icing physical model Iing_status based on an Icing and deicing model; constructing Icing power loss Icing_loss of the whole new energy electric field; constructing a historical power prediction data set; acquiring a historical actual measurement power data set; constructing a historical Icing prediction label, calling an Icing physical model based on the temperature, the humidity, the air pressure, the precipitation and the wind speed of weather set prediction data, and generating a plurality of groups of Icing prediction state data sets Iing_loss_prediction; generating a prediction Model model_Iing; obtaining Icing_k predicted under the Icing condition; and combining with the field Iing_loss to obtain a power prediction result under the Icing condition. The method has stronger applicability, greatly reduces the uncertainty of prediction, and couples power and icing prediction online to realize high-precision power prediction under the icing condition.
Description
Technical Field
The invention belongs to the technical field of new energy power, and particularly relates to a new energy power prediction method and device based on icing and power coupling modeling.
Background
Icing has serious influence on the electric field power prediction of new energy sources. Existing researches on icing are divided into two types, one is mainly based on theoretical icing research, and the other is focused on application of icing prediction.
Based on theoretical icing research, for example, a set of models is summarized for prediction aiming at the principle research and experimental simulation of icing, and finally, the monitoring result or the prediction result of icing is applied to monitoring or prediction to carry out output breakage. Models of such studies generally rely heavily on high quality, diverse sensor data, have poor applicability to different materials and different engineering environments, and the models cannot be applied to predictions for many days in the future.
The other category focuses on the application of ice coating prediction, and the focus of the research is also on the ice coating process itself, so that various classifications and statistical analysis of measured data are carried out on the ice coating, such as ice coating processes of rime, rime or dry growth, wet growth and the like, or different ice coating mechanisms are considered according to different condensation cores to respectively model. Such studies are less costly in terms of uncertainty assessment that exists when numerical weather forecast provides input to icing, and coupling to power prediction is weaker when power is compromised.
In consideration of the fact that the sensor for directly monitoring the icing is extremely small in deployment amount in an actual wind power plant, input variables required in many theoretical models, such as rime, rime and saturated water content of the icing, cannot be applied to large scale in actual production. Therefore, a set of physical model which is high in applicability and combines with other meteorological measured data and considers the actual icing physical process to replace the actual icing record is lacking.
The existing power prediction scheme combined with ice coating cannot solve the pain point of short-term power prediction, particularly the prediction precision of the short-term power prediction is different from that of ultra-short-term prediction in the day, the prediction level of the future atmospheric state is dependent on numerical weather prediction, the ice coating is more corresponding to extreme weather situations or more complex and changeable weather situations in many times, and uncertainty caused by microclimate with smaller scale needs to be considered, so that higher requirements are put forward on the short-term prediction difficulty of the ice coating, particularly the problem of uncertainty of temperature, humidity and wind speed, and one important method for solving the uncertainty is to provide an aggregate prediction result.
Finally, the icing damage solution model applied to short-term power prediction is simpler, and often the physical process prediction of the icing is finer, but the power prediction is not fully coupled. Therefore, the uncertainty of the power prediction and the uncertainty of the icing prediction cannot be comprehensively considered, so that the icing prediction precision and the power prediction precision can be high, but the advantages cannot be comprehensively exerted to provide more accurate power prediction under the icing condition.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a new energy power prediction method and device based on icing and power coupling modeling, which have stronger applicability, greatly reduce the uncertainty of prediction, and couple power and icing prediction online to realize high-precision power prediction under the icing condition.
In order to achieve the above purpose, the technical scheme of the invention is realized as follows:
a new energy power prediction method based on icing and power coupling modeling comprises the following steps:
s1, acquiring meteorological data of measured wind speed, temperature, humidity, air pressure and precipitation in a new energy electric field historical Icing period, and comprehensively obtaining an Icing physical model Icing_status based on an Icing and deicing model;
s2, obtaining theoretical power_thy and actual Power power_true of a new energy electric field in a historical Icing period, stopping and overhauling records, and constructing Icing Power breaking Iing_loss of the whole new energy electric field by using the theoretical Power and the actual Power;
；
s3, constructing an Icing power loss Iing_loss data set based on the step S2;
s4, performing back calculation of historical power prediction based on actual measurement data of the historical icing period, and constructing a historical power prediction data set；
S5, acquiring a historical actual measurement Power data set Power_true;
s6, constructing a historical Icing prediction tag, and calculating Icing prediction damage Iing_k as a prediction tag in a period corresponding to the Icing power damage Iing_loss data set;
；
s7, calling the Icing physical model in the step S1 based on the temperature, the humidity, the air pressure, the rainfall and the air speed of the weather set forecast data to generate an Icing prediction state data set Iing_loss_prediction;
s8, the Icing prediction state data set Iing_loss_prediction enters a statistical Model together with meteorological elements predicted by a meteorological set to carry out regression, and the regression target is Icing prediction damage Iing_k of a historical Icing period to generate a prediction Model model_Iing;
s9, based on the prediction Model model_Iing, combining real-time weather set prediction data to obtain real-time Icing prediction damage Iing_k;
s10, synthesizing a real-time power Prediction result prediction_norm and a real-time Icing Prediction break Iing_k to obtain a prediction_norm (1-Iing_k), and combining the on-site real-time Icing power break Iing_loss to obtain a power Prediction result under the Icing condition.
Further, in step S1, the ice-increasing model is as follows:
；
；
the ice melting model is as follows:
；
in the method, in the process of the invention,indicates the increase or decrease of ice coating, the coefficient->，/>Representing empirical coefficients related to the processes of ice-increasing and ice-melting, < ->Is the density of the liquid water content in the air, +.>For wind speed>The ratio of the molecular weights of water vapor and dry air, taken as a constant, was 0.622,for the density of dry air->Is the saturated water vapor pressure of air, +.>For the actual water vapour pressure +.>In order to provide a critical temperature at which ice coating begins,is the actual temperature; e, e 1 Is the vapor pressure of 2m near the surface; e, e 2 The water vapor pressure at the height of the hub; p is p 1 Is near surface 2m air pressure; p is p 2 The air pressure at the height of the hub is set;
and directly adding the result of the ice-adding model with the result of the ice-melting model to obtain an ice-coating physical model Icing_status.
Further, in step S8, the statistical model performs data enhancement and multiple regression model fusion by using a smote method.
Still further, the data enhancement also uses GAN to generate an antagonism network, or uses a physical model to model historical predictions to produce new samples.
Furthermore, the multiple regression models are fused to complete the regression task by adopting multiple machine learning regression models or adopting a deep neural network.
The invention also provides a new energy power prediction device based on icing and power coupling modeling, which comprises:
and the ice coating physical model module is as follows: acquiring meteorological data of measured wind speed, temperature, humidity, air pressure and precipitation in a new energy electric field historical Icing period, and comprehensively obtaining an Icing physical model Icing_status based on an Icing and deicing model;
the icing power breakage calculation module is used for: acquiring theoretical power_thy and actual Power power_true of a new energy electric field in a historical Icing period, stopping and overhauling records, and constructing Icing Power breaking Iing_loss of the whole new energy electric field by using the theoretical Power and the actual Power;
；
iing_loss dataset Module: constructing an Icing power loss Icing_loss data set based on the Icing power loss calculation module;
historical power prediction dataset module: performing back calculation of historical power prediction based on removing actual measurement data of historical icing time period, and constructing a historical power prediction data set；
Historical measured Power data Power_true data set module: acquiring a historical actual measurement Power data set Power_true;
historical icing prediction label module: constructing a historical Icing prediction tag, and calculating Icing prediction damage Iing_k as a prediction tag in a period corresponding to the Icing power damage Iing_loss data set;
；
ice coating prediction state data set module: calling the Icing physical model based on the temperature, humidity, air pressure, precipitation and wind speed of the weather set forecast data to generate an Icing prediction state data set Iing_loss_prediction;
a prediction model module: the Icing prediction state data set Iing_loss_prediction enters a statistical Model together with meteorological elements predicted by a meteorological set to carry out regression, and the regression target is Icing prediction breaking Iing_k of a historical Icing period to generate a prediction Model model_Iing;
iving_k prediction module: based on the prediction Model model_Iing, combining real-time weather set prediction data to obtain real-time Icing prediction damage Iing_k;
and a prediction result module: and combining a real-time power Prediction result prediction_norm and a real-time Icing Prediction breaking Iing_k to obtain a prediction_norm (1-Iing_k), and combining the on-site real-time Icing power breaking Iing_loss to obtain a power Prediction result under the Icing condition.
Further, in the icing physical model module, the icing model is as follows:
；
；
the ice melting model is as follows:
；
in the method, in the process of the invention,indicates the increase or decrease of ice coating, the coefficient->，/>Representing empirical coefficients related to the processes of ice-increasing and ice-melting, < ->Is the density of the liquid water content in the air, +.>For wind speed>The ratio of the molecular weights of water vapor and dry air, taken as a constant, was 0.622,for the density of dry air->Is the saturated water vapor pressure of air, +.>For the actual water vapour pressure +.>In order to provide a critical temperature at which ice coating begins,is the actual temperature; e, e 1 Is the vapor pressure of 2m near the surface; e, e 2 The water vapor pressure at the height of the hub; p is p 1 Is near surface 2m air pressure; p is p 2 The air pressure at the height of the hub is set;
and directly adding the result of the ice-adding model with the result of the ice-melting model to obtain an ice-coating physical model Icing_status.
Further, the statistical model in the prediction model module uses smote for data enhancement and fusion of multiple regression models.
Still further, the data enhancement in the predictive model module also uses GAN to generate a challenge network or uses a physical model to simulate historical predictions to produce new samples.
Furthermore, the multiple regression models in the prediction model module are fused to complete the regression task by adopting multiple machine learning regression models or adopting a deep neural network.
Compared with the prior art, the invention has the following beneficial effects:
the method combines the short-term icing prediction and the power prediction, particularly removes the actual measurement data of the icing to construct a short-term power prediction model as a basic model, and the regression target of the later prediction selects the difference between the basic model and the actual measurement power, so that the prediction result can not only effectively prompt the icing output loss risk of a station or a whole network, but also greatly improve the power prediction precision.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of the invention.
Detailed Description
It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other.
For the purpose of making the objects and features of the present invention more comprehensible, embodiments accompanied with figures are described in detail below.
Taking a wind farm as an example, as shown in fig. 1, the present invention specifically includes:
1) And obtaining meteorological data such as measured wind speed, temperature, humidity, air pressure, precipitation and the like of a wind power plant fan, and comprehensively obtaining the Icing state Iing_status based on the ice-adding and ice-melting models. Wherein the ice-increasing model is as follows:
；
；
the ice melting model is as follows:
；
in the method, in the process of the invention,indicates the increase or decrease of ice coating, the coefficient->，/>Representing empirical coefficients related to the processes of ice-increasing and ice-melting, < ->Is the density of the liquid water content in the air, +.>For wind speed>The ratio of the molecular weights of water vapor and dry air, taken as a constant, was 0.622,for the density of dry air->Is the saturated water vapor pressure of air, +.>For the actual water vapour pressure +.>In order to provide a critical temperature at which ice coating begins,e is the actual temperature e 1 Is the vapor pressure of 2m near the surface; e, e 2 The water vapor pressure at the height of the hub; p is p 1 Is near surface 2m air pressure; p is p 2 Is the air pressure at the height of the hub.
And the result of the ice-adding model is directly added with the result of the ice-melting model to obtain Iing_status, and the parameters are adjusted to obtain the ice-coating physical model applicable to the region.
In addition, other different solutions may be used for the icing physical model, such as a simplified or more classified model for icing and de-icing.
2) And acquiring theoretical power_thy and actual Power power_true of the historical icing period of the wind farm site, and recording shutdown and overhaul states. The ice coating power break Icing_loss of the whole wind power plant is constructed by using theoretical power and actual power.
；
Wherein, the historical data of Iing_loss is used for modeling, and the real-time Icing power damage data Iing_loss can be calculated according to the calculation of the above formula and used as initial value reference of short-term prediction.
3) And obtaining a data set of the historical Icing power loss Iing_loss based on the physical model and the calculation of the Iing_loss.
4) Performing back calculation of historical power prediction based on the actual measurement data of the ice coating period of the removal history (i.e. re-performing the prediction of the historical power by using the historical actual measurement data after the actual measurement data of the historical ice coating period is removed), and constructing a historical power prediction data set。
5) A historical measured Power dataset power_true is obtained.
6) Constructing a historical Icing prediction label, and calculating Icing prediction breaking Icing_k in an Icing_loss period to serve as the prediction label;
；
7) Calling the Icing physical model constructed in the step 1) to generate a plurality of groups of Icing prediction states Iing_loss_prediction based on the temperature, humidity, air pressure and air speed of the set prediction;
8) And (3) for the Iing_loss_prediction generated by the set prediction and the basic meteorological elements of the set prediction, entering a statistical Model to carry out regression, wherein the regression target is Iing_k at the time of historical actually measured loss data, and a scheme of fusion of data enhancement smote and a plurality of regression models is used for generating a prediction model_Iing.
Wherein the use of data augmentation is a technique that increases the amount of data and model generalization capability by making certain transformations to the raw data, as icing may belong to small sample events at certain wind farms. The aim is to create new and different data samples without changing the actual tag of the data. This can help the model learn different aspects of the data, reducing the likelihood of overfitting, and thus improving the performance of the model on unseen data.
The basic idea of smote mentioned in this embodiment is to generate new samples of a minority class by means of interpolation. Specifically, it will select a minority class sample, then randomly select a sample from its nearest neighbors, and finally randomly select a point on the line between the two samples as the new minority class sample. In this way, SMOTE generates some artificial minority class samples to achieve class balancing and increase model generalization ability.
In addition to the smote method, there are creating new data samples using a generation model such as GAN (generation countermeasure network), producing new samples using a physical model simulation history prediction, and the like.
The fusion of multiple regression models generally adopts multiple machine learning regression models, and if the data volume is large, a deep neural network can also be adopted to complete the regression task, so that model_Iing is generated, and the selection of the models is not the key point of the patent.
9) In actual prediction, based on the trained model_Iing combined with real-time weather set prediction data, the predicted power break Iing_k under the Icing condition is obtained.
10 Combining the real-time power Prediction result of removing the Icing with the real-time Prediction loss Iing_k of the Icing to obtain a prediction_norm (1-Iing_k), and combining the real-time Icing power loss data Iing_loss serving as a short-term Prediction initial value on site to obtain the power Prediction result under the Icing condition.
The root mean square precision of the short-term power prediction in the 5-month icing period in the southwest mountain region is improved by 0.5%, and the root mean square precision of a single day can be improved by more than 50% aiming at icing events under the condition of the cold and the tide in the China region. The prediction result can not only effectively prompt the icing output loss risk of a station or a whole network, but also greatly improve the power prediction precision.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (8)
1. The new energy power prediction method based on icing and power coupling modeling is characterized by comprising the following steps of:
s1, acquiring meteorological data of measured wind speed, temperature, humidity, air pressure and precipitation in a new energy electric field historical Icing period, and comprehensively obtaining an Icing physical model Icing_status based on an Icing and deicing model;
s2, acquiring theoretical power_thy and actual Power power_true of a new energy electric field in a historical Icing period, stopping and overhauling records, and constructing Icing Power breaking Iing_loss of the whole new energy electric field by using the theoretical Power and the actual Power;
；
s3, constructing an Icing power loss Iing_loss data set based on the step S2;
s4, performing back calculation of historical power prediction based on actual measurement data of the historical icing period, and constructing a historical power prediction data set；
S5, acquiring a historical actual measurement Power data set Power_true;
s6, constructing a historical Icing prediction tag, and calculating Icing prediction damage Iing_k as a prediction tag in a period corresponding to the Icing power damage Iing_loss data set;
；
s7, calling the Icing physical model in the step S1 based on the temperature, the humidity, the air pressure, the rainfall and the air speed of the weather set forecast data to generate an Icing prediction state data set Iing_loss_prediction;
s8, the Icing prediction state data set Iing_loss_prediction enters a statistical Model together with meteorological elements predicted by a meteorological set to carry out regression, and the regression target is Icing prediction damage Iing_k of a historical Icing period to generate a prediction Model model_Iing;
s9, based on the prediction Model model_Iing, combining real-time weather set prediction data to obtain real-time Icing prediction damage Iing_k;
s10, synthesizing a real-time power Prediction result prediction_norm and a real-time Icing Prediction break Iing_k to obtain a prediction_norm (1-Iing_k), and combining the on-site real-time Icing power break Iing_loss to obtain a power Prediction result under the Icing condition;
in step S1, the ice-increasing model is as follows:
；
；
the ice melting model is as follows:
；
in the method, in the process of the invention,indicates the increase or decrease of ice coating, the coefficient->，/>Representing the correlation of the processes of ice-increasing and ice-meltingEmpirical coefficient of>Is the density of the liquid water content in the air, +.>For wind speed>Taking the constant ratio of the molecular weight of water vapor and dry air as 0.622 +.>For the density of dry air->Is the saturated water vapor pressure of air, +.>For the actual water vapour pressure +.>For critical temperature at which icing starts, +.>Is the actual temperature; e, e 1 Is the vapor pressure of 2m near the surface; e, e 2 The water vapor pressure at the height of the hub; p is p 1 Is near surface 2m air pressure; p is p 2 The air pressure at the height of the hub is set;
and directly adding the result of the ice-adding model with the result of the ice-melting model to obtain an ice-coating physical model Icing_status.
2. The new energy power prediction method based on ice coating and power coupling modeling according to claim 1, wherein the statistical model in step S8 uses a smote method for data enhancement and multiple regression model fusion.
3. The new energy power prediction method based on ice coating and power coupling modeling of claim 2, wherein the data enhancement also uses GAN generation of an countermeasure network or physical model simulation history prediction to produce new samples.
4. The new energy power prediction method based on ice coating and power coupling modeling according to claim 2, wherein the multiple regression models are fused to complete regression tasks by adopting multiple machine learning regression models or by adopting a deep neural network.
5. The utility model provides a new forms of energy power prediction device based on icing and power coupling modeling which characterized in that includes:
and the ice coating physical model module is as follows: acquiring meteorological data of measured wind speed, temperature, humidity, air pressure and precipitation in a new energy electric field historical Icing period, and comprehensively obtaining an Icing physical model Icing_status based on an Icing and deicing model;
the icing power breakage calculation module is used for: acquiring theoretical power_thy and actual Power power_true of a new energy electric field in a historical Icing period, stopping and overhauling records, and constructing Icing Power breaking Iing_loss of the whole new energy electric field by using the theoretical Power and the actual Power;
；
iing_loss dataset Module: constructing an Icing power loss Icing_loss data set based on the Icing power loss calculation module;
historical power prediction dataset module: performing back calculation of historical power prediction based on removing actual measurement data of historical icing time period, and constructing a historical power prediction data set；
Historical measured Power data Power_true data set module: acquiring a historical actual measurement Power data set Power_true;
historical icing prediction label module: constructing a historical Icing prediction tag, and calculating Icing prediction damage Iing_k as a prediction tag in a period corresponding to the Icing power damage Iing_loss data set;
；
ice coating prediction state data set module: calling the Icing physical model based on the temperature, humidity, air pressure, precipitation and wind speed of the weather set forecast data to generate an Icing prediction state data set Iing_loss_prediction;
a prediction model module: the Icing prediction state data set Iing_loss_prediction enters a statistical Model together with meteorological elements predicted by a meteorological set to carry out regression, and the regression target is Icing prediction breaking Iing_k of a historical Icing period to generate a prediction Model model_Iing;
iving_k prediction module: based on the prediction Model model_Iing, combining real-time weather set prediction data to obtain real-time Icing prediction damage Iing_k;
and a prediction result module: combining a real-time power Prediction result prediction_norm and a real-time Icing Prediction break Iing_k to obtain a prediction_norm (1-Iing_k), and combining the on-site real-time Icing power break Iing_loss to obtain a power Prediction result under the Icing condition;
in the icing physical model module, the icing model is as follows:
；
；
the ice melting model is as follows:
；
in the method, in the process of the invention,indicates the increase or decrease of ice coating, the coefficient->，/>Representing empirical coefficients related to the processes of ice-increasing and ice-melting, < ->Is the density of the liquid water content in the air, +.>For wind speed>Taking the constant ratio of the molecular weight of water vapor and dry air as 0.622 +.>For the density of dry air->Is the saturated water vapor pressure of air, +.>For the actual water vapour pressure +.>For critical temperature at which icing starts, +.>Is the actual temperature; e, e 1 Is the vapor pressure of 2m near the surface; e, e 2 The water vapor pressure at the height of the hub; p is p 1 Is near surface 2m air pressure; p is p 2 The air pressure at the height of the hub is set;
and directly adding the result of the ice-adding model with the result of the ice-melting model to obtain an ice-coating physical model Icing_status.
6. The new energy power prediction device based on ice coating and power coupling modeling according to claim 5, wherein the statistical model in the prediction model module uses smote for data enhancement and multiple regression model fusion.
7. The new energy power prediction device based on ice coating and power coupling modeling of claim 6, wherein the data enhancement in the prediction model module also uses GAN generation of a countermeasure network or physical model simulation history prediction to produce new samples.
8. The new energy power prediction device based on ice coating and power coupling modeling according to claim 6, wherein the multiple regression models in the prediction model module are integrated to complete regression tasks by adopting multiple machine learning regression models or adopting a deep neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410100966.6A CN117638926B (en) | 2024-01-25 | 2024-01-25 | New energy power prediction method and device based on icing and power coupling modeling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410100966.6A CN117638926B (en) | 2024-01-25 | 2024-01-25 | New energy power prediction method and device based on icing and power coupling modeling |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117638926A CN117638926A (en) | 2024-03-01 |
| CN117638926B true CN117638926B (en) | 2024-04-05 |
Family
ID=90027230
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410100966.6A Active CN117638926B (en) | 2024-01-25 | 2024-01-25 | New energy power prediction method and device based on icing and power coupling modeling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117638926B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118153785B (en) * | 2024-05-11 | 2024-08-20 | 山东大学 | Wind-solar climbing event and extreme power prediction method and system under extreme weather |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114439706A (en) * | 2021-12-14 | 2022-05-06 | 电力规划总院有限公司 | Method for predicting icing state of fan blade of wind power plant |
| CN115438554A (en) * | 2022-11-04 | 2022-12-06 | 国网江西省电力有限公司电力科学研究院 | Wind power icing prediction method based on weather forecast |
| CN115828617A (en) * | 2022-12-20 | 2023-03-21 | 国网湖南省电力有限公司 | Fan blade icing and power loss experiment and calculation method thereof |
| WO2023063887A2 (en) * | 2021-10-14 | 2023-04-20 | Envision Digital International Pte. Ltd. | Method and apparatus for predicting state of wind turbine blade, and device and storage medium therefor |
| CN115994325A (en) * | 2023-03-24 | 2023-04-21 | 湖北省气象服务中心（湖北省专业气象服务台） | Fan icing power generation data enhancement method based on TimeGAN deep learning method |
| CN116307257A (en) * | 2023-05-08 | 2023-06-23 | 华北电力科学研究院有限责任公司 | Output power prediction method and device for wind farm under specific weather |
| CN116362382A (en) * | 2023-03-10 | 2023-06-30 | 广西电网有限责任公司 | Short-term power prediction method and system based on icing state of wind power plant |
| CN116662748A (en) * | 2023-05-11 | 2023-08-29 | 国网湖南省电力有限公司 | Prediction method and prediction device for fan icing thickness, storage medium and processor |
| CN116702957A (en) * | 2023-05-18 | 2023-09-05 | 国电南瑞科技股份有限公司 | New energy power prediction method, equipment and storage medium for extreme weather |
| CN116760009A (en) * | 2023-05-10 | 2023-09-15 | 国网湖南省电力有限公司 | Wind power icing shutdown off-grid risk prediction method, wind power icing shutdown off-grid risk prediction device and processor |
| CN117439075A (en) * | 2023-11-07 | 2024-01-23 | 国网重庆市电力公司电力科学研究院 | Wind power prediction method, device and medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120226485A1 (en) * | 2011-03-03 | 2012-09-06 | Inventus Holdings, Llc | Methods for predicting the formation of wind turbine blade ice |
| CN103673960B (en) * | 2012-08-30 | 2016-12-21 | 国际商业机器公司 | For the method and apparatus predicting the ice coating state on transmission line of electricity |
| CN109958588B (en) * | 2017-12-14 | 2020-08-07 | 北京金风科创风电设备有限公司 | Icing prediction method, icing prediction device, storage medium, model generation method and model generation device |
-
2024
- 2024-01-25 CN CN202410100966.6A patent/CN117638926B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023063887A2 (en) * | 2021-10-14 | 2023-04-20 | Envision Digital International Pte. Ltd. | Method and apparatus for predicting state of wind turbine blade, and device and storage medium therefor |
| CN114439706A (en) * | 2021-12-14 | 2022-05-06 | 电力规划总院有限公司 | Method for predicting icing state of fan blade of wind power plant |
| CN115438554A (en) * | 2022-11-04 | 2022-12-06 | 国网江西省电力有限公司电力科学研究院 | Wind power icing prediction method based on weather forecast |
| CN115828617A (en) * | 2022-12-20 | 2023-03-21 | 国网湖南省电力有限公司 | Fan blade icing and power loss experiment and calculation method thereof |
| CN116362382A (en) * | 2023-03-10 | 2023-06-30 | 广西电网有限责任公司 | Short-term power prediction method and system based on icing state of wind power plant |
| CN115994325A (en) * | 2023-03-24 | 2023-04-21 | 湖北省气象服务中心（湖北省专业气象服务台） | Fan icing power generation data enhancement method based on TimeGAN deep learning method |
| CN116307257A (en) * | 2023-05-08 | 2023-06-23 | 华北电力科学研究院有限责任公司 | Output power prediction method and device for wind farm under specific weather |
| CN116760009A (en) * | 2023-05-10 | 2023-09-15 | 国网湖南省电力有限公司 | Wind power icing shutdown off-grid risk prediction method, wind power icing shutdown off-grid risk prediction device and processor |
| CN116662748A (en) * | 2023-05-11 | 2023-08-29 | 国网湖南省电力有限公司 | Prediction method and prediction device for fan icing thickness, storage medium and processor |
| CN116702957A (en) * | 2023-05-18 | 2023-09-05 | 国电南瑞科技股份有限公司 | New energy power prediction method, equipment and storage medium for extreme weather |
| CN117439075A (en) * | 2023-11-07 | 2024-01-23 | 国网重庆市电力公司电力科学研究院 | Wind power prediction method, device and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117638926A (en) | 2024-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117638926B (en) | New energy power prediction method and device based on icing and power coupling modeling | |
| CN107341569B (en) | Photovoltaic power prediction method combining photovoltaic power physical model and data driving | |
| CN111310889B (en) | Evaporation waveguide profile estimation method based on deep neural network | |
| CN107766990B (en) | Method for predicting power generation power of photovoltaic power station | |
| CN109802430A (en) | A kind of wind-powered electricity generation power grid control method based on LSTM-Attention network | |
| CN111260126B (en) | Short-term photovoltaic power generation prediction method considering correlation degree of weather and meteorological factors | |
| CN111695724B (en) | Wind speed prediction method based on hybrid neural network model | |
| KR101476522B1 (en) | System for predicting energy producing quantity | |
| CN115345076B (en) | Wind speed correction processing method and device | |
| CN117290810B (en) | Short-time strong precipitation probability prediction fusion method based on cyclic convolutional neural network | |
| Li et al. | Deep spatio-temporal wind power forecasting | |
| CN107679687A (en) | A kind of photovoltaic output modeling method and Generation System Reliability appraisal procedure | |
| CN114462718A (en) | CNN-GRU wind power prediction method based on time sliding window | |
| CN115425680A (en) | Power prediction model construction and prediction method of multi-energy combined power generation system | |
| CN116307257B (en) | Output power prediction method and device for wind farm under specific weather | |
| CN115903086B (en) | Meteorological hydrologic element integrated numerical forecasting system of regional sea wave coupling mode and operation method thereof | |
| CN116502074A (en) | Model fusion-based photovoltaic power generation power prediction method and system | |
| Liu et al. | Research of photovoltaic power forecasting based on big data and mRMR feature reduction | |
| CN112580899A (en) | Medium-and-long-term wind power generation prediction method and system fused with machine learning model | |
| Cheng-Ping et al. | Research on hydrology time series prediction based on grey theory and [epsilon]-support vector regression | |
| CN116937565A (en) | Distributed photovoltaic power generation power prediction method, system, equipment and medium | |
| CN116073369A (en) | Copula photovoltaic output probability interval prediction method adopting Monte Carlo simulation improvement | |
| CN116167508A (en) | Short-term photovoltaic output rapid prediction method and system based on meteorological factor decomposition | |
| CN115912334A (en) | Method for establishing prediction model of output guarantee rate of wind power plant and prediction method | |
| CN113554203B (en) | Wind power prediction method and device based on high-dimensional meshing and LightGBM |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |