Wind power prediction method, device, equipment and medium
By acquiring meteorological and wind power data under cold wave weather scenarios, performing feature extraction and model prediction, the problems of accuracy and efficiency in wind power prediction under cold wave weather were solved, and rapid prediction at the minute level was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID HUNAN ELECTRIC POWER COMPANY LIMITED
- Filing Date
- 2025-02-11
- Publication Date
- 2026-08-04
AI Technical Summary
Traditional wind power forecasting methods have large prediction errors and low accuracy in cold wave weather scenarios, and cannot effectively adapt to the fluctuations in wind power.
We acquire meteorological data, wind power data, and cold wave data under cold wave weather scenarios, extract features, and use a pre-trained power prediction model to predict wind power by combining meteorological data and target cold wave dynamic characteristics.
It improves the accuracy and efficiency of wind power forecasting, enabling rapid forecasting at the minute level to meet the real-time operation requirements of the power system.
Smart Images
Figure CN119990455B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of wind power prediction technology, and in particular to a wind power prediction method, device, equipment and medium. Background Technology
[0002] The wind power output generated by wind power technology has strong fluctuations and uncertainties, which can seriously affect the stability and security of the power system. Therefore, wind power forecasting is very important.
[0003] In cold wave weather scenarios with strong winds, wind power fluctuates greatly and is influenced by many factors. Traditional wind power forecasting methods mainly rely on wind direction and speed data, which are not well adapted to cold wave weather scenarios, resulting in large prediction errors and relatively low accuracy. Therefore, how to achieve accurate wind power forecasting under cold wave weather scenarios has become an urgent problem to be solved. Summary of the Invention
[0004] To address the aforementioned technical problems, this disclosure provides a wind power prediction method, apparatus, equipment, and medium.
[0005] According to one aspect of this disclosure, a wind power prediction method is provided, comprising:
[0006] Acquire forecast data for wind power function prediction under cold wave weather scenarios, the forecast data including: meteorological data, wind power data and cold wave data;
[0007] Feature extraction is performed on the cold wave data to obtain the target cold wave dynamic characteristics;
[0008] Using a pre-trained power prediction model, the predicted wind power of the wind farm within a preset time period is determined based on the meteorological data, the wind power data, and the target cold wave dynamics.
[0009] According to another aspect of this disclosure, a wind power prediction device is also provided, comprising:
[0010] The data acquisition module is used to acquire forecast data for wind power function prediction under cold wave weather scenarios. The forecast data includes meteorological data, wind power data and cold wave data.
[0011] The feature extraction module is used to extract features from the cold wave data to obtain the target cold wave dynamic features;
[0012] The wind power prediction module is used to determine the predicted wind power of the wind farm within a preset time period based on the meteorological data, the wind power data, and the target cold wave dynamic characteristics using a pre-trained power prediction model.
[0013] According to another aspect of this disclosure, an electronic device is also provided, the electronic device comprising:
[0014] processor;
[0015] Memory used to store the processor's executable instructions;
[0016] The processor is configured to read the executable instructions from the memory and execute the instructions to implement the above method.
[0017] According to another aspect of this disclosure, a computer-readable storage medium is also provided, the storage medium storing a computer program for performing the above-described method.
[0018] The technical solution provided in this disclosure has the following advantages compared with the prior art:
[0019] The technical solution provided in this disclosure includes: acquiring prediction data for wind power function prediction under cold wave weather scenarios, the prediction data including meteorological data, wind power data, and cold wave data; extracting features from the cold wave data to obtain target cold wave dynamic features; and determining the predicted wind power value of the wind farm within a preset time period based on the meteorological data, wind power data, and target cold wave dynamic features using a pre-trained power prediction model. Compared with the prior art, this technical solution newly introduces cold wave data and the target cold wave dynamic features extracted therefrom, using the target cold wave dynamic features as a key factor in wind power prediction, and combining it with meteorological data and wind power data, which can effectively improve the accuracy and precision of wind power prediction. The use of a pre-trained power prediction model during prediction can balance prediction accuracy and efficiency, thereby providing minute-level rapid prediction to meet the real-time operation requirements of the power system. Therefore, this disclosure improves the accuracy and efficiency of wind power prediction. Attached Figure Description
[0020] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.
[0021] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.
[0022] Figure 1 This is a flowchart of the wind power prediction method described in the embodiments of this disclosure;
[0023] Figure 2This is a schematic diagram of the wind power prediction device described in the embodiments of this disclosure;
[0024] Figure 3 This is a schematic diagram of the structure of the electronic device described in an embodiment of this disclosure. Detailed Implementation
[0025] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.
[0026] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.
[0027] Figure 1 This is a flowchart illustrating a wind power prediction method provided in an embodiment of the present disclosure. This method is applicable to wind power prediction during cold wave weather scenarios. The wind power prediction method can be executed by a wind power prediction device, which can be implemented using software and / or hardware, such as electronic devices and servers.
[0028] like Figure 1 As shown, the wind power prediction method provided in this embodiment may include the following steps.
[0029] S102, acquire forecast data for wind power function prediction under cold wave weather scenarios. The forecast data includes: meteorological data, wind power data and cold wave data.
[0030] Meteorological data includes meteorological observation and reanalysis data related to the geographical area where the wind farm is located and cold wave weather scenarios, such as data on temperature, wind speed, air pressure, and wind direction. Specifically, this meteorological data can be obtained through ground observation stations, satellite remote sensing, and numerical weather prediction model data (such as ERA5 and GFS).
[0031] Wind power data includes historical wind speed, wind direction, and wind power output for the geographical area where the wind farm is located.
[0032] The cold wave data includes the spatial structure, propagation path, intensity, and spatiotemporal distribution of at least one historical cold wave event. The impact range of at least one historical cold wave event overlaps with the geographical area where the wind farm to be predicted is located.
[0033] This embodiment may include: preprocessing the forecast data; wherein, the preprocessing includes: data cleaning, and time matching of meteorological data and wind power data.
[0034] Specifically, data cleaning is performed on each forecast data point. This cleaning includes removing outliers and eliminating missing values. Data cleaning improves the accuracy of the forecast data.
[0035] Then, the meteorological data and wind power data in the forecast data are aligned and synchronized in the time dimension to achieve time matching.
[0036] In addition, initial feature extraction can be performed on the cold wave data, such as extracting data on temperature drop and wind speed change rate, and adding them to the cold wave data.
[0037] This embodiment performs subsequent steps based on the predicted data after the above preprocessing.
[0038] S104, feature extraction is performed on the cold wave data to obtain the target cold wave dynamic characteristics.
[0039] In this embodiment, the target time period for the cold wave event is first determined from the cold wave data based on the preset temperature change range and wind speed change range.
[0040] Specifically, the start and end of a cold wave process can be defined based on preset temperature and wind speed variation ranges. Based on this, the target time period for the cold wave event can be determined from the cold wave data, according to the temperature and wind speed variation ranges.
[0041] Secondly, target features are extracted from the cold wave data within the target time period to obtain the initial cold wave dynamics features; among them, the target features include: temporal features, spatial features, and key features to be identified for characterizing cold wave weather.
[0042] In this embodiment, a pre-trained feature extraction model can be used to extract target features from cold wave data within a target time period. The target features to be extracted include:
[0043] Temporal characteristics: arrival time lag of the cold wave, duration, etc.;
[0044] Spatial characteristics: the path of the cold wave, its affected area, and its relative position to wind farms;
[0045] Key indicators to be identified include: temperature gradient, southward velocity of cold air, wind disturbance, and pressure gradient.
[0046] Among them, cold wave data within the target time period can more accurately reflect the characteristics of the actual cold wave process, and has higher reference value when used for wind power prediction; at the same time, extracting target features only from cold wave data within the target time period can reduce the amount of data processing and improve the efficiency of feature extraction.
[0047] Then, target cold wave dynamic features related to changes in electrical power were selected from the initial cold wave dynamic features.
[0048] This embodiment can use feature engineering to screen out the target cold wave dynamic features most closely related to changes in electrical power from the initial cold wave dynamic features. The specific implementation method can be referred to below.
[0049] The correlation between each target feature and wind power in the initial cold wave dynamics was analyzed using the Pearson correlation coefficient; based on the correlation, target cold wave dynamics features related to changes in power power were selected from the initial cold wave dynamics.
[0050] Specifically, the Pearson correlation coefficient between each target feature and wind power in the initial cold wave dynamic characteristics can be calculated. The larger the absolute value of the Pearson correlation coefficient, the stronger the correlation between the two. Based on this, this embodiment samples the absolute value of the Pearson correlation coefficient to represent the correlation.
[0051] Based on the correlation between each target feature and wind power, features with a correlation higher than a preset correlation threshold are selected as the target cold wave dynamic features most closely related to changes in power.
[0052] In this embodiment, different target features in the initial cold wave dynamics have different effects on wind power. Analyzing the correlation between different features and wind power in advance and selecting closely related target cold wave dynamics features as inputs to the power prediction model has the following benefits:
[0053] By filtering target cold wave dynamics features related to changes in electrical power, the dimensionality of the power prediction model input can be reduced, computational complexity lowered, and prediction speed improved while saving computational and storage resources. This increases the amount of information captured by the model, contributing to improved prediction accuracy and reliability. Model interpretability is crucial for building trust and understanding how the model makes predictions; reducing the number of input features, especially retaining the most relevant target cold wave dynamics features, makes the model easier to understand and interpret.
[0054] S106 uses a pre-trained power prediction model to determine the predicted wind power of the wind farm within a preset time period based on meteorological data, wind power data, and the dynamic characteristics of the target cold wave.
[0055] In this embodiment, the power prediction model can be structured using algorithms such as neural networks, support vector machines, and decision tree regression models; no specific limitations are imposed. This power prediction model can estimate wind power by combining wind energy conversion formulas and atmospheric dynamics principles.
[0056] In this embodiment, meteorological data, wind power data, and target cold wave dynamics characteristics are input into the power prediction model. The meteorological data includes temperature, wind speed, wind direction, and air pressure, while the wind power data includes the height and length of the wind farm. The target tidal dynamics characteristics include temperature gradient, air pressure gradient, and wind speed change rate.
[0057] The power prediction model processes meteorological data, wind power data, and target cold wave dynamics characteristics based on the preset wind energy conversion formula and atmospheric dynamics principles, and outputs the predicted wind power of the wind farm in the future preset time period (such as 1 hour, 3 hours, 6 hours, 24 hours).
[0058] In one embodiment, the running speed of the power prediction model can be improved and the prediction timeliness optimized by using high-performance computing (HPC) or edge computing platforms to achieve wind power prediction at the minute level.
[0059] Based on the above embodiments, this disclosure may further include: obtaining the current wind power value, and generating risk warning information when the change between the predicted wind power value and the current wind power value exceeds a preset threshold.
[0060] In this embodiment, if the change between the predicted wind power value and the current wind power value exceeds a preset threshold, it indicates that the cold wave may cause a large fluctuation in wind power. In this case, risk warning information is generated to realize the warning function.
[0061] Regarding the power prediction model involved in the foregoing embodiments, in order for this power prediction model to be used for wind power prediction and to achieve the expected prediction results, model training is required in advance. Based on this, this embodiment provides a training method for a power prediction model, including:
[0062] Obtain the training dataset, which includes: sample meteorological data, sample wind power data, sample cold wave dynamics characteristics, and sample wind power values; iteratively train the power prediction model to be trained based on the training dataset to obtain the power prediction model.
[0063] In practice, the power prediction model to be trained can be iteratively trained based on the training dataset. During each training process, the loss value corresponding to the current training process is obtained, and the model parameters of the power prediction model are updated using the loss value until the model converges, thus obtaining the final power prediction model.
[0064] Taking the Kth (K≥1) training iteration as an example, during the Kth training iteration, sample meteorological data, sample wind power data, and sample cold wave dynamic characteristics are input into the power prediction model to be trained, obtaining intermediate power prediction values. It can be understood that when K equals 1, the power prediction model undergoes its first training, and the intermediate power prediction model at this point is the initial power prediction model to be trained. Thereafter, the power prediction model used each time is the model after the previous training parameter adjustment, i.e., the intermediate power prediction model.
[0065] Next, based on the preset error function (such as root mean square error RMSE, mean absolute percentage error MAPE, etc.), the loss value between the intermediate power prediction value and the sample wind power value is calculated, and it is checked whether the loss value is less than the preset loss value threshold.
[0066] If the loss value is not less than the loss value threshold, the intermediate power prediction model will be used as a new power prediction model to be trained, and the next round of training will continue.
[0067] If the loss value is less than the loss value threshold, the model is considered to have converged, and the intermediate power prediction model corresponding to the current iteration is taken as the completed power prediction model.
[0068] This embodiment iteratively trains the power prediction model based on the training dataset, enabling the trained power prediction model to accurately predict wind power, improve the adaptability of the power prediction model to characteristics under different cold wave weather scenarios, enhance the model's generalization ability, and better cope with complex weather systems.
[0069] Furthermore, in practical applications, the accuracy of the trained power prediction model can be verified based on real meteorological data, real electrical data, real cold wave dynamics characteristics, and real wind power values corresponding to past historical cold wave events. If the verification fails, the power prediction model can be retrained using data augmentation and other methods to improve its prediction accuracy.
[0070] In summary, the wind power prediction method provided in this disclosure includes: acquiring prediction data for wind power function prediction under cold wave weather scenarios, including meteorological data, wind power data, and cold wave data; extracting features from the cold wave data to obtain target cold wave dynamic features; and determining the predicted wind power value of the wind farm within a preset time period based on the meteorological data, wind power data, and target cold wave dynamic features using a pre-trained power prediction model. Compared to existing technologies, this technical solution newly introduces cold wave data and the target cold wave dynamic features extracted therefrom, using the target cold wave dynamic features as a key factor in wind power prediction, and combining it with meteorological and wind power data, which can effectively improve the accuracy and precision of wind power prediction. The use of a pre-trained power prediction model during prediction can balance prediction accuracy and efficiency, thereby providing minute-level rapid prediction to meet the real-time operation requirements of the power system. Therefore, this disclosure improves the accuracy and efficiency of wind power prediction.
[0071] Figure 2 This is a schematic diagram of a wind power prediction device provided in an embodiment of this disclosure. Figure 2 As shown, the wind power prediction device may include the following modules.
[0072] Data acquisition module 210 is used to acquire forecast data for wind power function prediction under cold wave weather scenarios. The forecast data includes: meteorological data, wind power data and cold wave data.
[0073] Feature extraction module 220 is used to extract features from the cold wave data to obtain the target cold wave dynamic features;
[0074] The wind power prediction module 230 is used to determine the predicted wind power of the wind farm within a preset time period based on the meteorological data, the wind power data, and the target cold wave dynamic characteristics using a pre-trained power prediction model.
[0075] In one embodiment, the feature extraction module 220 is further configured to:
[0076] Based on the preset temperature and wind speed variation ranges, the target time period for the cold wave event is determined from the cold wave data.
[0077] Target features are extracted from the cold wave data within the target time period to obtain initial cold wave dynamic features; wherein, the target features include: temporal features, spatial features, and key undefined features for characterizing cold wave weather;
[0078] Target cold wave dynamic features related to changes in electrical power are selected from the initial cold wave dynamic features.
[0079] The device provided in this embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.
[0080] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Figure 3 As shown, the electronic device 300 includes one or more processors 301 and memory 302.
[0081] The processor 301 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 300 to perform desired functions.
[0082] The memory 302 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 301 may execute the program instructions to implement the wind power prediction method of the embodiments of this disclosure described above and / or other desired functions. Various contents such as input signals, signal components, and noise components may also be stored in the computer-readable storage medium.
[0083] In one example, the electronic device 300 may also include an input device 303 and an output device 304, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).
[0084] In addition, the input device 303 may also include, for example, a keyboard, a mouse, etc.
[0085] The output device 303 can output various information to the outside, including determined distance information, direction information, etc. The output device 304 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.
[0086] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device 300 relevant to this disclosure are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device 300 may include any other suitable components depending on the specific application.
[0087] Furthermore, this embodiment also provides a computer-readable storage medium storing a computer program for executing the above-described wind power prediction method.
[0088] The computer program product of the wind power prediction method, apparatus, electronic device and medium provided in this disclosure includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.
[0089] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0090] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for predicting wind power output, characterized in that, include: Acquire forecast data for wind power function prediction under cold wave weather scenarios, the forecast data including: meteorological data, wind power data and cold wave data; Feature extraction is performed on the cold wave data to obtain the target cold wave dynamic characteristics; Using a pre-trained power prediction model, the predicted wind power of the wind farm within a preset time period is determined based on the meteorological data, the wind power data, and the target cold wave dynamics. The step of extracting features from the cold wave data to obtain the target cold wave dynamic features includes: Based on the preset temperature and wind speed variation ranges, the target time period for the cold wave event is determined from the cold wave data. The target features are extracted from the cold wave data within the target time period to obtain the initial cold wave dynamic features; wherein, the target features include: temporal features, spatial features and key features for characterizing cold wave weather, the key features include: temperature gradient, southward velocity of cold air, wind field disturbance and pressure gradient; Target cold wave dynamic features related to changes in electrical power are selected from the initial cold wave dynamic features.
2. The method according to claim 1, characterized in that, The process of selecting target cold wave dynamic features related to changes in electrical power from the initial cold wave dynamic features includes: The correlation between the target characteristics and wind power in the initial cold wave dynamics was analyzed using the Pearson correlation coefficient. Based on the correlation, target cold wave dynamic features related to changes in electrical power are selected from the initial cold wave dynamic features.
3. The method according to claim 1, characterized in that, The method further includes: The system acquires the current wind power value and generates a risk warning if the difference between the predicted wind power value and the current wind power value exceeds a preset threshold.
4. The method according to claim 1, characterized in that, The method further includes: Obtain a training dataset, wherein the training data in the training dataset includes: sample meteorological data, sample wind power data, sample cold wave dynamics characteristics, and sample wind power values; The power prediction model is obtained by iteratively training the power prediction model to be trained based on the training dataset.
5. The method according to claim 1, characterized in that, The method includes: The predicted data is preprocessed; wherein the preprocessing includes: data cleaning and time matching of the meteorological data and the wind power data.
6. A wind power prediction device, characterized in that, include: The data acquisition module is used to acquire forecast data for wind power function prediction under cold wave weather scenarios. The forecast data includes meteorological data, wind power data and cold wave data. The feature extraction module is used to extract features from the cold wave data to obtain the target cold wave dynamic features; The wind power prediction module is used to determine the predicted wind power of the wind farm within a preset time period based on the meteorological data, the wind power data, and the target cold wave dynamic characteristics using a pre-trained power prediction model. The feature extraction module is also used for: Based on the preset temperature and wind speed variation ranges, the target time period for the cold wave event is determined from the cold wave data. The target features are extracted from the cold wave data within the target time period to obtain the initial cold wave dynamic features; wherein, the target features include: temporal features, spatial features and key features for characterizing cold wave weather, the key features include: temperature gradient, southward velocity of cold air, wind field disturbance and pressure gradient; Target cold wave dynamic features related to changes in electrical power are selected from the initial cold wave dynamic features.
7. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1-5.
8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a terminal device, cause the terminal device to perform the method as described in any one of claims 1-5.