Power output prediction method and device based on wind speed data, terminal and medium
By constructing multiple ELM models and calculating confidence intervals, the shortcomings of traditional wind speed prediction models in terms of accuracy, reliability and real-time performance are solved, and more efficient wind speed prediction and power output prediction are achieved.
Patent Information
- Application Number
- CN202411589129.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-08
- Publication Date
- 2025-02-28
AI Technical Summary
Traditional wind speed prediction models have shortcomings in capturing wind speed changes, providing credibility assessment and real-time performance, resulting in low prediction accuracy and low confidence.
Using a prediction method based on wind speed data, a random sampling of N reentries was performed by obtaining historical wind speed and power data, multiple ELM models were constructed, and prediction was made using predicted wind speed data, and confidence intervals were calculated to evaluate the confidence of the prediction results.
Improves the accuracy and accuracy of wind speed prediction, provides higher credibility and real-time prediction results, helping operators better respond to risks and make decisions.
Smart Images
Figure CN119539161A_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to the field of wind power technology, and in particular to a method, device, terminal and medium for predicting power output based on wind speed data. [Background technology]
[0002] China's arid northwest region is mostly covered by deserts and Gobi, with rich land resources and sufficient wind energy resources, providing vast land resources for large-scale wind farm construction, and is therefore an ideal choice for building large-scale wind power projects. At the same time, the region has abundant wind energy resources and high wind speeds, allowing wind power systems to fully utilize wind energy to convert into electrical energy.
[0003] Compared with other power generation methods, wind power output is directly affected by wind speed and less affected by other environmental factors. However, wind power output is more volatile and is greatly affected by the randomness of wind speed. Therefore, in view of the development of wind power industry in Xinjiang and other northwestern regions, it is particularly important to accurately predict wind speed in order to better plan and dispatch wind power generation.
[0004] However, the traditional wind speed prediction model has the following problems:
[0005] (1) It is often difficult to accurately capture the changing trend of wind speed, especially under complex meteorological conditions, and its prediction accuracy is easily affected;
[0006] (2) It is often difficult to provide sufficient credibility assessment. Even if the prediction results are given, there is a lack of confidence assessment of the prediction results, resulting in low credibility of the prediction results.
[0007] (3) They often lack real-time performance and are unable to promptly reflect changes in meteorological conditions, resulting in large deviations between the forecast results and the actual situation.
[0008] In view of this, it is necessary to provide a method, device, terminal and medium for predicting power output based on wind speed data to overcome the above-mentioned defects. [Summary of the invention]
[0009] The purpose of the present invention is to provide a method, device, terminal and medium for predicting power output based on wind speed data, aiming to solve the problems of low prediction accuracy and low credibility of prediction results of existing wind speed prediction models, and effectively improve the accuracy and precision of model prediction.
[0010] In order to achieve the above object, the first aspect of the present invention provides a method for predicting power output based on wind speed data, comprising:
[0011] Step S10: Obtain historical wind speeds and corresponding power data of all wind farms that need to be predicted in historical data, and construct a data set;
[0012] Step S20: performing N random samplings with replacement on the data set, and saving the N sampling results to obtain N sets of training set data; wherein N is a positive integer greater than or equal to 2;
[0013] Step S30: construct N ELM models, and use N groups of training set data to train them one by one, and save the N trained ELM models;
[0014] Step S40: obtaining predicted wind speed data of wind farms involved in the wind zone that needs to be predicted;
[0015] Step S50: inputting the predicted wind speed data into the N trained ELM models respectively to obtain N groups of predicted power outputs;
[0016] Step S60: according to the N groups of predicted power outputs, respectively calculating corresponding confidence intervals for different outputs at each preset moment;
[0017] Step S70: output the confidence intervals at each preset time and the predicted power mean value calculated based on the N groups of predicted power outputs.
[0018] In a preferred embodiment, the step S10 includes the following sub-steps:
[0019] Process the null value data in the data set: change all the record values of the historical wind speed that are less than or equal to 0 to 0.
[0020] In a preferred embodiment, the step S20 includes:
[0021] Define the number of sampling rounds N and the number of samples in each round M; where N and M are both positive integers greater than or equal to 2;
[0022] In each round, M data are extracted from the data set to generate a new data set, and the new data set is added to the data set set;
[0023] Determine whether the traversal number of the sampling rounds is less than N. If it is less than N, continue to extract data from the data set; if it is equal to N, output a training set consisting of a data set set; wherein the data set set includes N groups of data sets, and each group of data sets includes M data.
[0024] In a preferred embodiment, the step S40 further includes:
[0025] The predicted wind speed data is corrected based on a preset ENKF method.
[0026] In a preferred embodiment, the step S40 further includes:
[0027] Processing the null value data of the predicted wind speed data: changing all recorded values less than or equal to 0 in the predicted wind speed to 0.
[0028] In a preferred embodiment, the step S60 includes:
[0029] Calculate the predicted power mean value based on the N groups of predicted power outputs;
[0030] Based on the target confidence level p at each moment, take the previous value of the predicted power mean After The percentiles of are used as the lower and upper limits of the confidence interval.
[0031] In a preferred embodiment, the step S60 further includes:
[0032] Compute confidence intervals for different gradients at each time step.
[0033] The present invention also provides a device for predicting power output based on wind speed data, comprising:
[0034] The data acquisition module is used to obtain the historical wind speed and corresponding power data of all wind farms that need to be predicted in the historical data and construct a data set;
[0035] A random sampling module, used to perform N random samplings with replacement on the data set, and save the N sampling results to obtain N sets of training set data; wherein N is a positive integer greater than or equal to 2;
[0036] The model building module is used to build N ELM models, and use N sets of training set data to train them one by one, and save the N trained ELM models;
[0037] A wind speed acquisition module is used to obtain the predicted wind speed data of the wind farm involved in the wind zone that needs to be predicted;
[0038] A power prediction module, used for inputting the predicted wind speed data into the N trained ELM models respectively to obtain N groups of predicted power outputs;
[0039] An interval calculation module, used to calculate corresponding confidence intervals for different outputs at each preset moment according to the N groups of predicted power outputs;
[0040] The prediction output module is used to output the confidence interval at each preset time and the predicted power mean calculated based on N groups of predicted power outputs.
[0041] The present invention also provides a terminal, which includes a memory, a processor, and a computer program stored in the memory. When the computer program is executed by the processor, the computer program implements the various steps of the method for predicting power output based on wind speed data as described in any one of the above implementation modes.
[0042] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the various steps of the method for predicting power output based on wind speed data as described in any of the above implementation modes.
[0043] The power output prediction method, device, terminal and medium based on wind speed data provided by the present invention reduce the weight of outliers through a random data sampling method with replacement, and train multiple ELM (Extreme Learning Machine) models respectively through N groups of different sampling results, so that N groups of predicted power outputs can be obtained according to the predicted wind speed data, and then the calculation of confidence intervals can be introduced according to the N groups of predicted values, so that the credibility of the prediction results can be evaluated, and multiple possible power value ranges can be provided. In this way, operators can better understand the uncertainty of the prediction results, so as to better deal with risks and make decisions. In addition, the present invention adopts an extreme learning machine (ELM) as the core algorithm of the prediction model. Compared with traditional neural network models, ELM has the advantages of fast training, efficient execution and good generalization ability, and can complete wind speed prediction tasks more quickly and improve prediction accuracy.
Brief Description of the Drawings
[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0045] Figure 1 A flow chart of a method for predicting power output based on wind speed data provided by the present invention;
[0046] Figure 2 An exemplary visualization chart of the actual value, confidence interval and actual power of the predicted power in a certain area by this method;
[0047] Figure 3 Another exemplary visualization chart of the actual value, confidence interval and actual power of the predicted power in a certain area by the method;
[0048] Figure 4A framework diagram of a power output prediction system based on wind speed data provided by the present invention. [Specific implementation method]
[0049] In order to make the purpose, technical solution and beneficial technical effect of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. It should be understood that the specific implementation methods described in this specification are only for explaining the present invention, not for limiting the present invention.
[0050] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.
[0051] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
[0052] Embodiment 1
[0053] In an embodiment of the present invention, a method for predicting power output based on wind speed data is provided, which is used to predict wind speed data of wind farms involved in the wind zone to be predicted (which can be obtained through a weather prediction model combined with current weather parameter predictions, or directly obtained through the weather forecast of the Meteorological Bureau), so as to predict the power generation of the wind farm for a period of time in the future, and then guide operating personnel to better deal with risks and make decisions.
[0054] like Figure 1 As shown, the method for predicting power output based on wind speed data includes steps S10-S70.
[0055] Step S10: Obtain historical wind speeds and corresponding power data of all wind farms that need to be predicted in historical data, and construct a data set.
[0056] This step also includes: processing the null value data in the data set, and changing all the record values of the historical wind speed that are less than or equal to 0 to 0.
[0057] Step S20: Perform N random samplings with replacement on the data set, and save the N sampling results to obtain N groups of training set data; wherein N is a positive integer greater than or equal to 2.
[0058] That is, Bootstrap sampling is performed on the original data set, and a sampling distribution can be formed by using random sampling with replacement of the original data set, which is convenient for subsequent estimation of population parameters (such as population mean, median, variance, quantile, etc.). Specifically, step S20 includes steps S21-S23.
[0059] Step S21: define the number of sampling rounds N and the number of samples in each round M; wherein N and M are both positive integers greater than or equal to 2.
[0060] Step S22: extract M data from the data set in each round, generate a new data set, and add the new data set to the data set set.
[0061] Step S23: Determine whether the traversal number of the sampling rounds is less than N. If it is less than N, continue to extract data from the data set; if it is equal to N, output a training set consisting of a data set set; wherein the data set set includes N groups of data sets, and each group of data sets includes M data.
[0062] Step S30: construct N ELM models, and use N groups of training set data to perform training respectively one by one, and save the N trained ELM models.
[0063] Therefore, the present invention adopts the extreme learning machine (ELM) as the core algorithm of the prediction model. Compared with the traditional neural network model, ELM has the advantages of fast training, efficient execution and good generalization ability, and can complete the wind speed prediction task more quickly and improve the prediction accuracy. Due to the difference in data distribution of each group of training set data, the internal parameters of the N ELM models trained separately are also different, so that the predicted power obtained by inputting the same wind speed data is also different.
[0064] Step S40: Obtain predicted wind speed data of wind farms involved in the wind zone that needs to be predicted.
[0065] Among them, this step also includes: processing the null value data of the predicted wind speed data: changing all record values less than or equal to 0 in the predicted wind speed to 0.
[0066] Furthermore, the method also includes the step of correcting the predicted wind speed data based on a preset ENKF (Ensemble Kalman Filter) method. That is, by combining the actual wind power forecasting business scenario with the analysis of the data situation, the wind speed range can be effectively predicted in actual cases by filtering and correcting the forecast wind speed by ENKF, which can effectively consider the observation error and model error, and improve the adaptability and robustness of the wind speed prediction model. Therefore, the uncertainty of the actual wind speed data is processed by Kalman filtering, which effectively improves the precision and accuracy of the model prediction. Kalman filtering can dynamically estimate the system state and adjust it according to the latest observation data to better reflect the actual changes in wind speed.
[0067] Step S50: inputting the predicted wind speed data into N trained ELM models respectively to obtain N groups of predicted power outputs.
[0068] Therefore, by training N ELM models, the weights of outliers in the training set can be effectively eliminated, and the mean obtained by integrating N prediction results is more realistic.
[0069] Step S60: According to the N groups of predicted power outputs, corresponding confidence intervals are calculated for different outputs at each preset moment.
[0070] Therefore, by introducing the calculation of the confidence interval of the predicted value, the present invention can evaluate the credibility of the predicted results and provide multiple possible power value ranges. In this way, operators can better understand the uncertainty of the predicted results, so as to better deal with risks and make decisions. Accurate wind power forecasting can help the power grid system to plan and dispatch energy more effectively, avoiding energy waste and unnecessary costs. In addition, the improvement of prediction accuracy can also reduce the demand for backup energy and improve the overall operating efficiency and economy of the system.
[0071] Specifically, step S60 includes: calculating the predicted power mean according to the N groups of predicted power outputs; taking the previous value of the predicted power mean based on the target confidence level p at each moment; After The percentiles of are used as the lower and upper limits of the confidence interval.
[0072] For example, when calculating a 95% confidence interval, the first 2.5% and last 2.5% percentiles of the predicted power mean are taken as the lower and upper limits of the 95% confidence interval.
[0073] Furthermore, based on the above steps, confidence intervals of different gradients may be calculated for each moment, such as 95% interval, 90% interval, 85% interval, 80% interval, etc.
[0074] Therefore, due to the complex changes in actual wind power forecasts, in order to eliminate the difficult-to-eliminate calculation errors caused by single-point predictions as much as possible and achieve a more accurate short-term prediction of future wind power, this method uses model data of different wind speed data as wind power data obtained by predictions of multiple different algorithms. Each model predicts the data of the predicted wind speed model separately, and finally calculates the confidence intervals of the results of these multiple models to obtain the wind power value range.
[0075] Step S70: output the confidence intervals at each preset time and the predicted power mean value calculated based on the N groups of predicted power outputs.
[0076] The following will combine a specific case to test the prediction results of this method. Figure 2 This is the result of power prediction based on the wind speed forecast data of certain days in the Urumqi area. The blue line is the power prediction data at different times, the orange-yellow blocks are the confidence intervals, and the black line is the actual measured power. Figure 3 This is a visualization chart of model power for some days in another exemplary region. The red line is the actual power measured, the black line is the model predicted power, and the gray block is the confidence interval. Figure 2 and Figure 3 In general, this method can make a relatively accurate prediction of wind power.
[0077] The present invention also provides a device 100 for predicting power output based on wind speed data, which is used to predict wind speed data of a wind farm involved in a wind zone to be predicted, thereby predicting the power generation of the wind farm in the future. It should be noted that the implementation principle and specific implementation method of the device 100 for predicting power output based on wind speed data can refer to the above-mentioned method for predicting power output based on wind speed data, and will not be repeated below.
[0078] like Figure 4 As shown, the power output prediction device 100 based on wind speed data includes:
[0079] The data acquisition module 10 is used to acquire the historical wind speed and corresponding power data of all wind farms that need to be predicted in the historical data, and construct a data set;
[0080] The random sampling module 20 is used to perform N random samplings with replacement on the data set, and save the N sampling results to obtain N sets of training set data; wherein N is a positive integer greater than or equal to 2;
[0081] The model building module 30 is used to build N ELM models, and use N sets of training set data to train them one by one, and save the N trained ELM models;
[0082] A wind speed acquisition module 40 is used to acquire predicted wind speed data of wind farms involved in the wind zone that needs to be predicted;
[0083] A power prediction module 50, used to input the predicted wind speed data into the N trained ELM models respectively to obtain N groups of predicted power outputs;
[0084] An interval calculation module 60, for calculating corresponding confidence intervals for different outputs at each preset moment according to the N groups of predicted power outputs;
[0085] The prediction output module 70 is used to output the confidence interval at each preset time and the predicted power mean value calculated based on the N groups of predicted power outputs.
[0086] In summary, the method, device, terminal and medium for predicting power output based on wind speed data provided by the present invention reduce the weight of outliers through a random data sampling method with replacement, and train multiple ELM models respectively through N groups of different sampling results, so that N groups of predicted power outputs can be obtained according to the predicted wind speed data, and then the calculation of confidence intervals can be introduced according to these N groups of predicted values, so that the credibility of the prediction results can be evaluated, and multiple possible power value ranges can be provided. In this way, operators can better understand the uncertainty of the prediction results, so as to better deal with risks and make decisions. In addition, the present invention adopts an extreme learning machine (ELM) as the core algorithm of the prediction model. Compared with traditional neural network models, ELM has the advantages of fast training, efficient execution and good generalization ability, and can complete wind speed prediction tasks more quickly and improve prediction accuracy.
[0087] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.
[0088] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0089] Those of ordinary skill in the art will appreciate that the units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0090] In the embodiments provided by the present invention, it should be understood that the disclosed systems or devices / terminal equipment and methods can be implemented in other ways. For example, the system or device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the system or unit can be electrical, mechanical or other forms.
[0091] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0092] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0093] The present invention is not limited to what is described in the specification and implementation modes, and therefore additional advantages and modifications can be easily realized by those skilled in the art. Therefore, without departing from the spirit and scope of the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details, representative devices, and illustrative examples shown and described herein.
Claims
1. A method for predicting power output based on wind speed data, characterized in that: include: Step S10: Obtain historical wind speeds and corresponding power data of all wind farms that need to be predicted in historical data, and construct a data set; Step S20: performing N random samplings with replacement on the data set, and saving the N sampling results to obtain N sets of training set data; wherein N is a positive integer greater than or equal to 2; Step S30: construct N ELM models, and use N groups of training set data to train them one by one, and save the N trained ELM models; Step S40: obtaining predicted wind speed data of wind farms involved in the wind zone that needs to be predicted; Step S50: inputting the predicted wind speed data into the N trained ELM models respectively to obtain N groups of predicted power outputs; Step S60: Calculating corresponding confidence intervals for different outputs at each preset moment according to the N groups of predicted power outputs; Step S70: output the confidence intervals at each preset time and the predicted power mean value calculated based on the N groups of predicted power outputs.
2. The method for predicting power output based on wind speed data according to claim 1, characterized in that: The step S10 includes the following sub-steps: Process the null value data in the data set: change all the record values of the historical wind speed that are less than or equal to 0 to 0.
3. The method for predicting power output based on wind speed data according to claim 1, characterized in that: The step S20 comprises: Define the number of sampling rounds N and the number of samples in each round M; where N and M are both positive integers greater than or equal to 2; In each round, M data are extracted from the data set to generate a new data set, and the new data set is added to the data set set; Determine whether the traversal number of the sampling rounds is less than N. If it is less than N, continue to extract data from the data set; if it is equal to N, output a training set consisting of a data set set; wherein the data set set includes N groups of data sets, and each group of data sets includes M data.
4. The method for predicting power output based on wind speed data according to claim 1, characterized in that: The step S40 further includes: The predicted wind speed data is corrected based on a preset ENKF method.
5. The method for predicting power output based on wind speed data according to claim 1, characterized in that: The step S40 further includes: Processing the null value data of the predicted wind speed data: changing all recorded values less than or equal to 0 in the predicted wind speed to 0.
6. The method for predicting power output based on wind speed data according to claim 1, characterized in that: The step S60 comprises: Calculate the predicted power mean value based on the N groups of predicted power outputs; Based on the target confidence level p at each moment, take the previous value of the predicted power mean After The percentiles of are used as the lower and upper limits of the confidence interval.
7. The method for predicting power output based on wind speed data according to claim 6, characterized in that: The step S60 also includes: Compute confidence intervals for different gradients at each time step.
8. A device for predicting power output based on wind speed data, characterized in that: include: The data acquisition module is used to obtain the historical wind speed and corresponding power data of all wind farms that need to be predicted in the historical data and construct a data set; A random sampling module, used to perform N random samplings with replacement on the data set, and save the N sampling results to obtain N sets of training set data; wherein N is a positive integer greater than or equal to 2; The model building module is used to build N ELM models, and use N sets of training set data to train them one by one, and save the N trained ELM models; A wind speed acquisition module is used to obtain the predicted wind speed data of the wind farm involved in the wind zone that needs to be predicted; A power prediction module, used for inputting the predicted wind speed data into the N trained ELM models respectively to obtain N groups of predicted power outputs; An interval calculation module, used to calculate corresponding confidence intervals for different outputs at each preset moment according to the N groups of predicted power outputs; The prediction output module is used to output the confidence interval at each preset time and the predicted power mean calculated based on N groups of predicted power outputs.
9. A terminal, characterized in that: The terminal includes a memory, a processor, and a computer program stored in the memory. When the computer program is executed by the processor, each step of the method for predicting power output based on wind speed data as described in any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, each step of the method for predicting power output based on wind speed data according to any one of claims 1 to 7 is implemented.