Power grid optimal scheduling method and device based on dynamic selection of wind power prediction model

Abstract

The invention discloses a power grid optimization scheduling method and device based on wind power prediction model dynamic selection, and the method comprises the steps: obtaining a prediction error probability density function of each wind power prediction model based on a plurality of wind power prediction models; acquiring real-time meteorological data of the wind turbine generator, performing power prediction based on a plurality of wind power prediction models to obtain a plurality of wind power real-time prediction values, and generating a wind power real-time prediction scene set in combination with a corresponding prediction error probability density function; and on the basis of a plurality of wind power real-time prediction scene sets, real-time control strategies corresponding to the scene sets are generated through a random unit combination model, and the real-time control strategy with the lowest sum of wind curtailment and load loss of wind power in a plurality of real-time control strategies is selected as a control strategy after optimization of the wind turbine generator. And carrying out random unit combination on the basis of scene sets generated by the multi-prediction model, generating a plurality of real-time control strategies in combination with the influence of meteorological factors, and finally selecting a strategy with the lowest value of the sum of wind curtailment and loss load of wind power as an optimization control strategy.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid control, and in particular to a power grid optimization dispatching method and device based on dynamic selection of a wind power prediction model. Background Art

[0002] For power systems containing wind power resources, traditional optimization dispatch models usually select a prediction model with the smallest error from multiple wind power prediction models, and use this prediction model in the random optimization dispatch model for a long time, resulting in the dispatch side only being able to "passively" receive one prediction result (prediction value and prediction error probability distribution). It should be noted that no prediction model can obtain better prediction results in all cases, and this model ignores other prediction results, resulting in the dispatch side losing potential decision results that may be more suitable for actual conditions. Summary of the invention

[0003] The purpose of the embodiment of the present invention is to provide a power grid optimization scheduling method and device based on the dynamic selection of a wind power prediction model. By fully considering the error characteristics of the wind power prediction model and the uncertainty of wind power affected by meteorological factors, a variety of wind power real-time prediction scenario sets are generated, and multiple real-time control strategies are formulated with the help of a random unit combination model. Finally, the strategy with the lowest sum of wind power abandonment and load loss is selected as the optimized control strategy, which effectively improves the utilization efficiency of wind power resources, enhances the ability of the power grid to cope with wind power fluctuations, optimizes power grid scheduling, and helps to achieve safe, stable and efficient operation of the power system.

[0004] To solve the above technical problems, a first aspect of an embodiment of the present invention provides a power grid optimization scheduling method based on dynamic selection of a wind power prediction model, comprising the following steps:

[0005] Based on multiple wind power prediction models, obtaining a prediction error probability density function of each wind power prediction model, and obtaining a plurality of wind power prediction error scenario sets;

[0006] Acquire real-time meteorological data of the wind turbines, perform power forecasts based on the multiple wind power forecast models, obtain a number of wind power real-time forecast values, and generate a number of wind power real-time forecast scenario sets in combination with the corresponding wind power forecast error scenario sets;

[0007] Based on the several wind power real-time prediction scenario sets, several real-time control strategies corresponding to the wind power real-time prediction scenario sets are generated one by one through a random unit combination model, and the real-time control strategy with the lowest value of the sum of wind power abandonment and load loss among the several real-time control strategies is selected as the optimized control strategy for the wind turbine set.

[0008] Furthermore, the types of real-time meteorological data of the wind turbine generator set include: wind speed, wind direction, temperature and / or air pressure.

[0009] Furthermore, the obtaining of a prediction error probability density function of each wind power prediction model based on a plurality of wind power prediction models includes:

[0010] Obtain historical forecast data and historical actual data of wind turbines;

[0011] Based on the historical prediction data, respectively establishing nonlinear mapping relationships between meteorological data and a plurality of wind power prediction models;

[0012] Based on the historical actual data, calculating the prediction error of each wind power prediction model;

[0013] Based on the prediction error, a prediction error probability density function of each wind power prediction model is obtained respectively by kernel density estimation.

[0014] Furthermore, the wind power prediction model includes: a neural network model, a support vector machine model and a random forest model;

[0015] The corresponding nonlinear mapping relationship of the neural network model is:

[0016]

[0017] The corresponding nonlinear mapping relationship of the support vector machine model is:

[0018]

[0019] The nonlinear mapping relationship corresponding to the random forest model is:

[0020]

[0021] Among them, f NN (·),f SVM (·) and f RF (·) are the mapping relationships between wind power prediction values ​​and meteorological data X based on neural network model, support vector machine model and random forest model, respectively. l , W l and are the activation function, weight vector and bias vector of the lth layer of the neural network, respectively, and l = 1, 2, 3, ..., L, L is the number of layers of the neural network, W and b NN are the weight matrix and bias matrix of the neural network, N SVM is the number of support vectors in the support vector machine model, α iis the Lagrange multiplier of the i-th support vector of the support vector machine model, y i is the true label of the i-th support vector of the support vector machine model, K SVM (X i ,X) is the kernel function of the support vector machine model, γ is a positive parameter, X i is the X corresponding to the i-th sample point, b SVM is the bias term of the support vector machine model, N RF is the number of decision trees in the random forest model, is the predicted value of the i-th decision tree for input X, p wind,NN 、p wind,SVM and p wind,RF These are the wind power prediction values ​​obtained based on the neural network model, support vector machine model and random forest model respectively.

[0022] Furthermore, the prediction error probability density function corresponding to the neural network model is:

[0023]

[0024] Δp i wind,NN =p i wind,0 -f NN (X i );

[0025] The corresponding prediction error probability density function of the support vector machine model is:

[0026]

[0027] Δp i wind,SVM =p i wind,0 -f SVM (X i );

[0028] The corresponding prediction error probability density function of the random forest model is:

[0029]

[0030] Δp i wind,RF =p i wind,0 -f RF (X i );

[0031] Among them, K(·) is the kernel density estimation function, h is the bandwidth, N S is the number of samples, Δp windis the wind power prediction error random variable, X i and p i wind,0 are the meteorological factor input vector and historical actual wind power of the i-th historical data sample point, Δp i wind,NN , Δp i wind,SVM and Δp i wind,RF are the wind power prediction errors of the i-th historical data sample point of the neural network model, support vector machine model and random forest model, respectively. and They are the prediction error probability density functions of the neural network model, support vector machine model and random forest model respectively.

[0032] Furthermore, the generating of wind power real-time prediction scenario sets respectively by combining the corresponding prediction error probability density functions includes:

[0033] Based on the prediction error probability density function, the corresponding wind power prediction error scenario set is generated by Latin hypercube sampling and Cholesky decomposition:

[0034]

[0035] Among them, Ω NN,err ,Ω SVM,err and Ω RF,err are the wind power prediction error scenario sets of the neural network model, support vector machine model and random forest model, (Δp wind,NN ) i 、(Δp wind,SVM ) i and (Δp wind,RF ) i The i-th wind power prediction error scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., N sg , N sg is the number of scenes.

[0036] Based on the wind power real-time prediction value and the wind power prediction error scenario set, a wind power real-time prediction scenario set is obtained:

[0037]

[0038] Among them, Ω NN ,Ω SVM and Ω RF The wind power real-time prediction scenario sets of neural network model, support vector machine model and random forest model, p wind,NN,pre , and p wind,RF,preare the real-time wind power prediction values ​​of the neural network model, support vector machine model and random forest model, respectively. (p wind,NN ) i 、(p wind,SVM ) i and (p wind,RF ) i The i-th wind power prediction scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., N sg , N sg is the number of scenes.

[0039] Furthermore, the wind power prediction model includes: a neural network model, a support vector machine model and a random forest model;

[0040] The random unit combination model is:

[0041]

[0042] Among them, f(p wind,output ,x) is the objective function of the random unit combination model, z(p wind,output ,x) is the objective function value, g(p wind,output ,x)≤0 is the constraint condition of the random unit model, including wind power output scenario constraint, power balance constraint and power flow constraint. Represents the prediction scenario set of the neural network model, support vector machine model and random forest model, p wind,output is the control strategy, and x is the output of other power generation equipment in the power system and the node power angle variable.

[0043] Furthermore, the selection formula of the real-time control strategy is:

[0044]

[0045] Among them, (p wind,output,NN ,x NN )、(p wind,output,SVM ,x SVM ) and (p wind,output,r ,x SVM ) are the real-time control strategies corresponding to the wind power real-time prediction scenario set obtained by power prediction based on the neural network model, support vector machine model and random forest model, and B is the final optimized control strategy. (p wind,output,r ,x r ) is the one with the lowest sum of wind power abandonment and load loss in the real-time control strategy corresponding to the wind power real-time prediction scenario set obtained by prediction based on the neural network model, support vector machine model and random forest model.

[0046] Accordingly, a second aspect of an embodiment of the present invention provides a power grid optimization dispatching system based on dynamic selection of a wind power prediction model, comprising:

[0047] A function construction module, which is used to obtain a prediction error probability density function of each wind power prediction model based on multiple wind power prediction models, and obtain a plurality of wind power prediction error scenario sets;

[0048] A scenario set generation module, which is used to obtain real-time meteorological data of wind turbines, perform power prediction based on the multiple wind power prediction models, obtain a number of wind power real-time prediction values, and generate a number of wind power real-time prediction scenario sets in combination with the corresponding wind power prediction error scenario sets;

[0049] The optimization scheduling module is used to generate a number of real-time control strategies corresponding to the wind power real-time prediction scenario sets through a random unit combination model based on the several wind power real-time prediction scenario sets, and select the real-time control strategy with the lowest value of the sum of wind power abandonment and load loss among the several real-time control strategies as the optimized control strategy for the wind turbine set.

[0050] Accordingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned power grid optimization scheduling method based on dynamic selection of a wind power prediction model.

[0051] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned power grid optimization scheduling method based on dynamic selection of a wind power prediction model.

[0052] The above technical solution of the embodiment of the present invention has the following beneficial technical effects:

[0053] 1. By using the historical forecast data and historical actual data of wind turbines, the nonlinear mapping relationship between various wind power forecast models (neural network model, support vector machine model and random forest model) and meteorological data is established respectively, and the forecast error is calculated to obtain the forecast error probability density function, which can accurately grasp the forecast characteristics and error distribution of each model. When combining real-time meteorological data for power forecasting and generating a real-time forecast scenario set, based on these accurate model characteristics and error analysis, the wind power real-time forecast scenario set generated by Latin hypercube sampling and Cholesky decomposition is more in line with the actual situation, and comprehensively covers the possible range of changes in wind power, providing reliable and comprehensive basic data for subsequent optimization and scheduling, and effectively improving the accuracy and rationality of wind power forecasting and scenario simulation;

[0054] 2. Based on multiple wind power real-time prediction scenario sets, a number of corresponding real-time control strategies are generated through a random unit combination model. The random unit combination model comprehensively considers multiple key factors such as wind power output scenario constraints, power balance constraints and flow constraints, ensuring that the generated control strategies can meet the basic requirements of power grid operation under different wind power scenarios. In addition, since corresponding control strategies are generated for different prediction model scenario sets, this multi-strategy approach greatly enhances the adaptability of power grid dispatching strategies to wind power uncertainty. Finally, by selecting the real-time control strategy with the lowest sum of wind power abandonment and load loss as the optimized control strategy, the power grid dispatch is optimized from the perspective of resource utilization efficiency, effectively reducing the waste of wind power resources and improving the economy and sustainability of the entire power system operation.

[0055] 3. By comprehensively using multiple wind power prediction models such as neural network models, support vector machine models and random forest models, the advantages of different models are fully utilized, such as the powerful learning and fitting capabilities of neural network models, the good performance of support vector machine models on small samples and nonlinear problems, and the characteristics of random forest models in resisting overfitting and processing high-dimensional data. This multi-model fusion method avoids the limitations of a single model, and can complement and verify each other in the entire power grid optimization and scheduling process, such as obtaining prediction error probability density functions, generating real-time wind power prediction scenario sets, and formulating real-time control strategies, thereby improving the overall performance of the entire power grid optimization and scheduling method, allowing the power grid to operate more robustly and efficiently when facing complex and changeable wind power conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart of a power grid optimization dispatching method based on dynamic selection of a wind power prediction model provided by an embodiment of the present invention;

[0057] Figure 2 It is a schematic diagram of a 500kV grid provided in an embodiment of the present invention;

[0058] Figure 3 It is a schematic diagram of wind power prediction value (NN, SVM, RF) and actual value of wind farm A provided by an embodiment of the present invention;

[0059] Figure 4 Schematic diagram of probability density distribution (NN, SVM, RF) of wind power prediction error of wind farm A provided by an embodiment of the present invention;

[0060] Figure 5 Schematic diagram of wind power prediction error time correlation (left) and wind power scenario set (right) provided by an embodiment of the present invention;

[0061] Figure 6 It is a schematic diagram of the total output of other power sources (thermal power, hydropower, and DC) based on the prediction results of NN, SVM, and RF provided in an embodiment of the present invention;

[0062] Figure 7 It is a block diagram of a module of a power grid optimization dispatching device based on dynamic selection of a wind power prediction model provided by an embodiment of the present invention.

[0063] Reference numerals:

[0064] 1. Function construction module, 2. Scenario set generation module, 3. Optimization scheduling module. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0066] Please refer to Figure 1 A first aspect of an embodiment of the present invention provides a method for optimizing power grid scheduling based on dynamic selection of a wind power prediction model, comprising the following steps:

[0067] Step S100 , based on a plurality of wind power prediction models, obtaining a prediction error probability density function of each wind power prediction model, and obtaining a plurality of wind power prediction error scenario sets.

[0068] Step S200, obtain real-time meteorological data of the wind turbine, perform power prediction based on multiple wind power prediction models, obtain several wind power real-time prediction values, and generate several wind power real-time prediction scenario sets in combination with corresponding several wind power prediction error scenario sets.

[0069] Step S300, based on several wind power real-time prediction scenario sets, generate several real-time control strategies corresponding to the wind power real-time prediction scenario sets through a random unit combination model, and select the real-time control strategy with the lowest value of the sum of wind power abandonment and load loss among the several real-time control strategies as the optimized control strategy for the wind turbine set.

[0070] First, the present invention uses the historical data of wind turbines to train multiple prediction models, thereby generating corresponding training models and calculating the prediction error probability density function of each model, and then obtaining several wind power prediction error scenario sets; then, based on the current actual operation data, these trained models are used to predict wind power, and a wind power real-time prediction scenario set is constructed accordingly. Finally, by systematically randomizing the real-time prediction scenario sets generated by different prediction models, with the goal of minimizing the sum of wind power abandonment and load loss, the optimal decision-making scheme is actively selected, thereby achieving the minimum sum of wind power abandonment and load loss in the wind power system.

[0071] The above method no longer selects the prediction method based on the prediction error evaluation index, but transmits multiple prediction results to the dispatching side of the power grid system at the same time, and obtains the corresponding decision results. Finally, the decision result is "actively" selected with the minimum sum of wind power abandonment and load loss as the guide. The above dispatching mode breaks the mutually exclusive relationship between the prediction models and forms a compatible relationship of mutual cooperation and mutual complementation. At the same time, it takes into account that the goal of making decisions in a random environment is to pursue the optimal result in a statistical sense.

[0072] In a specific implementation of an embodiment of the present invention, the types of real-time meteorological data of a wind turbine include: wind speed, wind direction, temperature and / or air pressure. Wind speed is the most critical factor affecting wind power generation, and its size directly determines the output power of the wind turbine. Wind speed is positively correlated with output power. Wind direction affects the wind energy capture efficiency of the wind turbine, and the yaw system needs to adjust the direction of the wind rotor in time to face the incoming wind. The layout of the wind farm also needs to consider wind direction to reduce the impact of wake. Changes in temperature indirectly affect power generation by changing air density. As the temperature rises, the air density decreases, and the output power decreases, and vice versa. Changes in air pressure cause changes in air density, which indirectly affects power generation. Air pressure is positively correlated with output power.

[0073] Specifically, in step S100, based on multiple wind power prediction models, obtaining the prediction error probability density function of each wind power prediction model includes:

[0074] Step S110, obtaining historical prediction data and historical actual data of the wind turbine generator system.

[0075] Historical forecast data usually comes from the forecast results obtained by using wind power forecast models in the past. These models may be built based on different algorithms, such as neural network models, support vector machine models, etc. Historical actual data is the real power output data generated by wind turbines during their actual operation in the past. These data can be collected through the monitoring system of the wind farm, which records the actual power generation of wind turbines at different time points, as well as relevant information such as the meteorological conditions at the time. The time span of data collection needs to be long enough to cover various meteorological conditions and wind power changes. For example, it may be necessary to collect data for months to years to fully reflect seasonal changes, wind power changes under different weather patterns, etc.

[0076] The accuracy and completeness of the data are crucial for subsequent steps. If there are large deviations in the historical forecast data, it may be due to unreasonable parameter settings of the previous model or incorrect input data. If the historical actual data is incomplete, such as the lack of certain key periods (such as data under extreme weather conditions), it will affect the grasp of the true changes in wind power. Therefore, after collecting the data, pre-processing such as data cleaning is required to remove outliers and erroneous data.

[0077] Step S120: establishing nonlinear mapping relationships between meteorological data and a plurality of wind power prediction models based on historical prediction data.

[0078] For neural network models, it has a strong nonlinear fitting ability. When establishing a mapping relationship, the neural network's neuron structure can automatically adjust weights and thresholds by learning patterns in historical forecast data to find the complex nonlinear relationship between meteorological data (such as wind speed, wind direction, temperature, air pressure, etc.) and wind power forecast values. For example, in a multi-layer perceptron (MLP) neural network, the input layer receives meteorological data, which is processed by neurons in the hidden layer, and finally the wind power forecast value is obtained in the output layer.

[0079] The support vector machine model divides the input space (meteorological data space) and the output space (wind power prediction value space) by finding an optimal hyperplane, which can transform nonlinear problems into linear problems to solve. The meteorological data is mapped to a high-dimensional feature space through a kernel function, and a linear relationship is established in this high-dimensional space. Commonly used kernel functions include Gaussian kernel functions, etc. This method can effectively handle the nonlinear mapping between meteorological data and wind power.

[0080] The random forest model is an integrated learning method based on decision trees. It learns the relationship between meteorological data and wind power by building multiple decision trees. Each decision tree is split according to different characteristics of meteorological data, and finally the wind power prediction value is obtained by integrating the results of multiple decision trees. This model can process high-dimensional meteorological data and has good anti-overfitting ability.

[0081] Establishing this mapping relationship can better reflect the complex impact of meteorological factors on wind power. Since there is no simple linear relationship between meteorological data and wind power, for example, wind speed and wind power may have an approximately quadratic relationship within a certain range, but in other cases, it will be complicated by the combined influence of factors such as wind direction and temperature. Through these nonlinear mapping relationships, the comprehensive effect of meteorological factors can be more accurately considered when making power forecasts.

[0082] Step S130: Calculate the prediction error of each wind power prediction model based on historical actual data.

[0083] By calculating the prediction error, we can intuitively understand the accuracy of each wind power prediction model. Different models may have different error performance under different meteorological conditions and power output ranges. For example, when the wind speed fluctuates greatly, some models may produce large errors, while when the wind speed is relatively stable, the errors may be smaller. Analyzing these errors can help discover the advantages and disadvantages of the model and provide a basis for subsequent improvement and selection.

[0084] Step S140: Based on the prediction error, a prediction error probability density function of each wind power prediction model is obtained by kernel density estimation.

[0085] Kernel density estimation is a non-parametric estimation method used to estimate the probability density function of a random variable. For the prediction error of the wind power prediction model, the kernel density estimation method is to regard each prediction error data point as the center of a probability distribution, and then use a kernel function (such as a Gaussian kernel function) to estimate the probability density around the point. After obtaining the prediction error probability density function, the distribution of the prediction error can be fully understood. For example, the probability density function can be used to determine the probability of the prediction error occurring in a certain interval, which is very important for evaluating the reliability of the wind power prediction model. If the prediction error probability density function of a model has a higher peak near zero and is more concentrated, it means that the prediction error of the model is small and relatively stable; conversely, if the probability density function is more dispersed, it means that the prediction error of the model fluctuates greatly and the reliability is low.

[0086] Correspondingly, the corresponding nonlinear mapping relationship of the neural network model is:

[0087]

[0088] The neural network model constructs a mapping relationship through a complex neuron structure, which can fit extremely complex nonlinear functions. In wind power forecasting, the relationship between meteorological data (including wind speed, wind direction, temperature, air pressure and other factors) and wind power is often nonlinear. For example, the impact of wind speed on power is not a simple linear relationship. In different wind speed ranges, its contribution to power is different, and it is also affected by the interaction of other meteorological factors. Neural networks can automatically learn these complex relationship patterns through a large amount of training data.

[0089] During the training process, the neural network can continuously adjust the connection weights between neurons according to the error back propagation algorithm, and can adaptively optimize the mapping relationship to better approximate the relationship between real meteorological data and wind power. For example, when new meteorological data and corresponding wind power data are added to the training set, the neural network can learn these new patterns by updating the weights, thereby improving the accuracy of the prediction.

[0090] Meteorological data is usually high-dimensional, and there may be interactions between various features. Neural networks can effectively handle this situation. They extract and combine features of high-dimensional data through neurons in the hidden layer, capture the interactions between different meteorological features, and generate more accurate wind power forecasts. For example, the combination of wind direction and wind speed may have different effects on wind power. Neural networks can learn this complex combination relationship and reflect it in the mapping relationship.

[0091] Accordingly, the corresponding nonlinear mapping relationship of the support vector machine model is:

[0092]

[0093] The meteorological data is mapped to a high-dimensional feature space through a kernel function. In this high-dimensional space, the relationship between meteorological data and wind power, which was originally nonlinear in a low-dimensional space, may become linearly separable. This conversion capability enables the support vector machine to effectively handle complex nonlinear relationships. For example, when the relationship between meteorological data and wind power presents a complex curve shape, through a suitable kernel function (such as a Gaussian kernel function), the support vector machine can find a hyperplane in the high-dimensional space to divide the data, thereby achieving accurate prediction.

[0094] When searching for the optimal partitioning hyperplane, the support vector machine model focuses on maximizing the classification interval and has good generalization performance. In wind power forecasting, when new meteorological data is input, the wind power can be predicted more accurately based on the learned mapping relationship. Even in the face of some meteorological conditions that slightly deviate from the distribution of training data, the support vector machine can give reasonable prediction results based on its generalization ability and reduce overfitting.

[0095] The support vector machine constructs the mapping relationship by selecting support vectors (sample points located on the classification boundary), which play a key role in determining the classification hyperplane. When dealing with wind power prediction problems, it can focus on the sample points that are most important for dividing the relationship between meteorological data and wind power, thereby making more effective use of limited training sample information, especially when there are fewer sample data. This advantage is more obvious.

[0096] Accordingly, the nonlinear mapping relationship of the random forest model is:

[0097]

[0098] The mapping relationship of the random forest model is achieved by integrating multiple decision trees, which can reduce the overfitting risk of a single decision tree and improve the accuracy and stability of the prediction. In wind power prediction, each decision tree can be regarded as a different way of dividing the relationship between meteorological data and wind power. By integrating the results of multiple decision trees, a more robust prediction value can be obtained. For example, when a decision tree is affected by abnormal meteorological data and produces deviations, other decision trees can play a role in balancing and correcting it.

[0099] Random forests can handle the complex distribution and noise of meteorological data. Since wind power is affected by a variety of meteorological factors, its data distribution may be complex and there may be some noise such as measurement errors. Random forests can adapt to this complex data environment through random sampling and feature selection of multiple decision trees. Each decision tree is trained on different sub-datasets and feature subsets, so that different data patterns and feature relationships can be captured, effectively handling noise in the data and improving the robustness of wind power prediction.

[0100] The random forest model has a certain degree of interpretability. In the process of building the mapping relationship, the relative importance of each feature in the meteorological data to the wind power prediction can be evaluated by analyzing the usage frequency and importance of the features in each decision tree. For example, the importance ranking of factors such as wind speed, wind direction, temperature, and air pressure in predicting wind power can be determined, which is of great significance for understanding the impact of meteorological factors and further optimizing the prediction model.

[0101] Among them, f NN(·),f SVM (·) and f RF (·) are the mapping relationships between wind power prediction values ​​and meteorological data X based on neural network model, support vector machine model and random forest model, respectively. l , W l and are the activation function, weight vector and bias vector of the lth layer of the neural network, respectively, and l = 1, 2, 3, ..., L, L is the number of layers of the neural network, W and b NN are the weight matrix and bias matrix of the neural network, N SVM is the number of support vectors in the support vector machine model, α i is the Lagrange multiplier of the i-th support vector of the support vector machine model, y i is the true label of the i-th support vector of the support vector machine model, K SVM (X i ,X) is the kernel function of the support vector machine model, γ is a positive parameter, X i is the X corresponding to the i-th sample point, b SVM is the bias term of the support vector machine model, N RF is the number of decision trees in the random forest model, is the predicted value of the i-th decision tree for input X, p wind,NN 、p wind,SVM and p wind,RF These are the wind power prediction values ​​obtained based on the neural network model, support vector machine model and random forest model respectively.

[0102] Furthermore, the corresponding prediction error probability density function of the neural network model is:

[0103]

[0104] Δp i wind,NN =p i wind,0 -f NN (X i );

[0105] The corresponding prediction error probability density function of the support vector machine model is:

[0106]

[0107] Δp i wind,SVM =p i wind,0 -f SVM (X i );

[0108] The corresponding prediction error probability density function of the random forest model is:

[0109]

[0110] Δp i wind,RF =p i wind,0 -f RF (X i );

[0111] Among them, K(·) is the kernel density estimation function, h is the bandwidth, N S is the number of samples, Δp wind is the wind power prediction error random variable, X i and p i wind,0 are the meteorological factor input vector and historical actual wind power of the i-th historical data sample point, Δp i wind,NN , Δp i wind,SVM and Δp i wind,RF are the wind power prediction errors of the i-th historical data sample point of the neural network model, support vector machine model and random forest model, respectively. and They are the prediction error probability density functions of the neural network model, support vector machine model and random forest model respectively.

[0112] Specifically, the step S200 of combining the corresponding wind power prediction error scenario sets to generate wind power real-time prediction scenario sets includes:

[0113] Step S201, based on the prediction error probability density function, a corresponding wind power prediction error scenario set is generated by Latin hypercube sampling and Cholesky decomposition:

[0114]

[0115] Among them, Ω NN,err ,Ω SVM,err and Ω RF,err are the wind power prediction error scenario sets of the neural network model, support vector machine model and random forest model, (Δp wind,NN ) i 、(Δp wind,SVM ) i and (Δp wind,RF ) i The i-th wind power prediction error scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., Nsg , N sg is the number of scenes.

[0116] Among them, Latin hypercube sampling is a stratified sampling technique. In the context of wind power forecasting scenario generation, the value range of each input variable (here, wind power real-time forecast value and its forecast error related variables) is divided into non-overlapping equal probability intervals. For example, for wind power forecast error, if its value range is from -100kW to 100kW, assuming that we want to divide it into 10 intervals, then the width of each interval is 20kW, and each interval has the same probability of being sampled. Then a sample point is randomly selected from each interval. Unlike simple random sampling, Latin hypercube sampling ensures that the samples are more evenly distributed throughout the value range, which can better cover the various possible values ​​of the variables. In wind power forecasting, this uniformly distributed sample can more comprehensively represent various possible combinations of wind power real-time forecast values ​​and errors. These sample points can serve as the basis for constructing wind power real-time forecasting scenarios. For example, by extracting different wind power real-time forecast values ​​and corresponding forecast error sample points, multiple different possible wind power states can be generated, thereby providing a diverse data source for constructing scenario sets.

[0117] These sample points are transformed by the lower triangular matrix obtained through Cholesky decomposition. This process injects correlation information into the sample points, so that the generated scenario set not only includes various possible values ​​of wind power real-time prediction value and prediction error, but also considers the relationship between the two. The final wind power real-time prediction scenario set is a collection of multiple scenarios, each of which represents a possible real-time change of wind power. These scenarios can be used to simulate the impact of wind power on power system operation under different circumstances, such as in the random unit combination model, to provide a basis for formulating different real-time control strategies.

[0118] Step S202: obtaining a wind power real-time prediction scenario set based on the wind power real-time prediction value and the wind power prediction error scenario set:

[0119]

[0120] Among them, Ω NN ,Ω SVM and Ω RF The wind power real-time prediction scenario sets of neural network model, support vector machine model and random forest model, p wind,NN,pre , and p wind,RF,pre are the real-time wind power prediction values ​​of the neural network model, support vector machine model and random forest model, respectively. (p wind,NN ) i 、(pwind,SVM ) i and (p wind,RF ) i The i-th wind power prediction scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., N sg , N sg is the number of scenes.

[0121] In addition, when the wind power prediction model includes: neural network model, support vector machine model and random forest model, the corresponding random unit combination model is:

[0122]

[0123] Among them, f(p wind,output ,x) is the objective function of the random unit combination model, z(p wind,output ,x) is the objective function value, g(p wind,output ,x)≤0 is the constraint condition of the random unit model, including wind power output scenario constraint, power balance constraint and power flow constraint. Represents the prediction scenario set of the neural network model, support vector machine model and random forest model, p wind,output is the control strategy, and x is the output of other power generation equipment in the power system and the node power angle variable.

[0124] The control strategy is the key variable for optimization. By adjusting the control strategy, it is possible to determine which units should be started, which should be shut down, and their respective power generation. Considering the uncertainty of wind power, the control strategy needs to be flexibly adjusted according to different wind power forecast scenarios to achieve the optimal objective function. For example, when the wind power forecast scenario shows that the wind power is high, the output of other power generation equipment can be appropriately reduced to increase the consumption of wind power.

[0125] In addition, the wind power output scenario constraint is a key constraint to consider the uncertainty of wind power. Since wind power will vary according to factors such as different meteorological conditions, Ω NN |Ω SVM |Ω RFThe wind power scenarios contained in these forecast scenario sets are different. The wind power output scenario constraint requires that in each scenario, the wind power output must comply with the actual physical laws and the range of the forecast model. For example, wind power cannot be negative, and its output must be within a reasonable range under different meteorological conditions such as wind speed. The power balance constraint is a basic constraint on the operation of the power system. In the random unit combination model, since multiple wind power forecast scenarios are considered, the power balance constraint needs to be met in each scenario. It is required that when formulating unit combinations and control strategies, the output of other power generation equipment should be reasonably arranged according to the changes in wind power in different scenarios. The flow constraint mainly considers the physical characteristics of the power grid, such as the capacity limit of the line, the upper and lower limits of the node voltage, etc.; in the random unit combination model, the flow constraint ensures that the power network can operate safely and stably under different wind power scenarios. For example, when the wind power suddenly increases or decreases, the flow distribution in the power grid will change. It is necessary to use reasonable unit combinations and control strategies to ensure that the flow is within the allowable range and avoid problems such as line overload.

[0126] Furthermore, the selection formula of the real-time control strategy is:

[0127]

[0128] Among them, (p wind,output,NN ,x NN )、(p wind,output,SVM ,x SVM ) and (p wind,output,r ,x SVM ) are the real-time control strategies corresponding to the wind power real-time prediction scenario set obtained by power prediction based on the neural network model, support vector machine model and random forest model, and B is the final optimized control strategy. (p wind,output,r ,x r ) is the one with the lowest sum of wind power abandonment and load loss in the real-time control strategy corresponding to the wind power real-time prediction scenario set obtained by prediction based on the neural network model, support vector machine model and random forest model.

[0129] Below, the specific process of the present invention is further described with an actual application scenario:

[0130] Based on a DC receiving-end power grid, the 500kV grid diagram of the provincial grid is as follows: Figure 2 As shown in Figure 1, the power generation units participating in the optimization dispatch of the provincial grid include 37 thermal power units, 10 cascade hydropower stations, and 12 wind farms. The grid structure studied is 220kV and above lines, which contains a total of 489 nodes after equivalent simplification.

[0131] The following is an example of a certain day: The prediction results of wind farm A on a certain day are as follows: Figure 3, where NN, SVM and RF represent neural network, support vector machine and random forest respectively.

[0132] Figure 4 The prediction error probability distribution of wind farm A under various prediction methods is shown. It can be seen that the prediction error distribution of each prediction method presents different "skewness", and the gap between the distributions is large, indicating that different prediction models with the same prediction level have obviously different "overestimation" or "underestimation" prediction "preferences".

[0133] Based on the prediction results obtained by the above three wind power prediction methods, the time correlation between the prediction errors is further analyzed, and based on Latin hypercube sampling and Cholesky decomposition, the wind power prediction error scenario sets of each prediction model are generated respectively in combination with the corresponding prediction error probability density function. Based on the wind power real-time prediction value of each prediction model and the corresponding wind power prediction error scenario set, the wind power real-time prediction scenario set of each prediction model is obtained, such as Figure 5 shown. Figure 5 It can be seen that in terms of the time correlation of the prediction error of each prediction method, as a whole, the correlation shows a trend of gradually decreasing as the time section interval increases; locally, there are some differences between the correlations of the prediction errors generated by different prediction methods. Based on the wind power prediction value and the wind power prediction error, a wind power scenario set can be generated. Since the scenario set has a great relationship with the wind power prediction value and the probability distribution of the prediction error, the wind power scenarios obtained by different prediction methods have obvious visible differences. However, on the whole, the scenario set obtained based on any prediction method can basically cover the fluctuation range of the actual wind power value.

[0134] In the stochastic optimization scheduling model considering the uncertainty of wind power, only the wind power output has uncertainty. Other power sources including thermal power, hydropower, direct current, etc. do not consider uncertainty and can all be output as planned. Figure 6 The total output of other power sources obtained based on the prediction results of the three prediction methods, NN, SVM and RF, is shown. It can be seen that the scheduling plans obtained based on different prediction results are quite different. Among them, at 14:00, the total output of other power sources based on SVM and the total output of other power sources based on RF reached 621.37MW, which is also the largest output gap between the three decision results.

[0135] Based on the 100-day scheduling plan and actual observations, the prediction model with the lowest expected system operation cost is "actively" selected from the three prediction models of neural network, support vector machine, and random forest for scheduling (dynamic scheduling mode).

[0136] Table 1 shows the number of times each prediction method was selected. It can be seen that the prediction results and decision results based on SVM were selected the most times, with 51 days selected in the 100-day calculation period. In other words, the decision based on the SVM prediction method had the lowest sum of system wind curtailment and load loss in these 51 days.

[0137] Table 1 The number of times each prediction method was selected (100 days)

[0138]

[0139] Accordingly, please refer to Figure 7 A second aspect of an embodiment of the present invention provides a power grid optimization dispatching system based on dynamic selection of a wind power prediction model, comprising:

[0140] Function construction module 1, which is used to obtain the prediction error probability density function of each wind power prediction model based on multiple wind power prediction models, and obtain several wind power prediction error scenario sets;

[0141] Scenario set generation module 2, which is used to obtain real-time meteorological data of wind turbines, perform power prediction based on multiple wind power prediction models, obtain several wind power real-time prediction values, and generate several wind power real-time prediction scenario sets in combination with corresponding several wind power prediction error scenario sets;

[0142] The optimization scheduling module 3 is used to generate a number of real-time control strategies corresponding to the wind power real-time prediction scenario sets based on a number of wind power real-time prediction scenario sets through a random unit combination model, and select the real-time control strategy with the lowest sum of wind power abandonment and load loss among the several real-time control strategies as the optimized control strategy for the wind turbine set.

[0143] Accordingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned power grid optimization scheduling method based on dynamic selection of a wind power prediction model.

[0144] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned power grid optimization scheduling method based on dynamic selection of a wind power prediction model.

[0145] The embodiment of the present invention aims to protect a method and device for optimizing and dispatching a power grid based on the dynamic selection of a wind power prediction model, comprising the following steps: based on multiple wind power prediction models, obtaining the prediction error probability density function of each wind power prediction model, and obtaining several wind power prediction error scenario sets; obtaining real-time meteorological data of wind turbines, and performing power predictions based on multiple wind power prediction models to obtain several wind power real-time prediction values, and combining the corresponding wind power prediction error scenario sets to generate several wind power real-time prediction scenario sets; based on several wind power real-time prediction scenario sets, generating several real-time control strategies corresponding to the wind power real-time prediction scenario sets through a random unit combination model, and selecting the real-time control strategy with the lowest sum of wind power abandonment and load loss among the several real-time control strategies as the optimized control strategy for the wind turbine. The above technical scheme has the following effects:

[0146] 1. By using the historical forecast data and historical actual data of wind turbines, the nonlinear mapping relationship between various wind power forecast models (neural network model, support vector machine model and random forest model) and meteorological data is established respectively, and the forecast error is calculated to obtain the forecast error probability density function, which can accurately grasp the forecast characteristics and error distribution of each model. When combining real-time meteorological data for power forecasting and generating a real-time forecast scenario set, based on these accurate model characteristics and error analysis, the wind power real-time forecast scenario set generated by Latin hypercube sampling and Cholesky decomposition is more in line with the actual situation, and comprehensively covers the possible range of changes in wind power, providing reliable and comprehensive basic data for subsequent optimization and scheduling, and effectively improving the accuracy and rationality of wind power forecasting and scenario simulation;

[0147] 2. Based on multiple wind power real-time prediction scenario sets, a number of corresponding real-time control strategies are generated through a random unit combination model. The random unit combination model comprehensively considers multiple key factors such as wind power output scenario constraints, power balance constraints and flow constraints, ensuring that the generated control strategies can meet the basic requirements of power grid operation under different wind power scenarios. In addition, since corresponding control strategies are generated for different prediction model scenario sets, this multi-strategy approach greatly enhances the adaptability of power grid dispatching strategies to wind power uncertainty. Finally, by selecting the real-time control strategy with the lowest sum of wind power abandonment and load loss as the optimized control strategy, the power grid dispatch is optimized from the perspective of resource utilization efficiency, effectively reducing the waste of wind power resources and improving the economy and sustainability of the entire power system operation.

[0148] 3. By comprehensively using multiple wind power prediction models such as neural network models, support vector machine models and random forest models, the advantages of different models are fully utilized, such as the powerful learning and fitting capabilities of neural network models, the good performance of support vector machine models on small samples and nonlinear problems, and the characteristics of random forest models in resisting overfitting and processing high-dimensional data. This multi-model fusion method avoids the limitations of a single model, and can complement and verify each other in the entire power grid optimization and scheduling process, such as obtaining prediction error probability density functions, generating real-time wind power prediction scenario sets, and formulating real-time control strategies, thereby improving the overall performance of the entire power grid optimization and scheduling method, allowing the power grid to operate more robustly and efficiently when facing complex and changeable wind power conditions.

[0149] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0150] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0151] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0152] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A power grid optimization dispatching method based on dynamic selection of wind power prediction model, characterized in that: The steps include: Based on multiple wind power prediction models, obtaining a prediction error probability density function of each wind power prediction model, and obtaining a plurality of wind power prediction error scenario sets; Acquire real-time meteorological data of the wind turbines, perform power forecasts based on the multiple wind power forecast models, obtain a number of wind power real-time forecast values, and generate a number of wind power real-time forecast scenario sets in combination with the corresponding wind power forecast error scenario sets; Based on the several wind power real-time prediction scenario sets, several real-time control strategies corresponding to the wind power real-time prediction scenario sets are generated one by one through a random unit combination model, and the real-time control strategy with the lowest value of the sum of wind power abandonment and load loss among the several real-time control strategies is selected as the optimized control strategy for the wind turbine set.

2. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to claim 1 is characterized in that: The types of real-time meteorological data of the wind turbine generator set include: wind speed, wind direction, temperature and / or air pressure.

3. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to claim 1 is characterized in that: The step of obtaining a prediction error probability density function of each wind power prediction model based on multiple wind power prediction models includes: Obtain historical forecast data and historical actual data of wind turbines; Based on the historical prediction data, respectively establishing nonlinear mapping relationships between meteorological data and a plurality of wind power prediction models; Based on the historical actual data, calculating the prediction error of each wind power prediction model; Based on the prediction error, a prediction error probability density function of each wind power prediction model is obtained respectively by kernel density estimation.

4. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to claim 3 is characterized in that: The wind power prediction model includes: a neural network model, a support vector machine model and a random forest model; The corresponding nonlinear mapping relationship of the neural network model is: The corresponding nonlinear mapping relationship of the support vector machine model is: K SVM (X i ,X)=exp(-γ||X i -X|| 2 ); The nonlinear mapping relationship corresponding to the random forest model is: Among them, f NN (·),f SVM (·) and f RF (·) are the mapping relationships between wind power prediction values ​​and meteorological data X based on neural network model, support vector machine model and random forest model, respectively. l , W l and are the activation function, weight vector and bias vector of the lth layer of the neural network, respectively, and l = 1, 2, 3, ..., L, L is the number of layers of the neural network, W and b NN are the weight matrix and bias matrix of the neural network, N SVM is the number of support vectors of the support vector machine model, α i is the Lagrange multiplier of the i-th support vector of the support vector machine model, y i is the true label of the i-th support vector of the support vector machine model, K SVM (X i ,X) is the kernel function of the support vector machine model, γ is a positive parameter, X i is the X corresponding to the i-th sample point, b SVM is the bias term of the support vector machine model, N RF is the number of decision trees in the random forest model, is the predicted value of the i-th decision tree for input X, p wind,NN 、p wind,SVM and p wind,RF These are the wind power prediction values ​​obtained based on the neural network model, support vector machine model and random forest model respectively.

5. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to claim 4 is characterized in that: The corresponding prediction error probability density function of the neural network model is: The corresponding prediction error probability density function of the support vector machine model is: The corresponding prediction error probability density function of the random forest model is: Among them, K(·) is the kernel density estimation function, h is the bandwidth, N S is the number of samples, Δp wind is the wind power prediction error random variable, X i and are the meteorological factor input vector and historical actual wind power of the i-th historical data sample point, and are the wind power prediction errors of the i-th historical data sample point of the neural network model, support vector machine model and random forest model, respectively. and They are the prediction error probability density functions of the neural network model, support vector machine model and random forest model respectively.

6. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to any one of claims 1 to 5, characterized in that: The generating of wind power real-time prediction scenario sets respectively by combining the corresponding wind power prediction error scenario sets comprises: Based on the prediction error probability density function, the corresponding wind power prediction error scenario set is generated by Latin hypercube sampling and Cholesky decomposition: Among them, Ω NN,err ,Ω SVM,err and Ω RF,err are the wind power prediction error scenario sets of the neural network model, support vector machine model and random forest model, (Δp wind,NN ) i 、(Δp wind,SVM ) i and (Δp wind,RF ) i The i-th wind power prediction error scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., N sg , N sg is the number of scenes; Based on the wind power real-time prediction value and the wind power prediction error scenario set, a wind power real-time prediction scenario set is obtained: Among them, Ω NN ,Ω SVM and Ω RF The wind power real-time prediction scenario sets of neural network model, support vector machine model and random forest model, p wind,NN,pre , and p wind,RF,pre are the real-time wind power prediction values ​​of the neural network model, support vector machine model and random forest model, (p wind,NN ) i 、(p wind,SVM ) i and (p wind,RF ) i The i-th wind power prediction scenario generated by the neural network model, support vector machine model and random forest model, i = 1, 2, 3, ..., N sg , N sg is the number of scenes.

7. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to any one of claims 1 to 5, characterized in that: The wind power prediction model includes: a neural network model, a support vector machine model and a random forest model; The random unit combination model is: Among them, f(p wind,output ,x) is the objective function of the random unit combination model, z(p wind,output ,x) is the objective function value, g(p wind,output ,x)≤0 is the constraint condition of the random unit model, including wind power output scenario constraint, power balance constraint and power flow constraint. Represents the prediction scenario set of the neural network model, support vector machine model and random forest model, p wind,output is the control strategy, and x is the output of other power generation equipment in the power system and the node power angle variable.

8. The power grid optimization dispatching method based on dynamic selection of wind power prediction model according to claim 7 is characterized in that: The selection formula of the real-time control strategy is: Among them, (p wind,output,NN ,x NN )、(p wind,output,SVM ,x SVM ) and (p wind,output,r ,x SVM ) are the real-time control strategies corresponding to the wind power real-time prediction scenario set obtained by power prediction based on the neural network model, support vector machine model and random forest model, and B is the final optimized control strategy. (p wind,output,r ,x r ) is the one with the lowest sum of wind power abandonment and load loss in the real-time control strategy corresponding to the wind power real-time prediction scenario set obtained by prediction based on the neural network model, support vector machine model and random forest model.

9. A power grid optimization dispatching system based on dynamic selection of wind power prediction model, characterized in that: include: A function construction module, which is used to obtain a prediction error probability density function of each wind power prediction model based on multiple wind power prediction models, and obtain a plurality of wind power prediction error scenario sets; A scenario set generation module, which is used to obtain real-time meteorological data of wind turbines, perform power prediction based on the multiple wind power prediction models, obtain a number of wind power real-time prediction values, and generate a number of wind power real-time prediction scenario sets in combination with the corresponding wind power prediction error scenario sets; The optimization scheduling module is used to generate a number of real-time control strategies corresponding to the wind power real-time prediction scenario sets through a random unit combination model based on the several wind power real-time prediction scenario sets, and select the real-time control strategy with the lowest value of the sum of wind power abandonment and load loss among the several real-time control strategies as the optimized control strategy for the wind turbine set.

10. An electronic device, characterized in that: include: at least one processor; And a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the power grid optimization scheduling method based on dynamic selection of wind power prediction model as described in any one of claims 1-8.

Images