An extreme weather new energy power generation multi-disaster influence analysis method and system

Abstract

The application discloses a new energy power generation multi-disaster influence analysis method and system under extreme weather, and the method comprises the following steps: constructing a random forest regression model according to meteorological data containing a target meteorological factor and normalized new energy power generation power data; introducing an improved Bayesian optimization method, taking the determination coefficient of the random forest regression model on the verification set as the objective function, optimizing the hyperparameters of the random forest regression model, the hyperparameters including the number of decision trees, the maximum depth and the minimum split sample number, and obtaining a disaster superposition influence rate model; inputting real-time meteorological data into the disaster superposition influence rate model, and outputting the new energy power generation power corresponding to the real-time meteorological data by the disaster superposition influence rate model. The influence of extreme weather on new energy power generation efficiency is comprehensively predicted, and the parameters in the random forest are optimized by introducing the improved Bayesian optimization, so that the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of renewable energy power generation, and in particular relates to a method, system and readable storage medium for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions. Background Art

[0002] The impact of extreme weather events on renewable energy power generation systems has become a key factor constraining their stable operation and development. For example, cold snaps can cause wind turbines to freeze, heavy fog can reduce solar radiation, and snowfall can cover photovoltaic panels, all of which severely impact power generation efficiency and system safety. Existing research has largely focused on the impact of single weather events on specific renewable energy generation methods, lacking a systematic, comprehensive analysis of multiple extreme weather conditions. For example, a single analysis of the impact of cold snaps, high winds, and snowfall on wind power generation has been conducted. The impact of extreme weather conditions on photovoltaic power generation has been discussed separately, including the impact of sandstorms, heavy rain, and snowfall on photovoltaic power generation.

[0003] The methods used above mostly analyze the impact of a single weather factor on renewable energy power generation. However, for the three extreme weather conditions of cold waves, heavy fog and snowfall, they are very likely to occur together rather than alone. If the method of analyzing a single weather factor is adopted, it is difficult to comprehensively evaluate and predict the comprehensive impact of extreme weather on the entire renewable energy power generation system. Summary of the Invention

[0004] The present invention provides a method and system for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions, which is used to solve the technical problem that it is difficult to comprehensively evaluate and predict the comprehensive impact of extreme weather on the entire renewable energy power generation system by analyzing a single weather factor.

[0005] In a first aspect, the present invention provides a method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions, comprising:

[0006] Acquiring meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation;

[0007] Determine the linear correlation between the meteorological factor and the renewable energy power generation according to the Pearson correlation coefficient, select the meteorological factor with an absolute value of the linear correlation greater than a preset threshold, and define it as the target meteorological factor;

[0008] Constructing a random forest regression model based on meteorological data including the target meteorological factor and normalized new energy power generation data;

[0009] An improved Bayesian optimization method was introduced, with the coefficient of determination of the random forest regression model on the validation set as the objective function. The hyperparameters of the random forest regression model were optimized, including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model.

[0010] The real-time meteorological data is input into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation corresponding to the real-time meteorological data.

[0011] In a second aspect, the present invention provides a system for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions, comprising:

[0012] an acquisition module configured to acquire meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation;

[0013] a screening module configured to determine a linear correlation between the meteorological factor and the renewable energy power generation according to a Pearson correlation coefficient, screen out meteorological factors whose absolute value of the linear correlation is greater than a preset threshold, and define them as target meteorological factors;

[0014] A construction module is configured to construct a random forest regression model based on the meteorological data including the target meteorological factor and the normalized new energy power generation data;

[0015] An optimization module is configured to introduce an improved Bayesian optimization method, using the coefficient of determination of the random forest regression model on the validation set as an objective function, to optimize the hyperparameters of the random forest regression model, including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model;

[0016] The output module is configured to input real-time meteorological data into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation power corresponding to the real-time meteorological data.

[0017] According to a third aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively 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 can execute the steps of the method for analyzing the impact of multiple disasters on new energy power generation under extreme weather conditions according to any embodiment of the present invention.

[0018] In a fourth aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon. When the program instructions are executed by a processor, the processor executes the steps of the method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions according to any embodiment of the present invention.

[0019] The method and system for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions in this application, based on actual meteorological data and corresponding renewable energy power generation data, completes the identification of extreme weather and establishes a mathematical model between extreme weather and renewable energy power generation, and analyzes the impact of extreme weather on renewable energy power generation. In addition, a random forest model will be established, using multiple meteorological factors and historical power generation data as basic data to comprehensively predict the impact of extreme weather on renewable energy power generation efficiency, and introduces improved Bayesian optimization to optimize the parameters in the random forest to improve the accuracy of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flowchart of a method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions provided by one embodiment of the present invention;

[0022] Figure 2 Four meteorological element numerical maps collected according to an embodiment of the present invention;

[0023] Figure 3 A graph of renewable energy power generation data collected according to an embodiment of the present invention;

[0024] Figure 4 This is a wind power generation random forest fitting and prediction diagram according to an embodiment of the present invention;

[0025] Figure 5 This is a random forest fitting and prediction graph of photovoltaic power generation according to an embodiment of the present invention;

[0026] Figure 6 This is a structural block diagram of a system for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions, provided by one embodiment of the present invention;

[0027] Figure 7 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0029] See also Figure 1 , which shows a flow chart of a method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions in this application.

[0030] like Figure 1 As shown in the figure, the analysis method of multiple disaster impacts on renewable energy power generation under extreme weather conditions specifically includes the following steps:

[0031] Step S101 : Acquire meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation.

[0032] In this step, data collection mainly includes the collection of meteorological data and power generation data. Among them, meteorological data includes historical meteorological data of cold waves, heavy fog and snowfall, such as air temperature, wind speed, visibility, snowfall, etc.; power generation data is the power value of wind power and photovoltaic power generation in the corresponding time period.

[0033] Data cleaning: Data cleaning primarily involves processing missing values ​​and outliers in the collected data. If the data is relatively small, you can directly delete the rows with missing values ​​and outliers. If the data is relatively large, you can fill it with the mean.

[0034] Step S102 : determining the linear correlation between the meteorological factor and the renewable energy power generation according to the Pearson correlation coefficient, screening out meteorological factors whose absolute values ​​of the linear correlation are greater than a preset threshold, and defining them as target meteorological factors.

[0035] In this step, the relationship between the meteorological element data and the renewable energy power generation data is analyzed. The Pearson correlation coefficient is used to measure the relationship between the two variables. The variable related to renewable energy power generation is selected (|Pearson correlation coefficient| ≥ a, a is 0.2-0.5). The calculation formula is:

[0036] ,

[0037] Where, is the Pearson correlation coefficient value, and For the corresponding meteorological data and new energy power generation data, and The Pearson correlation coefficient is a value between -1 and 1, with values ​​close to 1 or -1 indicating strong positive and negative correlations, and values ​​close to 0 indicating no linear correlation.

[0038] After completing the selection of meteorological factors, these related variables and renewable energy power generation data are normalized to facilitate subsequent modeling. The processing process is as follows:

[0039] ,

[0040] Where, is the normalized value, and are the minimum and maximum values ​​in the data, respectively.

[0041] Step S103: constructing a random forest regression model based on the meteorological data including the target meteorological factor and the normalized new energy power generation data.

[0042] In this step, before establishing a random forest regression model based on the obtained meteorological data and new energy power generation data, it is necessary to identify extreme weather conditions. The judgment criteria include:

[0043] The criteria for a cold wave in northern China is a drop of more than 10°C in 24 hours, or more than 12°C in 48 hours, with the minimum temperature below 4°C. The criteria for a cold wave in southern China is a drop of more than 8°C in 24 hours, or more than 10°C in 48 hours, with the minimum temperature below 5°C. The Guangdong Provincial Meteorological Bureau defines a cold wave as a drop of more than 8°C in the average daily temperature within a single day, or a sharp drop of 10°C in the average temperature within two days, with the minimum temperature dropping to 5°C or below.

[0044] Heavy fog is defined as fog when visibility is 500-1000 meters, dense fog when visibility is 50-500 meters, and very dense fog when visibility is less than 50 meters. Generally, heavy fog refers to dense fog and very dense fog when visibility is less than 500 meters.

[0045] For snowfall: the 24-hour precipitation is used as the classification standard, among which, precipitation of 0.1-2.4 mm is light snow, 1.3-3.7 mm is light to moderate snow, 2.5-4.9 mm is moderate snow, 3.8-7.4 mm is moderate to heavy snow, 5.0-9.9 mm is heavy snow, 7.5-14.9 mm is heavy to blizzard, and precipitation reaching or exceeding 10 mm is blizzard.

[0046] Building a random forest regression model specifically includes:

[0047] Sample selection: Multiple subsamples are randomly selected from meteorological data and renewable energy power generation data. Duplicates are allowed in multiple subsamples, that is, a data point appears multiple times in the same subsample.

[0048] Constructing a decision tree: For each randomly sampled subsample, a decision tree is independently constructed. When constructing each node of the decision tree, instead of selecting the best feature from all features, a portion of features is randomly selected, and then the best feature is selected from them for splitting.

[0049] Decision tree splitting: Use selected features to split each decision tree node. Common splitting criteria include information gain or Gini impurity. The expression of Gini impurity is:

[0050] ,

[0051] Where, is the Gini impurity, For the first The proportion of classes, is the total number of categories.

[0052] The expression of information gain is:

[0053] ,

[0054] Where, For a split dataset, the split attribute is For attributes All possible values ​​of For the current attribute Pick hour A subset of For the dataset Entropy before classification, Indicates that the attribute Pick hour The entropy of a subset of Indicates that the attribute The above dataset The information gain obtained by partitioning.

[0055] Result integration: For regression problems, the average of the predicted values ​​is calculated. For regression tasks, the average formula is as follows:

[0056] ,

[0057] Where, is the average of the predicted values, For the A decision tree for the sample The predicted value of is the total number of decision trees;

[0058] Combine the regression decision trees to get the corresponding random forest model, where each tree has a Give independent prediction values The final result is calculated by averaging the outputs of all trees, thereby improving the overall prediction accuracy and reducing the risk of overfitting. It represents the random forest input The integrated regression prediction process.

[0059] In step S104, an improved Bayesian optimization method is introduced, and the determination coefficient of the random forest regression model on the validation set is used as the objective function to optimize the hyperparameters of the random forest regression model, including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model.

[0060] In this step, the determination coefficient of random forest in the validation set is The maximum is the optimization goal, and the objective function is constructed;

[0061] From the hyperparameter space Randomly sample hyperparameter combinations in , respectively evaluate the corresponding function values , construct the initial data set, the expression is:

[0062] ,

[0063] Where, is the initial data set, is the initial sampling point number, For the Sampling hyperparameter combinations, for The response value at which the function is evaluated.

[0064] Leveraging Datasets , establish a Gaussian process regression model as an approximation of the objective function, the expression is:

[0065] ,

[0066] Where, For the objective function in the hyperparameter combination The response value under For the hyperparameter combination The mean and variance of the lower Gaussian process prediction, For the hyperparameter combination The average value under For the hyperparameter combination The variance under .

[0067] The expected improvement function is selected as the acquisition function to measure the value of sampling new points under the current proxy model;

[0068] Based on the improved acquisition function, the current optimal test point is selected, and the expression is:

[0069] ,

[0070] Where, To represent the optimal test point currently selected, even if the hyperparameter combination with the largest acquisition function value is To make the function Get the maximum value of the parameter , The value of the acquisition function that is expected to be improved is used to guide the selection of the next sampling point.

[0071] The sampled Applied to random forest regression model training and performance evaluation to obtain the true objective function value , and update the dataset, the expression is:

[0072] ,

[0073] Where, To represent the hyperparameter combination The true evaluation value of the objective function.

[0074] Repeat the modeling and sampling process, iteratively update the proxy model and data set, and finally obtain the optimal hyperparameter combination. Based on the optimal hyperparameter combination, a disaster superposition impact rate model is constructed. The expression of the optimal hyperparameter combination is:

[0075] ,

[0076] Where, To express the function Reaching the minimum Parameters, To express the optimal hyperparameter combination, even if the objective function Get the parameter with the minimum value.

[0077] It should be noted that the expression of the objective function is:

[0078] ,

[0079] Where, is the objective function value, For the The actual power generation value at the moment, for The model prediction value at time t, is the sample weight, For each weight, the parameters can be optimized by Bayesian optimization; It is a physical factor. For wind power generation, it is wind speed. For photovoltaic power generation, it is radiation. is an indicator function that takes a value of 1 when the condition is met and a value of 0 when it is not met. This method is used to avoid the situation where the power generation forecast is still positive when the physical conditions are insufficient. The optimization goal is to minimize the objective function.

[0080] The expression of the acquisition function is:

[0081] ,

[0082] ,

[0083] Where, is the optimal target at the current observation point, is the exploration function, which is used to control the trade-off between exploration and utilization. is the cumulative distribution function of the standard normal distribution, For the Gaussian process regression model at point The standard deviation of the prediction at is the probability density function of the standard normal distribution, Standardized variables, measured at the point The improvement in the model predictions at is the uncertainty adjustment factor, which can be selected through parameter optimization. For physical factors, wind speed is selected for wind power generation and radiation is selected for photovoltaic power generation; is an indicator function, which takes the value 1 when the condition is met and takes the value 0 when it is not met. is the predicted mean.

[0084] Step S105 , inputting the real-time meteorological data into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation corresponding to the real-time meteorological data.

[0085] In summary, the impact of various extreme weather conditions (cold waves, heavy fog, rainfall, etc.) on renewable energy power generation is comprehensively considered. Based on actual meteorological data and corresponding renewable energy power generation data, the method of this application identifies extreme weather conditions and establishes a mathematical model linking extreme weather and renewable energy power generation, analyzing the impact of extreme weather on renewable energy power generation. Furthermore, a random forest model is established, using multiple meteorological factors and historical power generation data as foundational data to comprehensively predict the impact of extreme weather on renewable energy power generation efficiency. Improved Bayesian optimization is introduced to optimize the parameters in the random forest to improve prediction accuracy.

[0086] In a specific embodiment, a total of 450 days of meteorological data and new energy power generation data are selected, and the meteorological data value graphs and new energy power generation data graphs are as follows: Figure 2 , Figure 3 When training the random forest model, the ratio of the training set to the validation set was 8:2. After Bayesian optimization, it showed good results in both wind power generation and photovoltaic power generation. In photovoltaic power generation, the mean square error (MSE) on the training set was 4.99, and the determination coefficient was 1. is 0.997, and the MSE on the test set is 56.78. is 0.969, and the optimal parameters are: n_estimators=73, max_depth=19, min_samples_split=3; in wind power generation, the mean square error (MSE) on the training set is 0.3878, and the determination coefficient is 0.95, and the MSE on the test set is 1.7574. The optimal parameters are n_estimators (the number of decision trees in the random forest) = 50, max_depth (maximum tree depth) = 13, and min_samples_split (minimum number of sample splits) = 2, which also shows good performance.

[0087] like Figure 4 and Figure 5 As shown in Figure 2, after Bayesian optimization, the random forest model showed good performance in wind power generation and photovoltaic power generation simulation and prediction. In the case of wind power generation, the mean square error (MSE) on the training set was 0.3878, and the determination coefficient was is 0.95, and the MSE on the test set is 1.7574. is 0.84; in photovoltaic power generation, the mean square error (MSE) on the training set is 4.99, and the determination coefficient is is 0.997, and the MSE on the test set is 56.78. It is 0.969.

[0088] See also Figure 6, which shows a structural block diagram of a multi-disaster impact analysis system for renewable energy power generation under extreme weather conditions of the present application.

[0089] like Figure 6 As shown, the multi-disaster impact analysis system 200 for renewable energy power generation under extreme weather conditions includes an acquisition module 210 , a screening module 220 , a construction module 230 , an optimization module 240 and an output module 250 .

[0090] Among them, the acquisition module 210 is configured to acquire meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors and the renewable energy power generation data includes renewable energy power generation; the screening module 220 is configured to determine the linear correlation between the meteorological factors and the renewable energy power generation according to the Pearson correlation coefficient, screen out meteorological factors with an absolute value of the linear correlation greater than a preset threshold, and define them as target meteorological factors; the construction module 230 is configured to construct a random forest regression model based on the meteorological data including the target meteorological factors and the normalized renewable energy power generation data; the optimization module 240 is configured to introduce an improved Bayesian optimization method, use the determination coefficient of the random forest regression model on the validation set as the objective function, and optimize the hyperparameters of the random forest regression model, the hyperparameters including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model; the output module 250 is configured to input real-time meteorological data into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the renewable energy power generation corresponding to the real-time meteorological data.

[0091] It should be understood that Figure 6 Modules and references documented in Figure 1 Therefore, the operations and features described above for the method and the corresponding technical effects also apply to Figure 6 The modules in it will not be described in detail here.

[0092] In other embodiments, embodiments of the present invention further provide a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor is caused to execute the method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions in any of the above method embodiments;

[0093] As an embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, and the computer-executable instructions are configured as follows:

[0094] Acquiring meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation;

[0095] Determine the linear correlation between the meteorological factor and the renewable energy power generation according to the Pearson correlation coefficient, select the meteorological factor with an absolute value of the linear correlation greater than a preset threshold, and define it as the target meteorological factor;

[0096] Constructing a random forest regression model based on meteorological data including the target meteorological factor and normalized new energy power generation data;

[0097] An improved Bayesian optimization method was introduced, with the coefficient of determination of the random forest regression model on the validation set as the objective function. The hyperparameters of the random forest regression model were optimized, including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model.

[0098] The real-time meteorological data is input into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation corresponding to the real-time meteorological data.

[0099] The computer-readable storage medium may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the system for analyzing the impact of multiple hazards on renewable energy power generation under extreme weather conditions, etc. Furthermore, the computer-readable storage medium may include high-speed random access memory and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some embodiments, the computer-readable storage medium may optionally include memory remotely located relative to the processor, and these remote memories may be connected to the system for analyzing the impact of multiple hazards on renewable energy power generation under extreme weather conditions via a network. Examples of the aforementioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0100] Figure 7 Schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 7 As shown, the device includes: a processor 310 and a memory 320. The electronic device may also include: an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330 and the output device 340 may be connected via a bus or other means. Figure 7The example of the bus connection is taken. The memory 320 is the computer-readable storage medium mentioned above. The processor 310 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 320, that is, implements the method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions in the above-mentioned method embodiment. The input device 330 can receive input digital or character information, and generate key signal input related to user settings and function control of the system for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions. The output device 340 may include a display device such as a display screen.

[0101] The electronic device can execute the method provided by the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided by the embodiment of the present invention.

[0102] As an embodiment, the electronic device is applied to a multi-hazard impact analysis system for renewable energy power generation under extreme weather conditions, and is used for a client, comprising: at least one processor; and a memory communicatively 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 to enable the at least one processor to:

[0103] Acquiring meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation;

[0104] Determine the linear correlation between the meteorological factor and the renewable energy power generation according to the Pearson correlation coefficient, select the meteorological factor with an absolute value of the linear correlation greater than a preset threshold, and define it as the target meteorological factor;

[0105] Constructing a random forest regression model based on meteorological data including the target meteorological factor and normalized new energy power generation data;

[0106] An improved Bayesian optimization method was introduced, with the coefficient of determination of the random forest regression model on the validation set as the objective function. The hyperparameters of the random forest regression model were optimized, including the number of decision trees, the maximum depth, and the minimum number of split samples, to obtain a disaster superposition impact rate model.

[0107] The real-time meteorological data is input into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation corresponding to the real-time meteorological data.

[0108] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods of each embodiment or certain portions of the embodiments.

[0109] 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions, characterized in that: include: Acquiring meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation; Determine the linear correlation between the meteorological factor and the renewable energy power generation according to the Pearson correlation coefficient, select the meteorological factor with an absolute value of the linear correlation greater than a preset threshold, and define it as the target meteorological factor; Constructing a random forest regression model based on meteorological data including the target meteorological factor and normalized new energy power generation data; An improved Bayesian optimization method is introduced, and the determination coefficient of the random forest regression model on the validation set is used as the objective function to optimize the hyperparameters of the random forest regression model. The hyperparameters include the number of decision trees, the maximum depth, and the minimum number of split samples, and a disaster superposition impact rate model is obtained. The improved Bayesian optimization method is introduced, and the determination coefficient of the random forest regression model on the validation set is used as the objective function to optimize the hyperparameters of the random forest regression model. The hyperparameters include the number of decision trees, the maximum depth, and the minimum number of split samples, and the disaster superposition impact rate model includes: The coefficient of determination of random forest on the validation set The maximum is the optimization goal, and the objective function is constructed; From the hyperparameter space Randomly sample hyperparameter combinations in , respectively evaluate the corresponding function values , construct the initial data set, the expression is: ， Where, is the initial data set, is the initial sampling point number, For the Sampling hyperparameter combinations, for The response value on the evaluation function; Leveraging Datasets , establish a Gaussian process regression model as an approximation of the objective function, the expression is: ， Where, For the objective function in the hyperparameter combination The response value under For the hyperparameter combination The mean and variance of the lower Gaussian process prediction, For the hyperparameter combination The average value under For the hyperparameter combination Variance under ; The expected improvement function is selected as the acquisition function to measure the value of sampling new points under the current proxy model; Based on the improved acquisition function, the current optimal test point is selected, and the expression is: ， Where, To represent the optimal test point currently selected, even if the hyperparameter combination with the largest acquisition function value is To make the function Get the maximum value of the parameter , is the value of the acquisition function that is expected to be improved, which is used to guide the selection of the next sampling point; The sampled Applied to random forest regression model training and performance evaluation to obtain the true objective function value , and update the dataset, the expression is: ， Where, To represent the hyperparameter combination The true evaluation value of the objective function; Repeat the modeling and sampling process, iteratively update the proxy model and data set, and finally obtain the optimal hyperparameter combination. Based on the optimal hyperparameter combination, a disaster superposition impact rate model is constructed. The expression of the optimal hyperparameter combination is: ， Where, To express the function Reaching the minimum Parameters, To express the optimal hyperparameter combination, even if the objective function Get the minimum value of the parameter; The expression of the objective function is: ， Where, is the objective function value, For the The actual power generation value at the moment, for The model prediction value at time t, is the sample weight, For each weight, the parameters can be optimized by Bayesian optimization; It is a physical factor. For wind power generation, it is wind speed. For photovoltaic power generation, it is radiation. is an indicator function, which takes the value 1 when the condition is met and takes the value 0 when it is not met; The expression of the acquisition function is: ， ， Where, is the optimal target at the current observation point, is the exploration function, which is used to control the trade-off between exploration and utilization. is the cumulative distribution function of the standard normal distribution, For the Gaussian process regression model at point The standard deviation of the prediction at is the probability density function of the standard normal distribution, is a standardized variable, measuring the The improvement in the model predictions at is the uncertainty adjustment factor, which can be selected through parameter optimization. For physical factors, wind speed is selected for wind power generation and radiation is selected for photovoltaic power generation; is an indicator function, which takes the value 1 when the condition is met and takes the value 0 when it is not met. is the predicted mean; The real-time meteorological data is input into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation corresponding to the real-time meteorological data.

2. The method for analyzing the impact of multiple disasters on renewable energy power generation under extreme weather conditions according to claim 1, characterized in that: The step of constructing a random forest regression model based on the meteorological data including the target meteorological factor and the normalized new energy power generation data includes: Sample selection: Multiple subsamples are randomly selected from meteorological data and renewable energy power generation data. Duplicates are allowed in multiple subsamples, that is, a data point appears multiple times in the same subsample. Constructing a decision tree: For each randomly sampled subsample, a decision tree is independently constructed. When constructing each node of the decision tree, instead of selecting the best feature from all features, a portion of features is randomly selected, and then the best feature is selected from them for splitting. Decision tree splitting: Use selected features to split each decision tree node. Common splitting criteria include information gain or Gini impurity. The expression of Gini impurity is: ， Where, is the Gini impurity, For the first The proportion of classes, is the total number of categories; The expression of information gain is: ， Where, For a split dataset, For the split attribute, For attributes All possible values ​​of For the current attribute Pick hour A subset of For the dataset Entropy before classification, Indicates that the attribute Pick hour The entropy of a subset of Indicates that the attribute The above dataset The information gain obtained by performing the partitioning; Result integration: For regression problems, the average of the predicted values ​​is calculated. For regression tasks, the average formula is as follows: ， Where, is the average of the predicted values, For the A decision tree for the sample The predicted value of is the total number of decision trees; Combine the regression decision trees to get the corresponding random forest model, where each tree has a Give independent prediction values .

3. A multi-disaster impact analysis system for renewable energy power generation under extreme weather conditions, characterized by: include: an acquisition module configured to acquire meteorological data and renewable energy power generation data, wherein the meteorological data includes meteorological factors, and the renewable energy power generation data includes renewable energy power generation; a screening module configured to determine a linear correlation between the meteorological factor and the renewable energy power generation according to a Pearson correlation coefficient, screen out meteorological factors whose absolute value of the linear correlation is greater than a preset threshold, and define them as target meteorological factors; A construction module is configured to construct a random forest regression model based on the meteorological data including the target meteorological factor and the normalized new energy power generation data; The optimization module is configured to introduce an improved Bayesian optimization method, use the determination coefficient of the random forest regression model on the validation set as the objective function, optimize the hyperparameters of the random forest regression model, and obtain a disaster superposition impact rate model. The improved Bayesian optimization method is introduced, use the determination coefficient of the random forest regression model on the validation set as the objective function, optimize the hyperparameters of the random forest regression model, and obtain a disaster superposition impact rate model including the number of decision trees, the maximum depth, and the minimum number of split samples. The coefficient of determination of random forest on the validation set The maximum is the optimization goal, and the objective function is constructed; From the hyperparameter space Randomly sample hyperparameter combinations in , respectively evaluate the corresponding function values , construct the initial data set, the expression is: ， Where, is the initial data set, is the number of initial sampling points, For the sampled hyperparameter combinations, for The response value on the evaluation function; Leveraging Datasets , establish a Gaussian process regression model as an approximation of the objective function, the expression is: ， Where, For the objective function in the hyperparameter combination The response value under For the hyperparameter combination The mean and variance of the lower Gaussian process prediction, For the hyperparameter combination The average value under For the hyperparameter combination Variance under ; The expected improvement function is selected as the acquisition function to measure the value of sampling new points under the current proxy model; Based on the improved acquisition function, the current optimal test point is selected, and the expression is: ， Where, To represent the optimal test point currently selected, even if the hyperparameter combination with the largest acquisition function value is To make the function Get the maximum value of the parameter , is the value of the acquisition function that is expected to be improved, which is used to guide the selection of the next sampling point; The sampled Applied to random forest regression model training and performance evaluation to obtain the true objective function value , and update the dataset, the expression is: ， Where, To represent the hyperparameter combination The true evaluation value of the objective function; Repeat the modeling and sampling process, iteratively update the proxy model and data set, and finally obtain the optimal hyperparameter combination. Based on the optimal hyperparameter combination, a disaster superposition impact rate model is constructed. The expression of the optimal hyperparameter combination is: ， Where, To express the function Reaching the minimum Parameters, To express the optimal hyperparameter combination, even if the objective function Get the minimum value of the parameter; The expression of the objective function is: ， Where, is the objective function value, For the The actual power generation value at the moment, for The model prediction value at time t, is the sample weight, For each weight, the parameters can be optimized by Bayesian optimization; It is a physical factor. For wind power generation, it is wind speed. For photovoltaic power generation, it is radiation. is an indicator function, which takes the value 1 when the condition is met and takes the value 0 when it is not met; The expression of the acquisition function is: ， ， Where, is the optimal target at the current observation point, is the exploration function, which is used to control the trade-off between exploration and utilization. is the cumulative distribution function of the standard normal distribution, For the Gaussian process regression model at point The standard deviation of the prediction at is the probability density function of the standard normal distribution, is a standardized variable, measuring the The improvement in the model predictions at is the uncertainty adjustment factor, which can be selected through parameter optimization. For physical factors, wind speed is selected for wind power generation and radiation is selected for photovoltaic power generation; is an indicator function, which takes the value 1 when the condition is met and takes the value 0 when it is not met. is the predicted mean; The output module is configured to input real-time meteorological data into the disaster superposition impact rate model, and the disaster superposition impact rate model outputs the new energy power generation power corresponding to the real-time meteorological data.

4. An electronic device, characterized in that: include: At least one processor, and a memory communicatively 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 to enable the at least one processor to perform the method according to any one of claims 1 to 2.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 2 is implemented.

Images