Intra-day wind power extreme scene simulation method, system and equipment based on extended sampling interval and improved sampling matrix and medium

Abstract

The invention discloses an intraday wind power extreme scene simulation method, system and device based on an extended sampling interval and an improved sampling matrix, and a medium, and belongs to the technical field of scene simulation, and the method comprises the steps: constructing a cumulative distribution function of wind power output based on historical wind power output data; a bilateral quantile is set, reverse operation is executed based on a cumulative distribution function of wind power output, and an extreme value sampling interval is obtained through calculation; performing normalized mapping on the sampling probability, and constructing an extreme value cumulative distribution function corresponding to the extreme value sampling interval; characteristic value fluctuation of historical wind power data is extracted, and an improved Latin hypercube sampling probability matrix is generated; performing inverse sampling to obtain an intra-day wind power output scene set containing a plurality of scenes; and screening the wind power output scene set based on the extreme value index, extracting scenes meeting extreme conditions, and forming a wind power intraday extreme scene set. According to the method, the problem of insufficient extreme scene simulation of a traditional method is solved, and the generated scene set is richer in an extreme region.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of scenario simulation, and in particular to a method, system, device and medium for simulating extreme scenarios of intraday wind power based on an extended sampling interval and an improved sampling matrix. BACKGROUND

[0002] Currently, the mainstream method for simulating wind power output scenarios mainly includes two steps of probability modeling-scenario sampling and scenario reduction. In terms of probability modeling-scenario sampling, traditional methods include time series models and their improved models, such as autoregressive conditional heteroscedasticity (ARCH) models, and scenario sampling methods such as Monte Carlo sampling (MCS) and Latin hypercube sampling (LHS). In terms of scenario reduction, technologies such as synchronous backward reduction and clustering are generally used to filter representative scenarios from a large-scale scenario set according to the similarity between scenarios. SUMMARY

[0003] In view of the above problems, the present application provides a method, system, device and medium for simulating extreme scenarios of intraday wind power based on an extended sampling interval and an improved sampling matrix.

[0004] Therefore, the technical problem solved by the present application is how to solve the problem that the existing synchronous backward reduction and clustering methods mainly rely on similarity criteria and lack specialized analysis from extreme characteristics themselves, resulting in their inability to accurately identify and extract scenarios most representative of extremes.

[0005] To solve the above technical problems, the present application provides the following technical solutions: a method for simulating extreme scenarios of intraday wind power based on an extended sampling interval and an improved sampling matrix, which includes: based on historical wind power output data, using a kernel density estimation method to construct a cumulative distribution function of wind power output; setting a double-sided quantile, performing an inverse operation based on the cumulative distribution function of wind power output to calculate an extreme sampling interval; normalizing and mapping the sampling probability to construct an extreme cumulative distribution function corresponding to the extreme sampling interval; extracting the eigenvalue fluctuation of historical wind power data to generate an improved Latin hypercube sampling probability matrix for simulating time continuity; using the extreme cumulative distribution function and the improved Latin hypercube sampling probability matrix to perform inverse sampling to obtain a set of intraday wind power output scenarios containing multiple scenarios; based on an extreme index, filtering the set of wind power output scenarios to extract scenarios that meet extreme conditions to form a set of intraday extreme scenarios of wind power.

[0006] As a preferred embodiment of the intraday extreme wind power scenario simulation method based on extended sampling interval and improved sampling matrix described in this invention, the step of constructing the cumulative distribution function of wind power output based on historical wind power output data using kernel density estimation includes: constructing a probability model based on historical wind power output data; performing distribution fitting on the probability model using kernel density estimation; estimating the probability density of wind power output data by selecting a kernel function; and obtaining the cumulative distribution function of wind power output based on the probability density calculation results.

[0007] As a preferred embodiment of the intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix described in this invention, the step of setting bilateral quantiles and performing a reverse operation based on the cumulative distribution function of wind power output to calculate the extreme value sampling interval includes: setting upper and lower bilateral quantiles to define the extreme output; performing a reverse operation using the cumulative distribution function of wind power output to calculate the upper and lower quantiles; and determining the extreme value sampling interval using the upper and lower quantiles as boundaries.

[0008] As a preferred embodiment of the intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix described in this invention, the step of normalizing the sampling probability and constructing the extreme value cumulative distribution function corresponding to the extreme value sampling interval includes: determining the original sample value based on the cumulative distribution function value corresponding to the sampling probability; normalizing the original sample value and converting it into the sampling result within the extreme value sampling interval; and sequentially associating the normalized sampling result with the sampling probability to construct the extreme value cumulative distribution function.

[0009] This preferred scheme effectively maintains the volatility and temporal continuity of historical data in the improved Latin hypercube sampling probability matrix, making the generated wind power output scenario more closely resemble the actual wind power operation behavior in terms of dynamic characteristics.

[0010] As a preferred embodiment of the intraday extreme wind power scenario simulation method based on extended sampling interval and improved sampling matrix described in this invention, the step of extracting the eigenvalue fluctuations of historical wind power data and generating an improved Latin hypercube sampling probability matrix for simulating time continuity includes: normalizing historical wind power output data and mapping it to a sampling counting interval; calculating the periodic fluctuations of the normalized wind power output data; calculating the standard deviation of the periodic fluctuations to obtain the degree of fluctuation in the simulated data; generating a random number matrix based on a normal distribution and setting the standard deviation as the degree of fluctuation in the simulated data; restricting random values ​​exceeding the maximum fluctuation threshold, and constructing the improved Latin hypercube sampling probability matrix.

[0011] This preferred scheme improves the sampling probability input of the Latin hypercube sampling probability matrix into the extreme cumulative distribution function and performs inverse function operation, so that the generated wind power output value strictly corresponds to the extreme sampling interval, effectively realizing the guiding construction of extreme wind power output scenarios. Combined with the time-series arrangement operation, the scenario set not only has extreme characteristics, but also has time series characteristics that can be used for intraday scheduling simulation.

[0012] As a preferred embodiment of the intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix described in this invention, the step of using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix for inverse sampling to obtain an intraday wind power output scenario set containing multiple scenarios includes: inputting each sampling probability in the improved Latin hypercube sampling probability matrix into the corresponding extreme value cumulative distribution function; performing an inverse function operation on the sampling probability to obtain the wind power output value; and arranging all the wind power output values ​​in chronological order to generate multiple intraday wind power output scenarios with temporal continuity.

[0013] This preferred solution filters the wind power output scenario set based on the extreme value index, targeting daily total output, instantaneous output value, continuous time period output value, and output ratio of a specified time period. This allows for a comprehensive measurement of the extreme nature of wind power output from multiple dimensions, identifying representative extreme scenarios under different conditions and improving the applicability and representativeness of the extreme scenario set in dispatch safety assessment and flexibility assessment.

[0014] As a preferred embodiment of the intraday extreme scenario simulation method for wind power based on extended sampling interval and improved sampling matrix described in this invention, the step of filtering the wind power output scenario set based on the extreme value index to extract scenarios that meet the extreme conditions and form an intraday extreme scenario set for wind power includes: calculating and filtering scenarios based on the extreme value index of the wind power output scenario set based on the total daily output; calculating and filtering scenarios based on the extreme value of the wind power output scenario set based on the instantaneous output value; calculating and filtering scenarios based on the extreme value of the wind power output scenario set based on the output value within a continuous time period; and calculating and filtering scenarios based on the output ratio within a specified time period.

[0015] This preferred solution filters the wind power output scenario set based on the extreme value index, targeting daily total output, instantaneous output value, continuous time period output value, and output ratio of a specified time period. This allows for a comprehensive measurement of the extreme nature of wind power output from multiple dimensions, identifying representative extreme scenarios under different conditions and improving the applicability and representativeness of the extreme scenario set in dispatch safety assessment and flexibility assessment.

[0016] This invention provides an intraday wind power extreme scenario simulation system based on extended sampling interval and improved sampling matrix.

[0017] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a system for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices, comprising: a probability distribution construction module, an extreme value sampling interval calculation module, a normalization mapping module, an improved sampling matrix generation module, an inverse sampling execution module, and an extreme value scenario screening module; the probability distribution construction module is used to construct the cumulative distribution function of wind power output based on historical wind power output data using a kernel density estimation method; the extreme value sampling interval calculation module is used to set bilateral quantiles, perform inverse operations based on the cumulative distribution function of wind power output, and calculate the extreme value sampling interval; the normalization mapping module is used to set bilateral quantiles, perform inverse operations based on the cumulative distribution function of wind power output, and calculate the extreme value sampling interval; the normalization mapping module is used to construct the cumulative distribution function of wind power output based on extended sampling intervals and improved sampling matrix, and perform inverse operations based on extended sampling intervals and improved sampling matrix, and perform inverse sampling execution module; the system for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrix, and performs ... The normalization mapping module is used to normalize the sampling probability and construct the extreme value cumulative distribution function corresponding to the extreme value sampling interval; the improved sampling matrix generation module is used to extract the feature value fluctuations of historical wind power data and generate an improved Latin hypercube sampling probability matrix for simulating time continuity; the inverse sampling execution module is used to perform inverse sampling using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios; the extreme value scenario filtering module is used to filter the wind power output scenario set based on the extreme value index, extract scenarios that meet the extreme conditions, and form a set of intraday extreme wind power scenarios.

[0018] The present invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the described method for simulating intraday extreme wind power scenarios based on extended sampling interval and improved sampling matrix.

[0019] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the described method for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices.

[0020] The beneficial effects of this invention are: This invention accurately focuses on extreme events and actively allocates sampling resources to extreme value intervals with lower probabilities by extending the sampling interval technology, which solves the problem of insufficient simulation of extreme scenarios by traditional methods and generates a richer set of scenarios in extreme regions.

[0021] By improving the sampling matrix technology, the temporal characteristics such as fluctuation amplitude and continuity in historical data are incorporated into the sampling process. The generated extreme scenarios are not only extreme in numerical terms, but also conform to the fluctuation patterns of real wind power in terms of temporal evolution, thus avoiding physically unrealistic abrupt changes.

[0022] A multi-dimensional extreme evaluation index system was established to quantify extremes from multiple perspectives, such as single-point extremes and continuous extremes, so that the scenario reduction and extraction process is based on evidence and the results are more representative.

[0023] It inherits the high sampling efficiency of LHS and, through targeted improvements, can obtain high-quality extreme scenario sets with relatively few samplings, which is superior to Monte Carlo simulations that require a large number of samples. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 The following is a general flowchart of an intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix, provided as an embodiment of the present invention. Detailed Implementation

[0026] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0027] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices, including: S1. Based on historical wind power output data, the cumulative distribution function of wind power output is constructed using the kernel density estimation method.

[0028] S2. Set the bilateral quantiles, perform the reverse operation based on the cumulative distribution function of wind power output, and calculate the extreme value sampling interval.

[0029] S3. Normalize the sampling probability and construct the cumulative distribution function of the extreme value corresponding to the extreme value sampling interval.

[0030] S4. Extract the eigenvalue fluctuations from historical wind power data to generate an improved Latin hypercube sampling probability matrix for simulating time continuity.

[0031] S5. Inverse sampling is performed using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios.

[0032] S6. Based on the extreme value index, the wind power output scenario set is screened, and the scenarios that meet the extreme conditions are extracted to form the wind power intraday extreme scenario set.

[0033] It should be noted that wind power output is highly uncertain and volatile, and extreme scenarios have a significant impact on the safe operation of the system in actual power system dispatch. However, due to the low frequency of extreme values, traditional random sampling is often insufficient to cover critical operating conditions.

[0034] This embodiment constructs a full-process mechanism from probabilistic modeling, extreme value guidance, continuous time simulation to scenario selection through steps S100 to S600. It can not only accurately simulate the changing behavior of wind power output under extreme conditions, but also effectively improve the support capability for extreme operating conditions in wind power dispatch analysis, flexibility assessment and operation safety prediction.

[0035] Example 2, an embodiment of the present invention, provides a method for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices, based on the previous embodiment, including: In this embodiment, the cumulative distribution function of wind power output can be calculated by using the kernel density estimation method to fit the distribution of wind power data based on historical wind power output data.

[0036] In an alternative implementation, the cumulative distribution function of wind power output can also be based on a cumulative distribution function constructed using histogram density estimation or other non-parametric probability estimation methods, which can also be used to characterize and support the sampling of wind power probability distribution.

[0037] In another alternative implementation, the cumulative distribution function of wind power output can also be constructed by statistically fitting historical wind power data and using a parametric distribution model, and can also be used for scenario simulation.

[0038] This invention constructs a cumulative distribution function of wind power output based on the kernel density estimation method, which can accurately capture the non-parametric fluctuation characteristics of wind power output. Compared with the traditional parametric model, it has better fitting ability at the boundary and can more realistically reflect the probability structure of wind power.

[0039] Furthermore, in step S1, based on historical wind power output data, a cumulative distribution function of wind power output is constructed using the kernel density estimation method, including the following steps A1-A4: A1. Construct a probabilistic model based on historical wind power output data.

[0040] A2. The kernel density estimation method is used to fit the distribution of the probability model.

[0041] A3. Select a kernel function to estimate the probability density of wind power output data.

[0042] A4. Based on the probability density calculation results, obtain the cumulative distribution function of wind power output.

[0043] In the embodiments of this application, the probability model can be a nonparametric probability model constructed based on historical wind power output data. The kernel density estimation method is used to fit the wind power output sequence to form a continuous probability density function, and then the cumulative distribution function is calculated to support subsequent extreme value sampling.

[0044] In an alternative implementation, the probability model can also be used to approximate the distribution of wind power output by constructing histograms or distribution fitting plots and employing statistical methods such as sliding windows to form a discrete probability model for estimating local probability distribution characteristics.

[0045] In another alternative implementation, the probability model can also employ statistical methods based on parameter assumptions to perform rapid cumulative distribution function derivation and sampling.

[0046] This invention constructs a probabilistic model based on kernel density estimation, which makes the modeling process independent of specific distribution assumptions and can more accurately describe the real fluctuation pattern of wind power output. It has a higher fitting ability, especially in the tail and multi-peak regions, thus providing a more reliable probabilistic basis for extreme value sampling and improving the accuracy and stability of wind power scenario simulation.

[0047] Specifically, maintaining the true volatility of wind power is the most critical criterion for evaluating the effectiveness of probabilistic modeling, and probability distribution is a key step in scenario simulation probabilistic modeling. First, kernel density estimation (KDE) is used to fit the probability distribution of historical wind power output data to more accurately capture its true volatility characteristics. The cumulative distribution function estimate using the Gaussian kernel function is calculated as follows: in, It is the cumulative distribution function The estimated value; The wind power output value to be estimated; For the first A sample of historical wind power output data; For bandwidth; For sample size; The cumulative distribution function of the kernel function. The input value for the kernel function, that is ; is the kernel density function.

[0048] The kernel density function of this invention is a Gaussian kernel, expressed as follows: in, The input values ​​are for the Gaussian kernel. It is a real number.

[0049] Furthermore, in step S2, the bilateral quantiles are set, and the inverse operation is performed based on the cumulative distribution function of the wind power output to calculate the extreme value sampling interval, including the following steps B1-B3: B1. Set the upper and lower quantiles used to define extreme outputs.

[0050] B2. Calculate the upper and lower quantiles by performing a reverse operation using the cumulative distribution function of wind power output.

[0051] B3. Use the upper and lower quantiles as boundaries to determine the extreme value sampling interval.

[0052] In the embodiments of this application, the upper and lower quantiles can be obtained by setting a bilateral quantile threshold on the cumulative distribution function and performing a reverse operation to obtain the corresponding probability position value. The upper quantile corresponds to a higher probability position, reflecting the critical point of high wind power output, and the lower quantile corresponds to a lower probability position, reflecting the boundary point of low wind power output.

[0053] In an optional implementation, the upper and lower quantile points can also be determined by boundary sampling points based on the sample sorting position after discretizing the wind power output probability distribution, which can be used to delineate the sampling range and approximately reflect the extreme operating condition range.

[0054] In another alternative implementation, the upper and lower quantiles can also be theoretical values ​​directly calculated from the set quantile probability values ​​in the parameterized cumulative distribution function, which can be used to reflect the sampling positions of boundary extreme values.

[0055] This invention constructs the upper and lower boundaries of the wind power output sampling interval by calculating the upper and lower quantiles based on the cumulative distribution function, ensuring that the sampling interval is close to the tail region of the wind power output distribution, effectively improving the generation efficiency of extreme value samples, and making the scenario sampling results more representative and comprehensive.

[0056] Specifically, based on the traditional LHS sampling probability matrix, this invention actively expands the sampling interval through inverse operation and normalization mapping, focusing the sampling emphasis on the extreme value range.

[0057] Obtained by kernel density estimation cumulative distribution function Specify bilateral quantiles (Set the extreme value threshold as needed, for example) =1%), and on Perform the reverse operation, and the result is represented as , ,in As a lower quantile, it is a lower Quantiles The upper quantile is relatively high. Quantiles, intervals It is considered an extreme value range.

[0058] In this embodiment, the sampling probability can be a uniformly distributed probability value generated in the Latin hypercube sampling probability matrix, which is used as an input variable. After normalization mapping and inverse operation steps, key probability parameters for wind power output scenario simulation are constructed.

[0059] In an alternative implementation, the sampling probability can also be generated by a probability value generated by a Monte Carlo sampling method or a random uniform sampling method, and used to replace Latin hypercube sampling to complete the sampling input of the cumulative distribution function.

[0060] In another alternative implementation, the sampling probability can also be based on probability location points extracted from the empirical distribution sequence of historical actual wind power output data, which are used to generate sampling inputs that fit the historical distribution characteristics.

[0061] This invention generates sampling probabilities with uniform distribution characteristics by using Latin hypercube sampling, and performs inverse sampling by combining normalized mapping and extreme value cumulative distribution function, which effectively improves the coverage density of extreme value intervals, so that the generated wind power scenario has statistical rationality while ensuring the expression of extreme values, and enhances the goal orientation and structural integrity of the scenario simulation.

[0062] Furthermore, in step S3, the sampling probabilities are normalized and mapped to construct the extreme value cumulative distribution function corresponding to the extreme value sampling interval, including the following steps C1-C3: C1. Determine the original sampled value based on the cumulative distribution function value corresponding to the sampling probability.

[0063] C2. Normalize the original sampled values ​​and convert them into sampling results within the extreme value sampling interval.

[0064] C3. Associate the normalized and mapped sampling results with the sampling probabilities in sequence to construct the extreme value cumulative distribution function.

[0065] Specifically, according to the first The sampling probability obtained from the second sampling right Perform sampling and represent the results as ,right The extreme sampling intervals are represented using a normalized map, as shown below: in, For all The maximum probability value in, For all The minimum probability value in; This represents the mapping result. .

[0066] By With sampling probability The order is relevant, and the cumulative distribution function corresponding to the extreme value sampling interval is derived. .

[0067] Furthermore, in step S4, eigenvalue fluctuations of historical wind power data are extracted to generate an improved Latin hypercube sampling probability matrix for simulating time continuity, including the following steps D1-D5: D1. Normalize the historical wind power output data and map it to the sampling counting interval.

[0068] D2. Calculate the periodic fluctuation of the normalized wind power output data.

[0069] D3. Calculate the standard deviation of the fluctuation during the period to obtain the degree of fluctuation of the simulated data.

[0070] D4. Generate a random number matrix based on the normal distribution and set the standard deviation to simulate the fluctuation of the data.

[0071] D5. Limit random values ​​that exceed the maximum fluctuation threshold and construct an improved Latin hypercube sampling probability matrix.

[0072] Specifically, to ensure the generated scenarios exhibit realistic temporal continuity, this invention improves the LHS sampling probability matrix. In probabilistic modeling for wind power fluctuation simulation, manually set parameters often differ from historical data. This invention extracts eigenvalue fluctuations from historical wind power data to enhance the LHS sampling probability matrix, simulating the temporal continuity and fluctuation characteristics of historical wind farm data through the following steps.

[0073] Historical data is mapped to sampling count intervals through normalization. As shown below: in, , The number of historical wind power output scenarios; This is non-historical wind power generation output data; For time series indexing; for Mapped to sampling count interval As a result, This represents the number of samples.

[0074] Calculate the fluctuation during the week As a basis for determining the fluctuation range of the simulated data, it is shown below: in, For the fluctuations during the week, , This represents the total time.

[0075] calculate Standard deviation , which represents the degree of data fluctuation in a random simulation.

[0076] Pick The absolute value of is expressed as Kernel density estimation is applied to obtain Then, its specified percentile (e.g., the 99th percentile) is calculated to derive the fluctuation threshold. To simulate time continuity.

[0077] Create a value with a mean of 0 and a standard deviation of 0. Normally distributed random number matrix , where the matrix The size is .

[0078] random number matrix Greater than Replace the value in with , will be less than Replace the value with .

[0079] By sampling the probability matrix of LHS (matrix Size is The columns of the array are randomly swapped to simulate resource fluctuation characteristics, as shown below: in, express of OK The sampling probability of the column, Represents a random matrix middle OK The data in the column, This represents the improved sampling matrix, with row transformations restricted to the interval [missing information]. .

[0080] Furthermore, in step S5, inverse sampling is performed using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios, including the following steps E1-E3: E1, improve the cumulative distribution function of the extreme values ​​corresponding to each sampling probability input in the Latin hypercube sampling probability matrix.

[0081] E2. Perform an inverse function operation on the sampling probability to obtain the wind power output value.

[0082] E3. Arrange all wind power output values ​​in chronological order to generate multiple intraday wind power output scenarios with temporal continuity.

[0083] Specifically, the extreme value index is used to extract extreme value scenarios. This invention first performs sampling calculations based on the wind power cumulative distribution function and the sampling probability matrix. The inverse operation formula is as follows: in, It is the inverse function of the cumulative distribution function. To be related to sampling probability The corresponding sampled values.

[0084] Use the extreme value cumulative distribution function constructed in step two. The improved sampling matrix obtained in step three is then subjected to inverse sampling operations to finally generate a sample matrix containing... The set of extreme scenarios for intraday wind power output is represented as follows: .

[0085] Furthermore, in step S6, the wind power output scenario set is screened based on the extreme value index to extract scenarios that meet the extreme conditions, forming a wind power intraday extreme scenario set, including the following steps F1-F4: F1. Calculate the extreme value index of the wind power output scenario set based on the total daily output and screen the scenarios.

[0086] F2. Calculate the extreme value index of the wind power output scenario set based on the instantaneous output value and filter the scenarios.

[0087] F3. Calculate the extreme value index of the wind power output scenario set based on the output values ​​within a continuous time period and then filter the scenarios.

[0088] F4. Calculate the extreme value index of the wind power output scenario set based on the output ratio within a specified time period and filter the scenarios.

[0089] Specifically, extreme daily output: a scenario where total daily output exceeds the 95th percentile of historical data.

[0090] Extreme output values: There are situations where the instantaneous output value exceeds the 95th percentile of historical data.

[0091] Extreme continuous output for 1 hour: There are scenarios where the output exceeds the 95th percentile of the corresponding value in the historical data for 1 hour.

[0092] Extreme 6-hour continuous output ratio: The scenario where the output for 6 consecutive hours accounts for more than 34% of the total output for the day.

[0093] Example 3 is an embodiment of the present invention. This embodiment provides an intraday wind power extreme scenario simulation system based on extended sampling interval and improved sampling matrix, including a probability distribution construction module, an extreme value sampling interval calculation module, a normalization mapping module, an improved sampling matrix generation module, an inverse sampling execution module, and an extreme value scenario screening module.

[0094] The probability distribution construction module is used to construct the cumulative distribution function of wind power output based on historical wind power output data using the kernel density estimation method.

[0095] The extreme value sampling interval calculation module is used to set the bilateral quantiles, perform the inverse operation based on the cumulative distribution function of wind power output, and calculate the extreme value sampling interval.

[0096] The normalization mapping module is used to normalize the sampling probabilities and construct the extreme value cumulative distribution function corresponding to the extreme value sampling interval.

[0097] An improved sampling matrix generation module is used to extract eigenvalue fluctuations from historical wind power data and generate an improved Latin hypercube sampling probability matrix for simulating time continuity.

[0098] The inverse sampling execution module is used to perform inverse sampling using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios.

[0099] The extreme scenario screening module is used to screen the wind power output scenario set based on the extreme value index, extract the scenarios that meet the extreme conditions, and form the wind power intraday extreme scenario set.

[0100] This embodiment also provides an electronic device applicable to a method for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices as proposed in the above embodiment.

[0101] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a method for simulating intraday extreme wind power scenarios based on an extended sampling interval and an improved sampling matrix, as proposed in the above embodiment.

[0102] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for simulating intraday extreme wind power scenarios based on extended sampling interval and improved sampling matrix proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0103] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part 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 a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0104] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for simulating intraday extreme wind power scenarios based on extended sampling interval and improved sampling matrix, characterized in that: include, Based on historical wind power output data, a cumulative distribution function of wind power output is constructed using the kernel density estimation method; By setting the bilateral quantiles and performing the inverse operation based on the cumulative distribution function of wind power output, the extreme value sampling interval is calculated. Normalize the sampling probabilities and construct the cumulative distribution function of extreme values ​​corresponding to the extreme value sampling interval; Extract eigenvalue fluctuations from historical wind power data to generate an improved Latin hypercube sampling probability matrix for simulating temporal continuity; Inverse sampling was performed using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios; The extreme value index is used to screen the wind power output scenario set and extract the scenarios that meet the extreme conditions to form the intraday extreme scenario set for wind power.

2. The intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 1, characterized in that: The method of constructing the cumulative distribution function of wind power output based on historical wind power output data using kernel density estimation includes: A probability model was constructed based on historical wind power output data. The kernel density estimation method is used to fit the distribution of the probability model; A kernel function is selected to estimate the probability density of wind power output data; Based on the probability density calculation results, the cumulative distribution function of wind power output is obtained.

3. The intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 2, characterized in that: The set bilateral quantiles, based on the cumulative distribution function of wind power output, perform a reverse operation to calculate the extreme value sampling interval, including: Set the upper and lower quantiles used to define extreme outputs; The upper and lower quantiles are calculated by performing a reverse operation using the cumulative distribution function of wind power output; The extreme value sampling interval is determined by using the upper and lower quantiles as boundaries.

4. The intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 3, characterized in that: The normalization mapping of sampling probabilities to construct the cumulative distribution function of extreme values ​​corresponding to the extreme value sampling interval includes, The original sampled value is determined based on the cumulative distribution function value corresponding to the sampling probability; The original sampled values ​​are normalized and mapped to convert them into sampling results within the extreme value sampling interval; The normalized sampling results are sequentially associated with the sampling probabilities to construct the extreme value cumulative distribution function.

5. The intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 4, characterized in that: The extraction of eigenvalue fluctuations from historical wind power data generates an improved Latin hypercube sampling probability matrix for simulating temporal continuity, including: The historical wind power output data is normalized and mapped to the sampling counting interval; Calculate the periodic fluctuation of the normalized wind power output data; The standard deviation of the fluctuations during the period is calculated to obtain the degree of fluctuation in the simulated data; A random number matrix is ​​generated based on a normal distribution, and the standard deviation is set to simulate the degree of data fluctuation. An improved Latin hypercube sampling probability matrix is ​​constructed by restricting random values ​​that exceed the maximum fluctuation threshold.

6. The intraday extreme wind power scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 5, characterized in that: The method of inverse sampling using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix yields a set of intraday wind power output scenarios containing multiple scenarios, including: The extreme value cumulative distribution function corresponding to each sampling probability input in the Latin hypercube sampling probability matrix will be improved. Perform an inverse function operation on the sampling probability to obtain the wind power output value; Arrange all wind power output values ​​in chronological order to generate multiple intraday wind power output scenarios with temporal continuity.

7. The intraday wind power extreme scenario simulation method based on extended sampling interval and improved sampling matrix as described in claim 6, characterized in that: The extreme value index is used to filter the wind power output scenario set, extracting scenarios that meet extreme conditions to form a daily extreme scenario set for wind power, including: Extreme value index calculation and scenario selection are performed on the wind power output scenario set based on total daily output. Extreme value index is calculated and scenarios are selected based on instantaneous output values ​​for the wind power output scenario set; The extreme value index of the wind power output scenario set is calculated and the scenario is screened based on the output value within a continuous time period. The extreme value index of the wind power output scenario set is calculated and the scenarios are selected based on the output ratio within a specified time period.

8. A system for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices, employing the intraday extreme wind power scenario simulation method based on extended sampling intervals and improved sampling matrices as described in any one of claims 1 to 7, characterized in that, include: The module includes a probability distribution construction module, an extreme value sampling interval calculation module, a normalization mapping module, an improved sampling matrix generation module, an inverse sampling execution module, and an extreme value scenario screening module. The probability distribution construction module is used to construct the cumulative distribution function of wind power output based on historical wind power output data and using the kernel density estimation method. The extreme value sampling interval calculation module is used to set the bilateral quantiles, perform the reverse operation based on the cumulative distribution function of wind power output, and calculate the extreme value sampling interval. The normalization mapping module is used to normalize the sampling probability and construct the extreme value cumulative distribution function corresponding to the extreme value sampling interval. The improved sampling matrix generation module is used to extract the eigenvalue fluctuations of historical wind power data and generate an improved Latin hypercube sampling probability matrix for simulating time continuity. The inverse sampling execution module is used to perform inverse sampling using the extreme value cumulative distribution function and the improved Latin hypercube sampling probability matrix to obtain a set of intraday wind power output scenarios containing multiple scenarios. The extreme scenario screening module is used to screen the wind power output scenario set based on the extreme value index, extract the scenarios that meet the extreme conditions, and form the wind power intraday extreme scenario set.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of any one of claims 1 to 7 for a method of simulating intraday extreme wind power scenarios based on extended sampling interval and improved sampling matrix.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of any one of claims 1 to 7 for simulating intraday extreme wind power scenarios based on extended sampling intervals and improved sampling matrices.

Images