Wind-solar combined output scene analysis method for power system

Abstract

The invention provides a wind-solar combined output scene analysis method for a power system. The method comprises the following steps: acquiring wind and light data samples; dividing the data samples according to time periods, performing nonparametric kernel density estimation on the wind and light data samples of each time period, and determining a probability density function and a probability distribution function of wind and light output of each time period; on the basis of the probability distribution function of the wind and light output in each time period, constructing a wind and light combined output distribution function in each time period through a Frank-Copula function; generating N initial wind-solar combined output scenes by using a Latin hypercube sampling method; and reducing the N initial wind and light combined output scenes to M by using a synchronous back substitution scene reduction technology of a Kantorovich distance to obtain typical output scenes with correlation between wind and light. According to the technical scheme, the correlation between wind and light is considered, and a scene reduction technology is introduced to reduce the number of initial samples, so that the sampling precision is ensured, and the calculation burden is reduced.

Description

Technical Field

[0001] This invention relates to a method for analyzing power system output scenarios, and more particularly to a method for analyzing power system combined wind and solar power output scenarios. Background Technology

[0002] Current research by scholars both domestically and internationally on the uncertainty of wind and solar power output focuses on improving the prediction accuracy by more rationally considering the correlation between wind and solar power and reducing its volatility. Furthermore, to maintain the temporal sequence of wind and solar power output, the Copula method is widely used in joint wind and solar power output prediction. Sampling these Copula functions can generate a large number of wind and solar power output scenarios. While this quantifies the uncertainty of wind and solar power output, it also requires a significant amount of computation. Therefore, it is difficult to simultaneously guarantee the accuracy requirements of wind and solar power prediction and meet the real-time computing needs of the power grid when performing wind and solar power output prediction. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a method for analyzing combined wind and solar power output scenarios in power systems, the technical solution of which is as follows:

[0004] The first aspect of the present invention provides a method for analyzing the combined wind and solar power output scenario of a power system, which includes the following steps:

[0005] Step 1: Obtain wind and solar data samples from wind farms and photovoltaic power plants;

[0006] Step 2: Divide the wind and solar data samples into time periods, perform nonparametric kernel density estimation on the wind and solar data samples of each time period, and determine the probability density function and probability distribution function of wind and solar output for each time period;

[0007] Step 3: Based on the probability distribution function of wind and solar power output in each time period, construct the wind and solar power output distribution function in each time period using the Frank-Copula function;

[0008] Step 4: Divide the wind-solar power output distribution function of each time period into N sub-intervals, take the midpoint of each sub-interval and perform an inverse function transformation to obtain N wind-solar power output sampling values, and generate N initial wind-solar power output scenarios;

[0009] Step 5: Set the number of target scenes M, where M is less than N, and calculate the Kantorovich distance D between each initial landscape-sunlight combined output scene. K (ω i ,ω j );

[0010] Step 6: Determine the relationship with scene ω k The closest scene ω m Calculate D K (ωk ,ω m ) and scene probability Pr{ω m The product of )};

[0011] Step 7: Repeat step 6 for each initial landscape-sunlight combined output scene, selecting PD. K (ω k ,ω m The scene with the smallest size is reduced, and the number of scenes N = N-1 is updated. The scene to be reduced is ω. s The probability of ' is superimposed on the nearest scene ω. s Above, i.e., Pr{ω s}=Pr{ω s}+Pr{ω s '};

[0012] Step 8: Repeat steps 5 to 7 for iterative processing until N = M, to obtain typical power output scenarios where wind and solar power are correlated.

[0013] Optionally, the wind and solar data samples in step 1 are measured data samples of wind speed and solar irradiance, respectively.

[0014] Optionally, in step 2, the wind data samples are divided into time periods, and nonparametric kernel density estimation is performed on the wind data samples of each time period to determine the probability density function and probability distribution function of wind output for each time period, as follows:

[0015] Based on the cut-in and cut-out characteristics of wind turbine generators, the power of a wind turbine generator is:

[0016]

[0017] In the formula, P WT,ca This indicates the real-time power and rated power of the wind turbine; v wind,in v wind,in Indicates the cut-in and cut-out wind speeds of the wind turbine; v wind,de Indicates the design wind speed of the wind turbine;

[0018] The nonparametric kernel density estimation establishes a probabilistic model for wind speed based on historical wind speed data as follows:

[0019]

[0020] In the formula, K(·) represents the kernel function; h is the window width, obtained through simulation calculation; n is the total number of historical wind speed samples, which increases as n approaches ∞. The probability distribution curve that converges to the actual wind speed;

[0021] Based on formulas (1) and (2), the output power of a single wind turbine is calculated. For a wind farm equipped with N wind turbines, the output power is:

[0022]

[0023] Optionally, in step 2, the optical data samples are divided into time periods, and nonparametric kernel density estimation is performed on the optical data samples of each time period to determine the probability density function and probability distribution function of the optical output for each time period. The method is as follows:

[0024] Photovoltaic array output power P s for:

[0025] P s =ηEA s (4)

[0026] In the formula, A s η represents the effective area of ​​the photovoltaic array and its photoelectric conversion efficiency;

[0027] The uncertainty model for the photovoltaic array output is established using nonparametric kernel density estimation as follows:

[0028]

[0029] In the formula: K(·) represents the kernel function; h is the window width, obtained through simulation calculation; n is the sample size, which is determined when n approaches ∞. It will converge to f(x).

[0030] Optionally, in step 3, based on the probability distribution functions of wind and solar power output for each time period, the process of constructing the combined wind and solar power output distribution function for a single time period using the Frank-Copula function is as follows:

[0031] The distribution function of the bivariate Frank Copula is as follows:

[0032]

[0033] In the formula, (x,y) is a two-dimensional random variable, and F(x) and F(y) are the marginal distribution functions of (x,y);

[0034] In the above formula, the correlation parameter λ∈(-∞,0)∪(0,+∞) has the following relationship with the Kendall rank correlation coefficient τ:

[0035]

[0036] The joint density function of wind and solar power output based on the Frank Copula function is:

[0037]

[0038] Optionally, the Kendall rank correlation coefficient τ is used as a measure of wind-solar correlation as follows:

[0039] τ=Pr{(X1-X2)(Y1-Y2)≥0}-Pr{(X1-X2)(Y1-Y2)<0} (9)

[0040] In the above formula, Pr{*} represents the probability of event * occurring; the positive and negative signs of (X1-X2) and (Y1-Y2) represent the consistency of the correlation between the values ​​of the two random variable samples.

[0041] Optionally, in step 5, the Kantorovich distance D between each initial wind-solar combined output scenario is calculated. K (ω i ,ω j The formula is:

[0042]

[0043] In the formula, The output value of wind power at each moment in the two scenarios; Let be the power output of photoelectric sensors at each moment in the two scenarios.

[0044] Optionally, in step 6, D is calculated. K (ω k ,ω m ) and scene probability Pr{ω m The product formula for} is as follows:

[0045] PD K (ω k ,ω m ) = D K (ω k ,ω m )·Pr{ω m} (11).

[0046] In a second aspect, the present invention provides a storage medium for storing the power system wind-solar combined output scenario analysis method described in any one of the present invention.

[0047] In a second aspect, the present invention provides a computer for executing the power system wind-solar combined output scenario analysis method described in any one of the claims.

[0048] Compared with the prior art, the technical solution of the present invention has the following advantages: considering the correlation between wind and light, scene reduction technology is introduced to reduce the number of initial samples, thereby ensuring sampling accuracy and reducing computational burden. Attached Figure Description

[0049] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0050] Figure 1 This is a schematic diagram of the overall process of an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of Latin hypercube sampling according to an embodiment of the present invention;

[0052] Figure 3 This is a simplified power curve diagram of a single fan in an embodiment of the present invention;

[0053] Figure 4 This is a distribution map of wind speed and solar irradiance data for a wind farm and a photovoltaic power station in a certain location in a certain year, according to an embodiment of the present invention.

[0054] Figure 5 This is a 3D bar chart of 1000 related wind and solar power output scenarios generated in an embodiment of the present invention;

[0055] Figure 6 These are line graphs of five typical wind and solar power output scenarios after reduction according to an embodiment of the present invention. Detailed Implementation

[0056] Example 1

[0057] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0058] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0059] like Figure 1 As shown in the figure, this embodiment provides a method for analyzing the combined wind and solar power output scenario of a power system, which includes the following steps:

[0060] Step 1: Obtain wind and solar data samples from wind farms and photovoltaic power plants;

[0061] Step 2: Divide the wind and solar data samples into time periods, perform nonparametric kernel density estimation on the wind and solar data samples of each time period, and determine the probability density function and probability distribution function of wind and solar output for each time period;

[0062] Step 3: Based on the probability distribution function of wind and solar power output in each time period, construct the wind and solar power output distribution function in each time period using the Frank-Copula function;

[0063] Step 4: Divide the wind-solar power output distribution function of each time period into N sub-intervals, take the midpoint of each sub-interval and perform an inverse function transformation to obtain N wind-solar power output sampling values, and generate N initial wind-solar power output scenarios;

[0064] Step 5: Set the number of target scenes M, where M is less than N, and calculate the Kantorovich distance D between each initial landscape-sunlight combined output scene. K (ω i ,ω j );

[0065] Step 6: Determine the relationship with scene ω k The closest scene ω m Then calculate D K (ω k ,ω m ) and scene probability Pr{ω m The product of )};

[0066] Step 7: Repeat step 6 for each initial landscape-sunlight combined output scene, selecting PD. K (ω k ,ω m Reduce the number of scenes to the minimum and update the number of scenes N = N-1, while also reducing the number of scenes to be reduced by ω. s The probability of ' is superimposed on the nearest scene ω. s Above, i.e., Pr{ω s}=Pr{ω s}+Pr{ω s '};

[0067] Step 8: Repeat steps 5 to 7 above for iterative processing until N = M, to obtain typical power output scenarios where wind and solar power are correlated.

[0068] In step 4, the combined wind and solar power output distribution function for each time period is divided into N sub-intervals, such as... Figure 2 As shown, the Latin hypercube sampling method is used for sampling. The Latin hypercube sampling method is a stratified sampling method based on Monte Carlo simulation. Compared to Monte Carlo, the Latin hypercube sampling method can achieve higher sampling accuracy with fewer sampling iterations. The basic idea of ​​the Latin hypercube sampling method is to divide the marginal distribution sampling space of each random variable into N subspaces, then extract the midpoints of these N subspaces and calculate the inverse function to obtain N sampling points. Finally, the sampled values ​​of the n random variables are sorted to form an n×N initial sample matrix. Essentially, it ensures that the sampling points cover the entire sampling area by stratifying the input probability distribution and extracting points from each layer.

[0069] The principle of steps 5 to 8 is the synchronous back-substitution scene reduction technique based on Kantorovich distance. Its basic principle is: determine the initial scene set and the empty set, put the scene in the initial scene set that is closest to other scenes into the empty set, so as to superimpose the scene with low probability onto the scene that is closest to it, and complete the reduction of the initial scene set. Repeat this process until the scene reduction is completed.

[0070] In some embodiments, the wind and solar data samples in step 1 are measured data samples of wind speed and solar irradiance, respectively.

[0071] In some embodiments, in step 2, the wind data samples are divided into time periods, and nonparametric kernel density estimation is performed on the wind data samples of each time period to determine the probability density function and probability distribution function of wind output for each time period. The method is as follows:

[0072] Based on the cut-in and cut-out characteristics of wind turbine generators, the power of a wind turbine generator is:

[0073]

[0074] In the formula, P WT,ca This indicates the real-time power and rated power of the wind turbine; v wind,in v wind,in Indicates the cut-in and cut-out wind speeds of the wind turbine; v wind,de Indicates the design wind speed of the wind turbine;

[0075] The nonparametric kernel density estimation establishes a probabilistic model for wind speed based on historical wind speed data as follows:

[0076]

[0077] In the formula, K(·) represents the kernel function; h is the window width, obtained through simulation calculation; n is the total number of historical wind speed samples, which increases as n approaches ∞. The probability distribution curve that converges to the actual wind speed;

[0078] Based on formulas (1) and (2), the output power of a single wind turbine is calculated. For a wind farm equipped with N wind turbines, the output power is:

[0079]

[0080] Wind energy is characterized by randomness and volatility, which leads to fluctuations in wind turbine output, potentially impacting power system stability. In actual wind farms, wind power output is affected by various factors, such as wind speed fluctuations, wind direction changes, equipment operating conditions, wind energy utilization coefficients of turbines, and wake effects between turbines. Therefore, wind farms are simplified in wind power output modeling, ignoring the internal structure of turbines and the complex connections between them.

[0081] The cut-in and cut-out characteristics of wind turbines refer to the fact that the active power output of wind turbines changes with wind speed. When the cut-in wind speed is not reached, the wind turbine will be in a stopped state and will not output power. If the real-time wind speed is between the cut-in wind speed and the rated wind speed, the turbine will start outputting power. To simplify calculations, the relationship between wind power and wind speed at this time is linearized as follows: Figure 3 As shown, when the real-time wind speed is between the rated wind speed and the cut-out wind speed, the wind turbine will output power according to the rated value; once the real-time wind speed exceeds the cut-out wind speed, the wind turbine will be forced to stop to prevent damage to its own structure.

[0082] Formula (1) is based on Figure 3 To express.

[0083] Among them, nonparametric kernel density estimation does not require the calculation of the probability distribution parameters of wind speed, and can establish a probability model of wind speed based on historical wind speed data, which can effectively reduce the impact of wind speed uncertainty.

[0084] Where n is the total number of historical wind speed samples, and when n approaches ∞, It will converge to the probability distribution curve of the actual wind speed, so there is no need to calculate its distribution parameters.

[0085] Since grid-connected wind farms are equipped with reactive power regulation devices, they can ensure zero reactive power exchange between the wind farm and the power grid. Therefore, the reactive power demand of the wind farm can be disregarded, and only its active power output needs to be considered.

[0086] Once the wind speed is obtained, the output of a single wind turbine can be calculated using formula (1).

[0087] In some embodiments, the optical data samples in step 2 are divided into time periods, and nonparametric kernel density estimation is performed on the optical data samples of each time period to determine the probability density function and probability distribution function of the optical output for each time period, as follows:

[0088] Photovoltaic array output power P s for:

[0089] P s =ηEA s (4)

[0090] In the formula, A sη represents the effective area of ​​the photovoltaic array and its photoelectric conversion efficiency;

[0091] The uncertainty model for the photovoltaic array output is established using nonparametric kernel density estimation as follows:

[0092]

[0093] In the formula: K(·) represents the kernel function; h is the window width, obtained through simulation calculation; n is the sample size, which is determined when n approaches ∞. It will converge to f(x).

[0094] Among them, the nonparametric kernel density estimation, which has better applicability, stronger stability and smaller error, is selected to establish the uncertainty model of photovoltaic array output.

[0095] Where n is the sample size, and when n approaches ∞, It will converge to f(x), so no distribution parameter is needed.

[0096] In some embodiments, the process of constructing the combined wind and solar power output distribution function for a single time period using the Frank-Copula function based on the probability distribution functions of wind and solar power output for each time period in step 3 is as follows:

[0097] The distribution function of the bivariate Frank Copula is as follows:

[0098]

[0099] In the formula, (x,y) is a two-dimensional random variable, and F(x) and F(y) are the marginal distribution functions of (x,y);

[0100] In the above formula, the correlation parameter λ∈(-∞,0)∪(0,+∞) has the following relationship with the Kendall rank correlation coefficient τ:

[0101]

[0102] The joint density function of wind and solar power output based on the Frank Copula function is:

[0103]

[0104] Copula functions are "connection functions" that link multiple non-normally related random variables together. Copula functions have no special requirements or restrictions on the type of marginal distribution functions of random variables and can maintain the correlation between random variables during transformation, making them widely used in random variable correlation modeling. This study mainly focuses on the uncertainty of wind and solar power; therefore, the correlation of wind and solar power output can be described by establishing a binary Copula joint distribution function. For a two-dimensional random variable (x, y), assuming its marginal distribution functions are F(x) and F(y), and their joint distribution function is denoted as H(x, y), then there exists a Copula function C(F(x), F(y)) such that the following equation holds:

[0105] H(x,y)=C(F(x),F(y)) (12)

[0106] If both F(x) and F(y) have inverse functions, then Equation 12 can be expressed as follows within their domain:

[0107] C(x,y)=F(F -1 (x),F -1 (y)) (13)

[0108] Taking the partial derivative of Equation 13 yields the joint probability density function of the bivariate distribution:

[0109]

[0110] In the above formula, C(F(x),F(y)) is the probability density function of the Copula function, and F(x) and F(y) represent the probability density functions of random variables x and y, respectively.

[0111] Among bivariate copula functions, the Archimedes copula function is more widely used. Within the Archimedes copula family, the Frank copula exhibits a uniform and symmetrical distribution, effectively describing both positive and negative correlations between random variables. Since wind power and solar power output in the same region often show a negative correlation, the Frank copula is used to construct their joint distribution function.

[0112] The rank correlation coefficient is a statistical indicator that measures the degree of correlation between two random variables. Commonly used rank correlation coefficients include Pearson's linear correlation coefficient, Spearman's rank correlation coefficient, and Kendall's rank correlation coefficient. Among them, Kendall's rank correlation coefficient analyzes the strength of the correlation coefficient by examining the relationship between sample data of random variables. It can reflect the consistency of change between two random variables and has better stability, monotonically increasing properties, and adaptability. Therefore, it is selected as a measure of the correlation between wind and solar power.

[0113] In some embodiments, the Kendall rank correlation coefficient τ is used as a measure of wind-solar correlation as follows:

[0114] τ=Pr{(X1-X2)(Y1-Y2)≥0}-Pr{(X1-X2)(Y1-Y2)<0} (9)

[0115] In the above formula, Pr{*} represents the probability of event * occurring; the positive and negative signs of (X1-X2) and (Y1-Y2) represent the consistency of the correlation between the values ​​of the two random variable samples.

[0116] In some embodiments, step 5 calculates the Kantorovich distance D between each initial wind-solar combined output scenario. K (ω i ,ω j The formula is:

[0117]

[0118] In the formula, The output value of wind power at each moment in the two scenarios; Let be the power output of photoelectric sensors at each moment in the two scenarios.

[0119] In some embodiments, D is calculated in step 6. K (ω k ,ω m ) and scene probability Pr{ω m The product formula for} is as follows:

[0120] PD K (ω k ,ω m ) = D K (ω k ,ω m )·Pr{ω m} (11).

[0121] Example 2

[0122] Using annual wind speed and solar irradiance data (sampling interval of 10 minutes) from wind farms and photovoltaic power plants in a certain area as historical samples, the corresponding power output sample sets of wind farms and photovoltaic power plants were calculated. The historical wind speed and solar irradiance distributions are shown below. Figure 4 .

[0123] The wind speed throughout the year is divided into four seasons based on the date: days 1-91 are spring, days 92-183 are summer, days 184-274 are autumn, and days 275-365 are winter.

[0124] Taking the hourly data of summer solar power output as an example, the scene generation method was verified according to the method in Example 1. The generated 1000 relevant solar power output scenes are as follows: Figure 5 As shown, the number of scenic output scenes has been reduced to 5. Figure 6 As shown.

[0125] from Figure 6 It can be observed that the five typical landscape power output scenarios after reduction are quite close to the actual landscape power output curves, indicating that the scenario generation and reduction methods in this chapter can effectively screen out typical landscape power output scenarios.

[0126] Example 3

[0127] Based on the same inventive concept, embodiments of the present invention also provide a computer device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor. When the processor executes the computer program, it implements the steps of the power system wind-solar combined output scenario analysis method described in either Embodiment 1 or Embodiment 2.

[0128] Example 4

[0129] Based on the same inventive concept, this embodiment of the invention also provides a computer storage medium storing a computer program, which, when executed by a processor, implements the steps of the power system wind-solar combined output scenario analysis method described in either Embodiment 1 or Embodiment 2.

[0130] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0132] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0133] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0134] In the description of this specification, the terms "one embodiment," "some embodiments," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for analyzing a wind-solar combined output scene of a power system, characterized in that, The method comprises the following steps: Step 1: obtaining wind and light data samples of a wind farm and a photovoltaic power station; Step 2: dividing the wind and light data samples according to time periods, performing non-parametric kernel density estimation on the wind and light data samples of each time period, and determining the probability density function and the probability distribution function of the wind and light output of each time period; Step 3: based on the probability distribution function of the wind and light output of each time period, constructing the wind-light joint output distribution function of each time period through a Frank-Copula function; Step 4: dividing the wind-light joint output distribution function of each time period into N subintervals, taking the midpoint of each subinterval and performing inverse function transformation to obtain N wind-light output sampling values, and generating N initial wind-light joint output scenarios; Step 5: Set the target scene number M, M is less than N, calculate the Kantorovich distance D between each initial wind and light combined output scene K (ω i ,ω j ) Step 6: Determine the scene ω k The closest scene ω m , compute the product of D K (ω k ,ω m ) and the scene probability Pr{ω m} Step 7: Repeat step 6 for each initial landscape-sunlight combined output scene, selecting PD. K (ω k ,ω m The scene with the smallest size is reduced, and the number of scenes N = N-1 is updated. The scene to be reduced is ω. s The probability of ' is superimposed on the nearest scene ω. s Above, i.e., Pr{ω s }=Pr{ω s }+Pr{ω s '}; Step 8: repeating steps 5 to 7 for repeated iteration until N=M, and obtaining typical output scenarios of wind and light with correlation.

2. The method of claim 1, wherein, The wind and light data samples in step 1 are wind speed and light intensity measured data samples.

3. The method of claim 1, wherein, In step 2, the wind data samples are divided according to time periods, non-parametric kernel density estimation is performed on the wind data samples of each time period, and the probability density function and the probability distribution function of the wind output of each time period are determined as follows: Based on the characteristics of wind turbine cut-in and cut-out, the power of the wind turbine is: In the formula, P WT,ca represents the real-time power and rated power of the wind turbine;v wind,in , v wind,in represents the cut-in and cut-out wind speed of the wind turbine;v wind,de represents the design wind speed of the wind turbine; The non-parametric kernel density estimation establishes the probability model of the wind speed based on the historical wind speed data as follows: In the formula, K(·) represents a kernel function; h is a window width, v is a wind speed, which is obtained through simulation calculation; n is a total amount of historical wind speed samples, v i is a wind speed of the i th sample; when n tends to ∞, converges to a probability distribution curve of an actual wind speed. Based on formulas (1) and (2), the output of a single wind turbine is the output power of a wind farm equipped with N wind turbines: where P w,i is the power output of the i-th wind turbine.

4. The method of claim 1, wherein, In step 2, the light data samples are divided according to time periods, non-parametric kernel density estimation is performed on the light data samples of each time period, and the probability density function and the probability distribution function of the light output of each time period are determined as follows: Photovoltaic array output power P s is: P s = ηEA s (4) where A s η is the effective area of the photovoltaic array and its photoelectric conversion efficiency, and E is the solar radiation intensity. The non-parametric kernel density estimation establishes the uncertainty model of the photovoltaic array output as follows: where K(·) represents the kernel function; h is the window width, which is obtained by simulation calculation; n is the sample size, P si is the output power of the photovoltaic array for the i th sample; when n tends to ∞, will converge to the probability distribution curve of the actual photovoltaic array output power.

5. The method of claim 1, wherein the method further comprises: In step 3, based on the probability distribution function of the wind and light output of each time period, the wind-light joint output distribution function of each single time period is constructed through a Frank-Copula function as follows: The distribution function of the binary Frank Copula is as follows: In the formula, (x, y) is a two-dimensional random variable, F(x) and F(y) are the edge distribution functions of (x, y); In the above formula, the correlation parameter λ∈(-∞, 0)∪(0, +∞), and the relationship with the Kendall rank correlation coefficient τ is as follows: The wind-light output joint density function based on the Frank Copula function is as follows: where P w is the output power of the wind turbine, P s is the output power of the photovoltaic cell array.

6. The method of claim 5, wherein, The Kendall rank correlation coefficient τ is used as a measure of the correlation between wind and light as follows: τ=Pr{(X1-X2)(Y1-Y2)≥0}-Pr{(X1-X2)(Y1-Y2)<0} (9) In the above formula, Pr* represents the probability of event *; the positive and negative of (X1-X2), (Y1-Y2) represent the consistency of the correlation between the two groups of random variable samples.

7. The method of claim 1, wherein the method further comprises: The step 5 calculates the Kantorovich distance D between each initial wind-solar combined output scenario K (ω i ,ω j ) is as follows: In the formula, Pwind(t) is the power output value of wind power at each time in the two scenarios; Ppv(t) is the power output value of photovoltaic power at each time in the two scenarios.

8. The method of claim 1, wherein the method further comprises: The product of the D K (ω k ,ω m ) and the scene probability Pr{ω m )} is calculated as follows: PD K (ω k ,ω m ) = D K (ω k ,ω m ) · Pr{ω m} (11).

9. A storage medium, characterized by A storage medium for storing the power system wind-light joint output scenario analysis method of any one of claims 1 to 8.

10. A computer, comprising: A storage medium for storing the power system wind-light joint output scenario analysis method of any one of claims 1 to 8.

Images