A method and system for generating a water, wind and light scene driven by multiple correlations and joint driving

Abstract

The present invention belongs to the technical field related to power scenario generation, and discloses a method and system for generating water-wind-solar scenarios driven by multiple correlations. The method includes: collecting historical data of three resources of water, wind and solar, determining the order of scenario generation, and extracting the existing variable information of each resource at time t-1 and time t; constructing the conditional distribution of variable x2 when variable x1 is known through a Gaussian mixture model, and optimizing the parameters of the Gaussian mixture model according to the existing variable information; for the first resource, optimizing the model parameters by using the existing variable information at time t-1 and time t; for the second and third resources, combining the existing variable information at time t of the previous resource, at time t-1 of this resource, and at time t of this resource to optimize the model parameters; and using the optimized Gaussian mixture model to generate scenarios for the corresponding resources. The present invention not only considers the autocorrelation relationship of a single resource but also considers the cross-correlation relationship between resources.

Description

Technical Field

[0001] The present invention belongs to the technical field related to power scenario generation, and more specifically, relates to a method and system for generating water-wind-solar scenarios driven by multiple correlations in combination. Background Art

[0002] In recent years, with the rapid development of renewable energy, the proportion of hydropower, wind power, and photovoltaic power generation in the global energy structure has been continuously increasing. These resources are not only clean and renewable but also can effectively reduce greenhouse gas emissions and contribute to achieving the carbon neutrality goal. However, hydropower, wind power, and photovoltaic power generation all have obvious randomness and volatility and are greatly affected by climate and natural conditions, resulting in large uncertainties in their power generation at different time scales. How to accurately quantify these uncertainties is the key to ensuring the safe and stable operation of the power system.

[0003] In this context, scenario generation technology has become an effective means to characterize uncertainties. By constructing a reasonable scenario set, various possible situations under different natural resource conditions can be simulated, providing a scientific basis for dispatching decisions and planning analysis. However, current research mainly focuses on the scenario generation of a single resource or only considers the simple correlations between single resources, lacking a systematic analysis and modeling of the complex interrelationships among hydropower, wind power, and photovoltaic power generation. Since these three resources are often affected by climate conditions in nature and have significant multiple correlations, ignoring this correlation will lead to deviations between the scenario generation results and the actual situation, thus affecting the accuracy of subsequent analysis. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement requirements of the prior art, the present invention provides a method and system for generating water-wind-solar scenarios driven by multiple correlations in combination, which are used to solve the problems that the existing scenario generation technology mainly focuses on the scenario generation of a single resource or only considers the simple correlations between single resources, lacking a systematic analysis and modeling of the complex interrelationships among hydropower, wind power, and photovoltaic power generation.

[0005] To achieve the above object, according to one aspect of the present invention, a method for generating water-wind-solar scenarios driven by multiple correlations in combination is provided, including:

[0006] Offline training stage:

[0007] S1, collect historical data on water resource runoff, wind resource wind power output, and light resource photovoltaic power output, calculate the interrelationships between water-wind-solar pairwise resources according to the historical data, determine the order of water-wind-solar scenario generation based on the interrelationships, and sequentially take the first resource, the second resource, and the third resource in sequence; and extract the existing variable information of each resource at the t-1 period and the t period according to the historical data;

[0008] S2. Construct Gaussian mixture models corresponding to three resources respectively: construct the two-dimensional joint distribution of variable x1 and variable x2 and the conditional distribution of variable x2 when variable x1 is known through the Gaussian mixture model, then generate the simulated value of variable x2 according to variable x1, and optimize the Gaussian mixture model parameters corresponding to each resource by establishing a sample data set based on the existing variable information, so as to obtain the Gaussian mixture models with optimized parameters corresponding to each resource respectively;

[0009] Among them, for the first resource, the existing variable information in the (t - 1) period is corresponded to variable x1, and the existing variable information in the t period is corresponded to variable x2, and the Gaussian mixture model parameters are optimized;

[0010] For any one of the second resource and the third resource, after combining the existing variable information in the t period of the resources ranked before this any one resource and the existing variable information in the (t - 1) period of this any one resource and corresponding it to variable x1, and corresponding the existing variable information in the t period of this any one resource to variable x2, the Gaussian mixture model parameters corresponding to this any one resource are optimized;

[0011] Online scenario generation stage:

[0012] S3. Use the Gaussian mixture models with optimized parameters corresponding to each resource to generate scenarios for the corresponding resources.

[0013] According to the multi - correlated joint - driven water - wind - light scenario generation method provided by the present invention, in S2, optimizing the Gaussian mixture model parameters corresponding to each resource by establishing a sample data set according to the existing variable information specifically includes:

[0014] For the second resource, after arranging the existing variable information in the t period of the first resource and the existing variable information in the (t - 1) period of the second resource in sequence and corresponding it to variable x1, and corresponding the existing variable information in the t period of the second resource to variable x2, the Gaussian mixture model parameters are optimized;

[0015] For the third resource, after arranging the existing variable information in the t period of the first resource, the existing variable information in the t period of the second resource and the existing variable information in the (t - 1) period of the third resource in sequence and corresponding it to variable x1, and corresponding the existing variable information in the t period of the third resource to variable x2, the Gaussian mixture model parameters are optimized.

[0016] According to the multi - correlated joint - driven water - wind - light scenario generation method provided by the present invention, S3 specifically includes:

[0017] For the first resource, the average value of the existing variable information in the maximum period is used as variable x1 to generate the first simulation value of variable x2, and the first simulation value of variable x2 is used as the first generated value of the scenario. Then, the previous generated value is used as the new variable x1 to generate the current variable x2;

[0018] For the second resource, the first generated value of the first resource is combined with the average value of the existing variable information of the second resource's maximum period as variable x1 to generate the first simulated value of variable x2. The first simulated value of variable x2 is used as the first generated value of the scenario. Then, the current generated value of the first resource is combined with the previous generated value of the second resource as the new variable x1 to generate the current variable x2.

[0019] For the third resource, the first generated value of the first resource, the first generated value of the second resource and the average value of the existing variable information of the third resource's maximum time period are combined as variable x1 to generate the first simulated value of variable x2. The first simulated value of variable x2 is used as the first generated value of the scene. Then, the current generated value of the first resource, the current generated value of the second resource and the previous generated value of the third resource are combined as the new variable x1 to generate the current variable x2.

[0020] According to the method for generating water, wind and light scenes driven by multiple correlation joint drive provided by the present invention, in S1, the mutual relationship between water, wind and light resources is calculated according to historical data, and the order of generating water, wind and light scenes is determined according to the mutual relationship, which specifically includes:

[0021] Calculate the mutual information between the water, wind and light resources based on historical data, use the mutual information to determine the relationship between the water, wind and light resources, and obtain the mutual information value between the water, wind and light resources;

[0022] Calculate the sum of the mutual information values ​​between any resource and other resources, obtain the comprehensive mutual information index of any resource, and determine the order of generating the three resource scenarios according to the comprehensive mutual information index.

[0023] According to the method for generating water, scenery and light scenes driven by multiple correlations provided by the present invention, in S1, the existing variable information of each resource in the t-1 period and the t period is extracted according to historical data, specifically:

[0024] ;

[0025] ;

[0026] ;

[0027] Among them, DH is the multi-year historical data of the water resource runoff h, DW is the multi-year historical data of the wind resource wind power output w, and DP is the multi-year historical data of the light resource photovoltaic power output p; the existing variable information of the three resources of water, wind and light at the t-1 time period and the t time period are h t-1 and h t 、w t-1 and w t 、p t-1 and p t ; , and N is determined according to the target time scale.

[0028] According to the method for generating water, wind and light scenarios driven by multiple correlations jointly provided by the present invention, in S2, the Gaussian mixture model parameters corresponding to each resource are optimized by establishing a sample data set according to the existing variable information, and the Gaussian mixture models with optimized parameters corresponding to each resource are obtained, specifically including:

[0029] For any resource, the variables x1 and x2 change with the increase of the time period t, and the parameters of the Gaussian mixture model are optimized respectively by using the existing variable information of each time period t, and N models with optimized parameters corresponding to any resource are obtained for generating scenarios of N time periods of any resource.

[0030] According to the method for generating water, wind and light scenarios driven by multiple correlations jointly provided by the present invention, the Gaussian mixture model in S2 is specifically:

[0031] ;

[0032] ;

[0033] In the formula: is the two-dimensional joint distribution of the variables x1 and x2, is the conditional distribution of the variable x2 when the variable x1 is known, is the binary probability density function, is the conditional probability density function of the variable x2 when the variable x1 is known;

[0034] Among them, the binary probability density function , and the mathematical expression is:

[0035] ;

[0036] In the formula: represents the binary Gaussian probability density function; K is the number of components of the Gaussian mixture model, which is a hyperparameter; is the weight of the k-th component; are the mean vector and covariance matrix of the k-th component respectively.

[0037] According to the method for generating a water-wind-light scene driven by multiple correlations provided by the present invention, the optimization of any resource corresponding to the Gaussian mixture model for each resource in S2 is specifically to optimize the hyperparameters K, weights , mean vector and covariance matrix in the Gaussian mixture model, and the specific optimization steps include:

[0038] Preset several different hyperparameter K values;

[0039] For any preset K value, use the clustering method to divide the sample data set into K clusters; take the center of each cluster as the mean vector of each component in the Gaussian mixture model; calculate the difference between the data points in each cluster and the cluster center to obtain the covariance matrix ; calculate the proportion of the number of samples in each cluster to the total number of samples to obtain the weight ;

[0040] Take the obtained parameters as the initial values, use the expectation maximization algorithm to optimize the model parameters, and obtain the optimized mean vector, covariance matrix, and weight parameter values of the model corresponding to any preset K value;

[0041] Use the comprehensive strategy of the Akaike information criterion and the Bayesian information criterion to evaluate the Gaussian mixture models under different preset K values, and select the preset K value corresponding to the Gaussian mixture model with the smallest comprehensive strategy value as the best K value;

[0042] Perform the Kolmogorov-Smirnov test on the theoretical distribution of the Gaussian mixture model corresponding to the best K value and the empirical distribution of the sample data set, where the significance level is set to 0.2. If the test passes, the model training is completed. If the test fails, adjust the preset K value and re-optimize the model parameters.

[0043] According to the method for generating a water-wind-light scene driven by multiple correlations provided by the present invention, the Akaike information criterion is specifically:

[0044] ;

[0045] The Bayesian information criterion is specifically:

[0046] ;

[0047] Where: is the number of parameters of the Gaussian mixture model under K components, representing the model complexity; M is the amount of data in the sample data set; the comprehensive strategy value is the weighted sum of the Akaike information criterion and the Bayesian information criterion.

[0048] According to another aspect of the present invention, there is provided a water-wind-solar scene generation system driven by multiple correlations and associations, the system comprising a memory and a processor, the memory storing a computer program, and the processor, when executing the computer program, executing the water-wind-solar scene generation method according to any one of the above.

[0049] Generally speaking, compared with the prior art by the above technical solutions conceived by the present invention, the water-wind-solar scene generation method and system provided by the present invention:

[0050] 1. Not only consider the autocorrelation relationship of a single resource but also consider the cross-correlation relationship between resources. The sample data set during the Gaussian mixture model training associates the relevant resources, and the Gaussian mixture model is used to directly describe the joint distribution and conditional distribution between the water-wind-solar resources without pre-supposing a specific distribution form of the water-wind-solar. It has strong adaptability, can effectively depict the autocorrelation relationship of a single resource and the cross-correlation relationship between resources. The scene generated by the optimized Gaussian mixture model can comprehensively depict the multiple correlations of the water-wind-solar, which helps to more realistically reflect the fluctuation characteristics of multiple renewable energy sources, and further provides more accurate basic data for the scheduling and planning of the power system;

[0051] 2. By embedding the joint relationship and conditional distribution between each resource into the scene generation process, gradually generating the time series scenes of the three resources of water, wind, and solar, it can dynamically simulate the linkage changes between resources, making the generated scenes more conform to the resource fluctuation characteristics in the actual operating environment, thereby significantly improving the adaptability and operating stability of the water-wind-solar multi-energy complementary system;

[0052] 3. Using the conditional distribution to decompose the high-dimensional water-wind-solar data generation problem into multiple low-dimensional generation problems, effectively avoiding the difficulty of high-dimensional model parameter estimation; combining the nesting and step-by-step generation strategy of multiple model results, truly reflecting the mutual influence relationship between each resource, not only reducing the computational complexity but also retaining the multiple correlations between data, further improving the overall computational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flowchart of the water-wind-solar scene generation method provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0055] Please refer to Figure 1 , Embodiment 1 provides a method for generating a water-wind-solar scenario with multiple correlation joint driving. The method for generating a water-wind-solar scenario includes:

[0056] Offline training stage:

[0057] S1. Collect historical data on water resource runoff, wind resource wind power output, and solar resource photovoltaic power output. Calculate the mutual relationship between any two of the water-wind-solar resources based on the historical data. Determine the order of generating the water-wind-solar scenario according to the mutual relationship, and sequentially take them as the first resource, the second resource, and the third resource in the order; and extract the existing variable information of each resource at the (t - 1)th period and the tth period according to the historical data.

[0058] S2. Construct Gaussian mixture models corresponding to the three resources respectively: Construct the two-dimensional joint distribution of variable x1 and variable x2 and the conditional distribution of variable x2 when variable x1 is known through the Gaussian mixture model, and then generate the simulated value of variable x2 according to variable x1, and optimize the parameters of the Gaussian mixture model corresponding to each resource by establishing a sample data set based on the existing variable information, and obtain the Gaussian mixture models with optimized parameters corresponding to each resource.

[0059] Among them, for the first resource, the existing variable information at the (t - 1)th period is corresponding to variable x1, and the existing variable information at the tth period is corresponding to variable x2, and optimize the parameters of the Gaussian mixture model; that is, for the first resource, combine the existing variable information at the (t - 1)th period and the existing variable information at the tth period as the sample data set input to the Gaussian mixture model corresponding to the tth period.

[0060] For any one of the second resource and the third resource, combine the existing variable information at the tth period of the resource ranked before this any one resource and the existing variable information at the (t - 1)th period of this any one resource, and then correspond it to variable x1, and correspond the existing variable information at the tth period of this any one resource to variable x2, and optimize the parameters of the Gaussian mixture model corresponding to this any one resource; that is, for any one of the second resource and the third resource, combine the existing variable information at the tth period of the resource ranked before this any one resource, the existing variable information at the (t - 1)th period of this any one resource, and the existing variable information at the tth period of this any one resource as the sample data set input to the Gaussian mixture model corresponding to the tth period of this any one resource.

[0061] Online scenario generation stage:

[0062] S3. Use the Gaussian mixture models with optimized parameters corresponding to each resource to generate scenarios for the corresponding resources.

[0063] Further, the optimization of the Gaussian mixture model parameters corresponding to each resource by establishing a sample data set according to the existing variable information in S2 specifically includes:

[0064] For the second resource, the existing variable information of the first resource at time t and the existing variable information of the second resource at time t - 1 are arranged in sequence, that is, after being arranged in order and then corresponding to the variable x1, and the existing variable information of the second resource at time t corresponds to the variable x2, and the Gaussian mixture model parameters are optimized;

[0065] For the third resource, the existing variable information of the first resource at time t, the existing variable information of the second resource at time t, and the existing variable information of the third resource at time t - 1 are arranged in sequence, that is, after being arranged in order and then corresponding to the variable x1, and the existing variable information of the third resource at time t corresponds to the variable x2, and the Gaussian mixture model parameters are optimized.

[0066] Further, S3 specifically includes:

[0067] For the first resource, the average value of the existing variable information when the maximum time period, that is, t takes the maximum value, is used as the variable x1 to generate the first simulated value of the variable x2. The first simulated value of the variable x2 is used as the first generated value of the scenario. After that, the previous generated value is used as the new variable x1 to generate the current variable x2;

[0068] For the second resource, the combination of the first generated value of the first resource and the average value of the existing variable information when the maximum time period of the second resource, that is, t takes the maximum value, is used as the variable x1 to generate the first simulated value of the variable x2. The first simulated value of the variable x2 is used as the first generated value of the scenario. After that, the combination of the current generated value of the first resource and the previous generated value of the second resource is used as the new variable x1 to generate the current variable x2;

[0069] For the third resource, the combination of the first generated value of the first resource, the first generated value of the second resource, and the average value of the existing variable information when the maximum time period of the third resource, that is, t takes the maximum value, is used as the variable x1 to generate the first simulated value of the variable x2. The first simulated value of the variable x2 is used as the first generated value of the scenario. After that, the combination of the current generated value of the first resource, the current generated value of the second resource, and the previous generated value of the third resource is used as the new variable x1 to generate the current variable x2.

[0070] In some specific embodiments, calculating the mutual relationship between the water, wind, and light resources according to the historical data in S1, and determining the order of generating the water, wind, and light scenarios based on the mutual relationship specifically includes:

[0071] Calculate the mutual information between any two of the water, wind, and light resources based on historical data, determine the mutual relationship between any two of the water, wind, and light resources using the mutual information, and obtain the mutual information values between any two of the water, wind, and light resources;

[0072] Calculate the sum of the mutual information values between any one resource and the other resources, obtain the comprehensive mutual information index of any one resource, and determine the sequence of generation of the three resource scenarios according to the comprehensive mutual information index.

[0073] The mutual information values between any two of the water, wind, and light resources are specifically as follows:

[0074]

[0075] In the formula, is the mutual information value between random variables X and Y (such as water and wind, wind and light, light and water), x and y are the observed values of random variables X and Y, and they represent the specific data points in the random experiment or dataset. is the joint probability distribution of x and y, are the marginal probability distributions of x and y respectively.

[0076] The comprehensive mutual information index of a resource, which represents the overall correlation between a certain resource and the other two resources; determine the sequence of scenario generation according to the magnitude of the comprehensive mutual information index value of each resource, and the larger the comprehensive mutual information index, the higher the priority. The formula for the comprehensive mutual information index is as follows:

[0077]

[0078] In the formula, is the comprehensive mutual information index of random variable X.

[0079] Furthermore, the historical measured water, wind, and light data are all matrices D of M rows and N columns:

[0080]

[0081] In the formula, D represents the historical data of resources over the years. Let DH be the historical data of the water resource runoff h over the years, DW be the historical data of the wind resource wind power output w over the years, and DP be the historical data of the light resource photovoltaic power output p over the years. M is the number of years; if it is daily-scale data, N is 365; if it is dekadal-scale data, N is 36; if it is monthly-scale data, N is 12.

[0082] In S1, according to the historical data, the specific information of the existing variables in the t-1 period and the t period of each resource is to extract the relevant variables according to the historical measured water, wind, and light data of the target time scale. The six existing variable information related to water, wind, and light in the t-1 and t periods is as follows:

[0083]

[0084] Among them, the existing variable information of the three resources of water, wind, and light at time t-1 and time t are h t-1 and h t 、w t-1 and w t 、p t-1 and p t ; , N is determined according to the target time scale. That is, the existing variable information of any time t is a column of data corresponding to that time t in the historical data including multiple years.

[0085] Specifically, in S2, the Gaussian mixture model parameters corresponding to each resource are optimized by establishing a sample data set based on the existing variable information, and the Gaussian mixture models with optimized parameters corresponding to each resource are obtained, which specifically include:

[0086] For any resource, the variables x1 and x2 in the Gaussian mixture model change with the increase of time t. The parameters of the Gaussian mixture model are optimized using the existing variable information of each time t, and N models with optimized parameters corresponding to any resource are obtained for scenario generation of N time periods of that resource. That is, for any resource, when t takes values from 1 to N, it corresponds to N Gaussian mixture models. The N Gaussian mixture models are respectively input into the corresponding sample data sets for optimization. When generating scenarios, the optimized Gaussian mixture models corresponding to each of the N time periods are used for scenario generation.

[0087] The Gaussian mixture model in S2 is specifically:

[0088]

[0089] In the formula: is the two-dimensional joint distribution of variables x1 and x2, is the conditional distribution of variable x2 when variable x1 is known, is the binary probability density function, is the conditional probability density function of variable x2 when variable x1 is known;

[0090] Among them, the binary probability density function , the mathematical expression is:

[0091]

[0092] In the formula: represents the binary Gaussian probability density function; K is the number of components of the Gaussian mixture model, which is a hyperparameter; is the weight of the k-th component; are the mean vector and covariance matrix of the k-th component respectively.

[0093] The optimization of the Gaussian mixture model parameters corresponding to each resource in S2 specifically refers to the optimization of the hyperparameters K, weights , mean vectors and covariance matrices to determine. The specific optimization steps for each parameter include:

[0094] S21, preset several different hyperparameter K values in advance; for example, the preset K values can be taken as 1, 2, 3... 20 respectively, that is, 20 K values are preset in advance. The specific preset value size of the K value can be selected according to experience and is not specifically limited;

[0095] S22, for any preset K value, use the K-means clustering method to divide the sample data set into K clusters; take the center of each cluster as the mean vector of each component in the Gaussian mixture model ; calculate the difference between the data points in each cluster and the cluster center to obtain the covariance matrix ; calculate the proportion of the number of samples in each cluster to the total number of samples to obtain the weight ;

[0096] S23, take the parameters obtained in S22 as the initial values, use the expectation maximization (EM) algorithm to optimize the model parameters, and update iteratively until the model converges to obtain the optimized mean vector, covariance matrix, and weight parameter values corresponding to any preset K value;

[0097] S24, use the comprehensive strategy of the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) to evaluate the Gaussian mixture model under different preset K values, and select the preset K value corresponding to the Gaussian mixture model with the smallest comprehensive strategy value as the best K value;

[0098] S25, perform the Kolmogorov-Smirnov (KS) test on the theoretical distribution of the Gaussian mixture model corresponding to the best K value and the empirical distribution of the sample data set to evaluate its goodness of fit, where the significance level is set to 0.2. If the test passes, it is considered that the model fits the data well and the parameter values are determined, and the model training is completed. If the test fails, adjust the preset K value and re-optimize the model parameters, that is, re-perform S21 - S25.

[0099] Among them, the Akaike information criterion is specifically:

[0100]

[0101] The Bayesian information criterion is specifically:

[0102]

[0103] Where: is the number of parameters of the Gaussian mixture model under K components, representing the complexity of the model; M is the amount of data in the sample data set, i.e. the number of years; the comprehensive strategy value is the weighted sum of the Akaike information criterion and the Bayesian information standard.

[0104] A compromise strategy is adopted to select the model with the smallest score, taking into account the model's predictive performance, simplicity, and ability to prevent overfitting. :

[0105]

[0106] After the parameters in the Gaussian mixture model are optimized and determined, the conditional probability density function of variable x2 when variable x1 is known is , the mathematical expression is:

[0107]

[0108] Where: They are the conditional mean and conditional covariance of variable x2 when variable x1 is known, and the expressions are as follows:

[0109]

[0110] Where: and are the mean vectors of variables x1 and x2 respectively; is the covariance matrix between variables x1 and x1; is the covariance matrix between variables x2 and x2; and Consistent, is the covariance matrix between variables x1 and x2; is the covariance matrix of variable x2 when variable x1 is known.

[0111] Substitute the result of formula (13) into formula (8) to calculate the conditional distribution of variable x2 when variable x1 is known; when each scene is generated, obtain a random number between 0 and 1 ,make and Equal, that is, according to the conditional distribution of variable x2 when variable x1 is known and the conditional distribution value of variable x2 when variable x1 is known, variable x2 when variable x1 is known can be obtained, that is, the simulation value of variable x2 is the generated scenario.

[0112] This embodiment provides a method for generating a water-landscape scene driven by multiple correlations, aiming to solve the problems that the existing scene generation method only considers the generation of a single resource or the generation of a single correlation. The method includes the following steps:

[0113] Collect the historical data of three resources: water resource runoff h, wind resource wind power output w, and solar resource photovoltaic power output p, calculate the mutual relationship between any two of water, wind, and solar, and determine the order of generating water-wind-solar scenarios according to the calculation results. In this embodiment, it is assumed that w > p > h.

[0114] According to the order of scenario generation, first generate the scenario of the w variable. In this step, only the first-order autocorrelation relationship of the w variable is considered, that is, at this time, the sample data set of the Gaussian mixture model corresponding to the t time period is x1 = {w t-1}, x2 = {w t}. Optimize the model parameters according to the steps of S21 - S25 based on the sample data set, obtain the optimized Gaussian mixture model, and then use the model to generate scenarios.

[0115] Specifically: Substitute the sample data set x1 = {w t-1}, x2 = {w t} into the Gaussian mixture model corresponding to the t time period for training and optimizing the model, that is, determine the model parameters;

[0116] Use the above-trained and optimized model for scenario generation. The scenario data generated by the wind resource is denoted as . When generating the first data of the first year, that is, when m = 1 and t = 1, at this time, let the known variable x1 be equal to the average value of w t-1 corresponding to t = 1, that is, the average value of the existing variable information in the maximum period, to generate the simulated value of the first variable x2, that is, the first data of the first-year scenario; then let t = t + 1, and use the calculated as the new . At this time, the known variable x1 is equal to ; when the simulation of the Nth data of the first year is completed, that is, when t = N, the generation of the first-year scenario is completed, and the simulation of the first data of the next year is entered, that is, m = m + 1;

[0117] When simulating the first data of the new year, that is, , t = 1, let ; that is, when simulating the first data of the next year, let the known variable x1 be equal to the last generated data of the previous year, and then continue to generate the data of the new year until t is equal to the target time number N;

[0118] Repeat the generation of the scenario data of each year until m is equal to the target number of years to complete the generation of the scenario of the w variable.

[0119] Next, generate the scenario of the p variable according to the generation order of the scenarios. In this step, consider the first-order autocorrelation relationship of the p variable and the cross-correlation relationship between the p variable and the w variable. That is, at this time, in the sample dataset corresponding to the Gaussian mixture model at time t, x1 = {w t , p t-1}, that is, arrange the existing variable information of the first resource at time t and the existing variable information of the second resource at time t - 1 together in sequence as x1, x2 = {p t}. Optimize the model parameters according to the steps of S21 - S25 based on the sample dataset to obtain the optimized Gaussian mixture model, and then use the model to generate scenarios.

[0120] Specifically: Substitute the sample dataset x1 = {w t , p t-1}, x2 = {p t} into the Gaussian mixture model corresponding to time t for training and optimizing the model, that is, determine the model parameters;

[0121] Use the above-trained and optimized model for scenario generation. The scenario data generated by the optical resource is denoted as . When simulating the first data of the first year, that is, when m = 1, t = 1, at this time, let the known variable x1 be equal to , where is the average value of p t-1 corresponding to t = 1, that is, the average value of the existing variable information in the maximum period, is the generated value of the corresponding scenario of the first resource. Arrange and together in sequence as the simulated value of the variable x1 to generate the first data of the scenario in the first year, that is, the first variable x2; then let t = t + 1, and use the calculated as the new , and at this time, the known variable x1 is equal to ; when the simulation of the Nth data in the first year is completed, that is, when t = N, then the generation of the scenario in the first year is completed, and enter the simulation of the first data in the next year, that is, m = m + 1;

[0122] When simulating the first data in the new year, that is, , when t = 1, let ; that is, when simulating the first data in the next year, let the known variable x1 be equal to the data arranged in sequence of the generated value of the corresponding scenario of the first resource and the last generated data of the second resource in the previous year, and then continue to generate the data in the new year until t is equal to the target time number N;

[0123] Repeat the generation of the scenario data for each year until m is equal to the target number of years to complete the generation of the scenario of the p variable.

[0124] According to the order of scenario generation, the scenario that finally generates the h variable is considered. In this step, the first-order autocorrelation relationship of the h variable and the cross-correlation relationships between the h variable and the w variable and the p variable are considered. That is, the sample data set of the Gaussian mixture model corresponding to the t period at this time has x1 = {w t , p t , h t-1}. That is, the existing variable information of the first resource in the t period, the existing variable information of the second resource in the t period, and the existing variable information of the third resource in the (t - 1) period are arranged in sequence as x1, and x2 = {h t}. According to the sample data set, optimize the model parameters according to the steps of S21 - S25 to obtain the optimized Gaussian mixture model, and then use the model to generate scenarios.

[0125] Specifically: Substitute the sample data set x1 = {w t , p t , h t-1} and x2 = {h t} into the Gaussian mixture model corresponding to the t period for training and optimizing the model, that is, determining the model parameters;

[0126] Use the above-trained and optimized model for scenario generation. The scenario data generated by water resources is denoted as . When simulating the first data of the first year, that is, when m = 1 and t = 1, at this time, let the known variable x1 be equal to , where is the average value of h t-1 corresponding to t = 1, that is, the average value of the existing variable information in the maximum period, is the generated value of the corresponding scenario of the first resource, is the generated value of the corresponding scenario of the second resource. Arrange , and in sequence as the variable x1 to generate the simulated value of the first variable x2, that is, the first data of the scenario in the first year; then let t = t + 1, and use the calculated as the new . At this time, the known variable x1 is equal to ; when the simulation of the Nth data in the first year is completed, that is, when t = N, the generation of the scenario in the first year is completed, and the simulation of the first data in the next year is entered, that is, m = m + 1;

[0127] When simulating the first data in the new year, that is, , when t = 1, let That is, in the first data simulation of the next year, the known variable x1 is equal to the data arranged in sequence of the generated value of the scenario corresponding to the first resource, the generated value of the scenario corresponding to the second resource, and the last generated data of the third resource in the previous year, and then the generation of data in the new year continues until t is equal to the target time number N;

[0128] Repeat the generation of scenario data for each year until m is equal to the target number of years, and complete the generation of the scenario of the h variable.

[0129] Furthermore, Embodiment 2 of the present invention provides a water-wind-solar scenario generation system driven by multiple correlations and associations. The system includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it executes the water-wind-solar scenario generation method described in any one of the above embodiments.

[0130] Taking the daily-scale historical hydrological data of the Xiluodu Hydropower Station and the daily-scale historical wind speeds and photovoltaic intensities in the planned installed capacity areas of wind power and photovoltaic in its nearby areas as the research objects, the historical wind speeds and photovoltaic intensities are converted into the corresponding wind power outputs and photovoltaic outputs according to the planned installed capacity. The time series of runoff, wind power output, and photovoltaic output all cover the period from 1940 to 2021. According to the above time series, the water-wind-solar scenario generation is carried out by using the method of the present invention, and 1000-year scenario sequences are generated for each. During the implementation process, statistical indicators (mean, standard deviation, skewness coefficient, kurtosis coefficient), time correlation indicators (Pearson correlation coefficient), and spatial correlation indicators (average absolute error of Kendall correlation coefficient) are used to evaluate the statistical characteristics and multiple correlations of the generated scenarios.

[0131] Tables 1 to 3 respectively show the root mean square error (RMSE), mean absolute percentage error (MAPE), and mean absolute error (MAE) between the statistical characteristics and correlation coefficients of the daily-scale water, wind, and light scenarios generated by the method of the present invention, the two-dimensional Copula method, and the seasonal autoregressive model method (SAR method), and the statistical characteristics and correlation coefficients of the historical measured daily-scale water-wind-solar scenarios. Table 4 shows the spatial correlation indicators between the daily-scale wind-solar, water-wind, and water-light scenarios generated by these three methods. The smaller the values of the root mean square error, mean absolute percentage error, and mean absolute error, the higher the fitting degree of the generated scenarios to the historical measured scenarios in terms of statistical characteristics and autocorrelation; the smaller the spatial correlation indicator, the more accurately the simulation sequence describes the mutual correlation characteristics between various resources. In addition, the logarithmic values in Tables 1 to 4 are bolded to identify the best results in each method;

[0132] ;

[0133] ;

[0134] ;

[0135] 。

[0136] In the comparison method, the two-dimensional Copula method only considers the mutual correlation between resources, while the SAR method only focuses on the first-order autocorrelation of a single resource. As can be seen from Tables 1 to 3, the method of the present invention can more comprehensively describe the statistical characteristics of historical measured scenarios, and its simulation effects in terms of statistical features and first-order autocorrelation are better than those of the other two comparison methods. In addition, as can be seen from Table 4, the method of the present invention also shows a better effect in simulating the mutual correlation between wind-solar, water-wind, and water-solar sequences, significantly better than the two-dimensional Copula method and the SAR method.

[0137] It is easy for those skilled in the art to understand that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for generating water-landscape scenes driven by multiple correlations, characterized in that: include: Offline training phase: S1, collect historical data on water resource runoff, wind resource wind power output, and light resource photovoltaic output, calculate the relationship between water, wind, and light resources based on the historical data, and determine the order of generating water, wind, and light scenes based on the relationship, which are the first resource, the second resource, and the third resource in order of priority; and extract the existing variable information of each resource in the t-1 period and the t period based on the historical data; S2, construct Gaussian mixture models corresponding to the three resources respectively: construct the two-dimensional joint distribution of variables x1 and variables x2 and the conditional distribution of variable x2 when variable x1 is known through the Gaussian mixture model, and then generate the simulated value of variable x2 according to variable x1, and establish a sample data set according to the existing variable information to optimize the parameters of the Gaussian mixture model corresponding to each resource, and obtain the Gaussian mixture model with optimized parameters corresponding to each resource; For the first resource, the existing variable information in the time period t-1 is made to correspond to the variable x1, and the existing variable information in the time period t is made to correspond to the variable x2, and the Gaussian mixture model parameters are optimized; For any resource among the second resource and the third resource, the existing variable information of the time period t of the resource ranked before the any resource and the existing variable information of the time period t-1 of the any resource are combined and corresponded to the variable x1, the existing variable information of the time period t of the any resource is corresponded to the variable x2, and the Gaussian mixture model parameters corresponding to the any resource are optimized; Online scene generation phase: S3, using the Gaussian mixture model with optimized parameters corresponding to each resource to generate scenes for the corresponding resources.

2. The method for generating water-landscape scenes driven by multiple correlation joint driving according to claim 1, characterized in that: In S2, a sample data set is established based on the existing variable information to optimize the Gaussian mixture model parameters corresponding to each resource, specifically including: For the second resource, the existing variable information of the first resource in the t period and the existing variable information of the second resource in the t-1 period are arranged in sequence and correspond to the variable x1, and the existing variable information of the second resource in the t period is corresponded to the variable x2, and the Gaussian mixture model parameters are optimized; For the third resource, the existing variable information of the first resource in time period t, the existing variable information of the second resource in time period t, and the existing variable information of the third resource in time period t-1 are arranged in sequence and correspond to the variable x1, and the existing variable information of the third resource in time period t is corresponded to the variable x2, and the Gaussian mixture model parameters are optimized.

3. The method for generating water-landscape scenes driven by multiple correlation joint driving according to claim 1, characterized in that: S3 specifically includes: For the first resource, the average value of the existing variable information in the maximum period is used as variable x1 to generate the first simulation value of variable x2, and the first simulation value of variable x2 is used as the first generated value of the scenario. Then, the previous generated value is used as the new variable x1 to generate the current variable x2; For the second resource, the first generated value of the first resource is combined with the average value of the existing variable information of the second resource's maximum period as variable x1 to generate the first simulated value of variable x2. The first simulated value of variable x2 is used as the first generated value of the scenario. Then, the current generated value of the first resource is combined with the previous generated value of the second resource as the new variable x1 to generate the current variable x2. For the third resource, the first generated value of the first resource, the first generated value of the second resource and the average value of the existing variable information of the third resource's maximum time period are combined as variable x1 to generate the first simulated value of variable x2. The first simulated value of variable x2 is used as the first generated value of the scene. Then, the current generated value of the first resource, the current generated value of the second resource and the previous generated value of the third resource are combined as the new variable x1 to generate the current variable x2.

4. The method for generating water-landscape scenes driven by multiple correlation joint driving according to claim 1, characterized in that: In S1, the relationship between water, wind and light resources is calculated based on historical data, and the order of generating water, wind and light scenes is determined based on the relationship, including: Calculate the mutual information between the water, wind and light resources based on historical data, use the mutual information to determine the relationship between the water, wind and light resources, and obtain the mutual information value between the water, wind and light resources; Calculate the sum of the mutual information values ​​between any resource and other resources, obtain the comprehensive mutual information index of any resource, and determine the order of generating the three resource scenarios according to the comprehensive mutual information index.

5. The method for generating water-landscape scenes driven by multiple correlation joint driving according to claim 1, characterized in that: In S1, according to historical data, the existing variable information of each resource in period t-1 and period t is extracted as follows: ； ； ； Among them, DH is the multi-year historical data of water resource runoff h, DW is the multi-year historical data of wind resource wind power output w, and DP is the multi-year historical data of light resource photovoltaic output p; the existing variable information of the three resources of water, wind and light in the t-1 period and the t period are h t-1 With h t 、w t-1 With w t 、p t-1 With p t ; , N is determined according to the target time scale.

6. The method for generating water-landscape-light scenes by multiple correlation joint driving according to claim 5, characterized in that: In S2, a sample data set is established based on the existing variable information to optimize the Gaussian mixture model parameters corresponding to each resource, and the Gaussian mixture model after the parameter optimization corresponding to each resource is obtained, which specifically includes: For any resource, variables x1 and x2 change with the increase of time period t. The existing variable information of each time period t is used to optimize the parameters of the Gaussian mixture model, and N models with optimized parameters corresponding to any resource are obtained for scene generation of N time periods for any resource.

7. The method for generating water-landscape scene with multiple correlation joint drive according to any one of claims 1 to 6, characterized in that: The Gaussian mixture model in S2 is specifically: ； ； Where: is the two-dimensional joint distribution of variables x1 and x2, is the conditional distribution of variable x2 when variable x1 is known, is the binary probability density function, is the conditional probability density function of variable x2 when variable x1 is known; The bivariate probability density function , the mathematical expression is: ； Where: represents the binary Gaussian probability density function; K is the number of components of the Gaussian mixture model, which is a hyperparameter; is the weight of the kth component; are the mean vector and covariance matrix of the kth component respectively.

8. The method for generating water-landscape scenes driven by multiple correlation joint driving according to claim 7, characterized in that: In S2, the Gaussian mixture model parameters corresponding to each resource are optimized, specifically, the hyperparameters K, weights in the Gaussian mixture model are optimized. , mean vector and the covariance matrix Optimize and determine, the specific optimization steps include: Pre-set several different hyperparameter K values; For any preset K value, the sample data set is divided into K clusters using clustering methods; the center of each cluster is used as the mean vector of each component in the Gaussian mixture model ; Calculate the difference between the data points in each cluster and the cluster center to obtain the covariance matrix ; Calculate the proportion of the number of samples in each cluster to the total number of samples and obtain the weight ; The obtained parameters are used as initial values, and the expectation maximization algorithm is used to optimize the model parameters to obtain the mean vector, covariance matrix and weight parameter value of the model after optimization corresponding to any preset K value; The comprehensive strategy of Akaike information criterion and Bayesian information criterion is used to evaluate the Gaussian mixture model under different preset K values, and the preset K value corresponding to the Gaussian mixture model with the smallest comprehensive strategy value is selected as the optimal K value; The Kolmogorov-Smilov test was performed on the theoretical distribution of the Gaussian mixture model corresponding to the optimal K value and the empirical distribution of the sample data set, with a significance level of Set to 0.

2. If the test passes, the model training is completed. If the test fails, adjust the preset K value and re-optimize the model parameters.

9. The method for generating water-landscape-light scenes by multiple correlation joint driving according to claim 8, characterized in that: Akaike Information Criterion Specifically: ； Bayesian Information Criterion Specifically: ； Where: is the number of parameters of the Gaussian mixture model under K components, representing the complexity of the model; M is the amount of data in the sample data set; the comprehensive strategy value is the weighted sum of the Akaike information criterion and the Bayesian information criterion.

10. A water-landscape scene generation system driven by multiple correlations, characterized in that: The system includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method for generating water, landscape and light scenes driven by multiple correlation joint drives described in any one of claims 1 to 9 is executed.

Images