A method for partitioning offshore multi-energy complementary resources based on C-Vine Copula and K-Means clustering

Abstract

The application belongs to the technical field of offshore multi-energy complementary resource partitioning, and provides an offshore multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering, comprising: acquiring wind speed, wave height and solar radiation data, fitting the edge distribution after preprocessing, and generating three optimal edge distribution models; constructing a three-dimensional C-Vine Copula joint distribution model with wind speed as the root node, and quantifying the dependence relationship of the three resources; extracting the resource combination characteristics under the joint probability, and generating the spatial grid feature vector; using K-Means clustering to partition each spatial grid, and generating the resource combination partitioning result; combining the water depth data to evaluate the development potential of each partition, and outputting the final partitioning result. The application can represent the nonlinear dependence relationship between multiple resources, and take into account the water depth constraint, thereby improving the accuracy and practicability of the multi-energy complementary development potential evaluation.

Description

Technical Field

[0001] This application belongs to the field of marine multi-energy complementary resource zoning technology, and particularly relates to a marine multi-energy complementary resource zoning method based on C-Vine Copula and K-Means clustering. Background Technology

[0002] The complementary development of multiple energy sources, such as offshore wind power, wave power, and offshore photovoltaic power, is an important direction for the utilization of marine energy. Wind speed, wave height, and solar radiation resources vary significantly in their spatiotemporal distribution across different sea areas, and these three resources exhibit complex dependencies. Existing marine resource zoning methods typically assess each resource type independently, such as dividing suitable areas for wind farms based on wind speed distribution or wave energy development areas based on wave height distribution; a few studies involving multi-energy complementarity use simple resource overlay or weighted scoring methods for zoning.

[0003] However, the above methods failed to effectively characterize the nonlinear dependence among wind speed, wave height, and solar radiation, resulting in the multi-energy complementary resource zoning results failing to truly reflect the differences in resource combination characteristics in different sea areas, thus affecting the accuracy of subsequent development potential assessments. Summary of the Invention

[0004] This application provides a method, system, terminal equipment, and storage medium for marine multi-energy complementary resource partitioning based on C-Vine Copula and K-Means clustering. It can solve the problem that current methods fail to effectively characterize the nonlinear dependence between wind speed, wave height, and solar radiation, resulting in partitioning results that cannot truly reflect resource combination characteristics and affecting the accuracy of subsequent development potential assessment.

[0005] In a first aspect, embodiments of this application provide a method for marine multi-energy complementary resource partitioning based on C-Vine Copula and K-Means clustering, including: S1, acquiring three types of resource data for a target sea area, including wind speed resource data, wave height resource data, and solar radiation resource data; preprocessing the three types of resource data and fitting the marginal distributions of the three types of resource data respectively; estimating the model parameters of the probability distribution model corresponding to each resource data; and generating optimal marginal distribution models for wind speed, wave height, and solar radiation; S2, based on the optimal marginal distribution models for wind speed, wave height, and solar radiation, constructing a three-dimensional C-Vine Copula joint probability distribution model with wind speed resource as the root node; the three-dimensional C-Vine Copula joint probability distribution model is used to quantify the dependency relationship between the three types of resources; S3, based on the three-dimensional C-Vine Copula... The Copula joint probability distribution model extracts resource combination feature data under the target joint probability conditions and generates a resource combination feature vector; S4, using the resource combination feature vector as input, the K-Means clustering algorithm is used to perform cluster analysis on each spatial grid point of the target sea area to generate resource combination partitioning results, which are used to classify different resource combination types; S5, the resource combination partitioning results and the corresponding water depth data for each partition are obtained, and the multi-energy complementary development potential of each partition is determined based on the water depth data to generate the final resource partitioning results.

[0006] In one possible implementation of the first aspect, the marginal distributions of the three resource data are fitted in S1 above, the model parameters of the probability distribution model corresponding to each resource data are estimated, and the optimal marginal distribution model for wind speed, wave height, and solar radiation are generated, including:

[0007] For each type of resource data, multiple candidate marginal distribution models are used for fitting, and the model parameters of each candidate marginal distribution model are estimated. The goodness-of-fit value of each candidate marginal distribution model is calculated. The candidate marginal distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data, thus obtaining the optimal marginal distribution model for wind speed, wave height, and solar radiation.

[0008] Optionally, in another possible implementation of the first aspect, for each type of resource data, multiple candidate marginal distribution models are used for fitting, and the goodness-of-fit value of each candidate marginal distribution model is calculated; the candidate marginal distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data, resulting in the optimal marginal distribution model for wind speed, wave height, and solar radiation, including:

[0009] Based on wind speed time series data in wind speed resource data, the Weibull distribution model is used for parameter fitting, the goodness-of-fit value of the Weibull distribution model is calculated, and the Weibull distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for wind speed.

[0010] Based on the wave height time series data in the wave height resource data, the Gamma distribution model is used for parameter fitting, the goodness of fit value of the Gamma distribution model is calculated, and the Gamma distribution model with the smallest goodness of fit value is selected as the optimal marginal distribution model for wave height.

[0011] Based on solar radiation time series data in solar radiation resource data, a nonlinear mixed-weight zero-inflation Beta distribution model is used for parameter fitting. The goodness-of-fit value of the nonlinear mixed-weight zero-inflation Beta distribution model is calculated, and the nonlinear mixed-weight zero-inflation Beta distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model of solar radiation.

[0012] Optionally, in another possible implementation of the first aspect, the above-mentioned S2, based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, constructs a three-dimensional C-Vine Copula joint probability distribution model with wind speed resource as the root node, including:

[0013] S21. Based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, various Copula functions are used to fit the three resource pairs of wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation respectively. The optimal pairwise Copula functions are selected for each resource pair through the preset screening criteria.

[0014] S22. Taking wind speed resource as the root node, the optimal pairwise Copula functions of wind speed and wave height resource pairs and the optimal pairwise Copula functions of wind speed and solar radiation resource pairs are used as the first layer connection structure, and the optimal pairwise Copula functions of wave height and solar radiation resource pairs are used as the second layer connection structure to construct a three-dimensional C-Vine Copula joint probability distribution model.

[0015] Optionally, in another possible implementation of the first aspect, in S21 above, based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, various Copula functions are used to fit the three resource pairs of wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation, respectively. The optimal pairwise Copula functions are selected for each resource pair according to a preset screening criterion, including:

[0016] For wind speed and wave height resource pairs, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the Gumbel Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for wind speed and wave height resource pairs after comparing the information criterion values.

[0017] For wind speed and solar radiation resource pairs, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for wind speed and solar radiation resource pairs after comparing the information criterion values.

[0018] For the pair of wave height and solar radiation resources, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for the pair of wave height and solar radiation resources.

[0019] Optionally, in another possible implementation of the first aspect, S3 above, based on the three-dimensional C-VineCopula joint probability distribution model, extracts resource combination feature data under the target joint probability condition and generates a resource combination feature vector, including:

[0020] S31. Based on the three-dimensional C-Vine Copula joint probability distribution model, calculate the resource combination feature data on the equiprobability surface corresponding to the preset joint probability value.

[0021] S32. Extract the resource combination feature data of each spatial grid point on the equiprobability surface, and combine them to form a resource combination feature vector.

[0022] Optionally, in another possible implementation of the first aspect, S4 above, using the resource combination feature vector as input, employs the K-Means clustering algorithm to perform cluster analysis on each spatial grid point of the target sea area, generating resource combination partitioning results, including:

[0023] S41. Obtain the resource combination feature vector of each spatial grid point to form a feature vector set;

[0024] S42. Randomly select a preset number of feature vectors from the feature vector set as initial cluster centers;

[0025] S43. Calculate the distance between the feature vector of each spatial grid point and each initial cluster center, and assign each spatial grid point to the cluster to which the initial cluster center with the smallest distance belongs;

[0026] S44. Calculate the mean of all feature vectors within each cluster, and update the mean to the new cluster center of the corresponding cluster;

[0027] S45. Repeat S43 and S44 until the cluster centers of each cluster no longer change. Output the cluster label to which each spatial grid point belongs and generate the resource combination partitioning result.

[0028] Optionally, in another possible implementation of the first aspect, S5 above, obtaining the resource combination partitioning results and the water depth data corresponding to each partition, determining the multi-energy complementary development potential of each partition based on the water depth data, and generating the final resource partitioning results, includes:

[0029] S51. Obtain the resource combination partitioning results and the water depth data for each spatial grid point;

[0030] S52. For each partition in the resource combination partitioning result, statistically analyze the water depth data distribution characteristics of all spatial grid points within the partition.

[0031] S53. Calculate the proportion of the number of spatial grid points in each partition whose water depth data is within the preset exploitable water depth range to the total number of spatial grid points in the partition, and mark the partitions with a proportion greater than the preset proportion threshold as priority development areas.

[0032] S54. Output the final resource partitioning results, including priority development area markers.

[0033] Secondly, embodiments of this application provide a marine multi-energy complementary resource zoning system based on C-Vine Copula and K-Means clustering, comprising: a data acquisition and edge distribution fitting module, used to acquire three types of resource data for a target sea area, including wind speed resource data, wave height resource data, and solar radiation resource data; preprocessing the three types of resource data and fitting the edge distributions of the three types of resource data respectively; estimating the model parameters of the probability distribution models corresponding to each resource data; and generating optimal edge distribution models for wind speed, wave height, and solar radiation; a three-dimensional C-Vine Copula joint modeling module, used to construct a three-dimensional C-Vine Copula joint probability distribution model with wind speed resource as the root node based on the optimal edge distribution models for wind speed, wave height, and solar radiation; and a resource combination feature extraction module, used to extract features based on the three-dimensional C-Vine Copula... The Copula joint probability distribution model extracts resource combination feature data under the target joint probability conditions and generates resource combination feature vectors. The K-Means clustering partitioning module uses the resource combination feature vectors as input and employs the K-Means clustering algorithm to perform clustering analysis on each spatial grid point of the target sea area to generate resource combination partitioning results. The resource combination partitioning results are used to classify different resource combination types. The partitioning potential assessment module obtains the resource combination partitioning results and the corresponding water depth data for each partition. Based on the water depth data, it determines the multi-energy complementary development potential of each partition and generates the final resource partitioning results.

[0034] Thirdly, embodiments of this application provide a terminal device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements a marine multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering as described above.

[0035] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned method for partitioning marine multi-energy complementary resources based on C-Vine Copula and K-Means clustering.

[0036] Beneficial Effects: This application acquires wind speed, wave height, and solar radiation resource data, fits marginal distributions to each, and generates three optimal marginal distribution models, providing accurate input for joint probability modeling. Based on the three optimal marginal distribution models, a three-dimensional C-Vine Copula joint probability distribution model with wind speed as the root node is constructed, accurately depicting the nonlinear dependencies and tail correlations among the three resources, overcoming the shortcomings of traditional methods in characterizing the nonlinear interdependence of multiple resources. Resource combination feature data under preset joint probability conditions are extracted from this model to generate resource combination feature vectors for each spatial grid point, ensuring a unified measurement standard for resource combination features across different grid points in the joint probability space, eliminating the uncertainty of dimensional differences and subjective weighting inherent in traditional methods. Using the feature vectors as input, the K-Means clustering algorithm is used to perform unsupervised clustering of each spatial grid point, generating resource combination partitioning results, avoiding the subjectivity of manually set boundaries. Water depth data for each partition is acquired, and partitions that meet engineering water depth conditions are marked as priority development areas, ensuring that the partitioning results possess both resource complementarity suitability and engineering feasibility, providing reliable technical support for site selection of offshore multi-energy complementary projects. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart illustrating a method for partitioning marine multi-energy complementary resources based on C-Vine Copula and K-Means clustering, provided in an embodiment of this application.

[0039] Figure 2 This is a schematic diagram of the structure of the terminal device provided in the embodiments of this application;

[0040] Figure 3 This is a schematic diagram of the structure of a marine multi-energy complementary resource zoning system based on C-Vine Copula and K-Means clustering, provided in one embodiment of this application. Detailed Implementation

[0041] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0042] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0043] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0044] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0045] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0046] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0047] The following description, with reference to the accompanying drawings, details a method, system, terminal equipment, and storage medium for marine multi-energy complementary resource partitioning based on C-Vine Copula and K-Means clustering, as provided in this application.

[0048] Figure 1 The illustration shows a flowchart of a marine multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering provided in an embodiment of this application.

[0049] like Figure 1As shown, this method for partitioning marine multi-energy complementary resources based on C-Vine Copula and K-Means clustering includes the following steps:

[0050] S1. Obtain three types of resource data for the target sea area, including wind speed resource data, wave height resource data, and solar radiation resource data. Preprocess the three types of resource data and fit the marginal distribution of each type of resource data respectively. Estimate the model parameters of the probability distribution model corresponding to each resource data and generate the optimal marginal distribution model of wind speed, the optimal marginal distribution model of wave height, and the optimal marginal distribution model of solar radiation.

[0051] As one possible approach, wind speed resource data can be derived from meteorological observation stations within the target sea area, satellite remote sensing inversion data, or time-series data output from numerical weather prediction models. Wave height resource data can be derived from wave buoy observations, ocean wave model simulation data, or satellite altimeter inversion data. Solar radiation resource data can be derived from radiation observation stations, solar radiation products retrieved from satellite cloud images, or downward shortwave radiation data from reanalysis data.

[0052] As one possible implementation, the preprocessing described above includes missing value handling, outlier removal, and time series alignment. Missing value handling can be achieved using linear interpolation or the mean of adjacent time points. Outlier removal can employ a standard deviation threshold-based method, marking data points exceeding a certain standard deviation from the mean as outliers and removing them. Time series alignment refers to unifying the three types of resource data to the same time resolution and the same time points, ensuring that subsequent joint analysis simultaneously provides observations of wind speed, wave height, and solar radiation at each moment.

[0053] In this embodiment of the application, fitting of the marginal distribution refers to selecting a suitable probability distribution model and estimating the model parameters for each type of resource data, so that the probability distribution model can describe the statistical regularity of the resource data itself.

[0054] Furthermore, in this embodiment of the application, the step S1 above, which involves fitting the marginal distributions of the three types of resource data, estimating the model parameters of the probability distribution model corresponding to each resource data, and generating the optimal marginal distribution model for wind speed, wave height, and solar radiation, includes:

[0055] S11. For each type of resource data, multiple candidate marginal distribution models are used for fitting, and the model parameters of each candidate marginal distribution model are estimated; the goodness-of-fit value of each candidate marginal distribution model is calculated; the candidate marginal distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data, thus obtaining the optimal marginal distribution model for wind speed, wave height, and solar radiation.

[0056] As one possible implementation, the above calculation of the goodness-of-fit value of each candidate marginal distribution model can be achieved by calculating the AIC value and selecting the candidate marginal distribution model with the smallest AIC value as the optimal marginal distribution model for the corresponding resource data. This application does not limit this approach.

[0057] As one possible implementation, candidate marginal distribution models can be pre-selected based on the characteristics of the resource data. For wind speed resource data, candidate models can include Weibull distribution, Rayleigh distribution, log-normal distribution, and gamma distribution. For wave height resource data, candidate models can include gamma distribution, Weibull distribution, log-normal distribution, and generalized extreme value distribution. For solar radiation resource data, since solar radiation data often contains a large number of zero values ​​or near-zero values, candidate models can include beta distribution, zero-inflated beta distribution, mixed beta distribution, and nonlinear mixed-weight zero-inflated beta distribution.

[0058] As one possible implementation, the goodness-of-fit value can be calculated using the Kolmogorov-Smirnov test statistic, the Anderson-Darling test statistic, or the Akaike information criterion. A smaller goodness-of-fit value indicates a better model fit. Alternatively, the negative log-likelihood value or the Bayesian information criterion can be used, selecting the model with the smallest value as the optimal model. From all candidate marginal distribution models, the candidate marginal distribution model with the best goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data.

[0059] Furthermore, in the embodiments of this application, the above-mentioned S11 specifically includes:

[0060] S111. Based on the wind speed time series data in the wind speed resource data, the Weibull distribution model is used for parameter fitting, the goodness-of-fit value of the Weibull distribution model is calculated, and the Weibull distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for wind speed.

[0061] Among them, wind speed time series data are continuous time observations obtained from the target sea area, such as wind speed values ​​recorded once per hour.

[0062] The Weibull distribution is a continuous probability distribution whose probability density function includes shape and scale parameters. The shape parameter controls the skewness of the distribution curve, while the scale parameter controls the dispersion of the wind speed values.

[0063] In one embodiment, the maximum likelihood estimation method or the method of moments is used to estimate the two parameters of the Weibull distribution. Maximum likelihood estimation is performed by constructing a likelihood function and finding the parameter value that maximizes the likelihood function. After parameter estimation, a set of specific shape and scale parameter values ​​are obtained. These values ​​are then substituted into the probability density function or cumulative distribution function expression of the Weibull distribution to form the optimal marginal distribution model of wind speed for the target sea area's wind speed resource data.

[0064] In the embodiments of this application, the Weibull distribution is suitable for describing natural phenomena such as wind speed, which have non-negative values ​​and a right-skewed distribution. It can accurately capture the high-frequency characteristics of low-wind-speed areas and the distribution pattern of the tail.

[0065] S112. Based on the wave height time series data in the wave height resource data, the Gamma distribution model is used for parameter fitting, the goodness-of-fit value of the Gamma distribution model is calculated, and the Gamma distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for wave height.

[0066] Among them, wave height time series data are usually derived from the effective wave height values ​​output by wave buoys or wave models.

[0067] The Gamma distribution is a two-parameter continuous probability distribution, containing shape and scale parameters, or equivalently shape and rate parameters. The domain of the gamma distribution is non-negative real numbers, making it suitable for describing statistical characteristics such as wave height, which exhibits non-negative values, high frequency of small wave heights, and low frequency of large wave heights.

[0068] In one embodiment, parameter estimation can employ the maximum likelihood estimation method, which uses an iterative algorithm to find the maximum value of the likelihood function to obtain the optimal estimates of the shape and scale parameters. Substituting the estimated parameter values ​​into the cumulative distribution function expression of the gamma distribution yields the optimal marginal distribution model of wave height that characterizes the cumulative probability law of wave height data in the sea area.

[0069] In the embodiments of this application, the gamma distribution can flexibly adjust its shape to adapt to the kurtosis and skewness characteristics of wave height distribution in different sea areas.

[0070] S113. Based on the solar radiation time series data in the solar radiation resource data, a nonlinear mixed weight zero-inflation Beta distribution model is used for parameter fitting. The goodness-of-fit value of the nonlinear mixed weight zero-inflation Beta distribution model is calculated, and the nonlinear mixed weight zero-inflation Beta distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model of solar radiation.

[0071] It should be noted that solar radiation time series data have significant characteristics: during nighttime or periods of severe cloud cover, solar radiation values ​​are close to zero; during clear periods, solar radiation exhibits a continuous positive distribution; the positive values ​​are often distributed within a unit interval or, after normalization, are distributed between zero and one.

[0072] In one embodiment, the aforementioned nonlinear hybrid weighted zero-inflation Beta distribution model is designed specifically for the characteristics of solar radiation time series data. This model comprises three components: a discrete point mass distribution describing the probability of zero solar radiation, a beta distribution describing the continuous distribution of solar radiation over open intervals, and a nonlinear hybrid weighting function regulating the probability allocation between zero and positive values. The nonlinear hybrid weights can be logistic functions or other nonlinear functions with multiple parameters. Parameter estimation employs the expectation-maximization algorithm or the Markov chain Monte Carlo method, iteratively optimizing to obtain the zero-inflation parameters, the two shape parameters of the beta distribution, and the parameters of the hybrid weighting function. Integrating these parameter values ​​yields the optimal marginal distribution model of solar radiation.

[0073] In the embodiments of this application, the nonlinear mixed-weight zero-inflation Beta distribution model can simultaneously and accurately fit the zero-value accumulation phenomenon and the distribution pattern of the continuous positive value part in solar radiation, and has a stronger expressive power than the ordinary beta distribution or the zero-inflation beta distribution.

[0074] S2. Based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, a three-dimensional C-Vine Copula joint probability distribution model with wind speed resource as the root node is constructed. The three-dimensional C-Vine Copula joint probability distribution model is used to quantify the dependency relationship between the three resources.

[0075] The C-Vine Copula is a vine-structured copula suitable for scenarios where there are primary and secondary relationships between variables. With wind speed as the root node, it represents wind speed as the core variable, forming a first-level dependency with wave height and solar radiation. The dependency between wave height and solar radiation, given wind speed, is characterized by a second-level dependency. The three-dimensional C-Vine Copula joint probability distribution model is used to quantify the dependencies among wind speed, wave height, and solar radiation data.

[0076] In this embodiment, the marginal distribution functions of three resources are used as input. Appropriate binary Copula functions are selected to describe the correlations between wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation, respectively. These are then nested and combined according to a C-Vine structure to obtain a joint probability distribution model. Wind speed, wave height, and solar radiation are physically related; for example, wind speed drives wave generation, and clouds influence solar radiation while also being associated with wind speed changes. These relationships cannot be accurately described by the assumption of independence or simple linear relationships. The C-Vine Copula can decompose the multivariate joint distribution into the product of multiple binary Copulas, thereby flexibly capturing specific dependency patterns between each pair of variables.

[0077] Furthermore, in this embodiment of the application, step S2 includes:

[0078] S21. Based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, various Copula functions are used to fit the three resource pairs of wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation respectively. The optimal pairwise Copula functions are selected for each resource pair through the preset screening criteria.

[0079] In one embodiment, based on the optimal marginal distribution models for wind speed, wave height, and solar radiation, various Copula functions are used to fit three resource pairs: wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation. Copula functions connect a multivariate joint distribution to its respective marginal distribution. Commonly used families of binary Copula functions include the Gaussian Copula and t-Copula functions in the elliptic family, and the Gumbel Copula, Clayton Copula, and Frank Copula functions in the Archimedes family. For each resource pair, the two sets of resource data are transformed by their respective marginal distribution cumulative distribution functions to obtain a uniformly distributed data sequence. Then, each Copula function is used to fit the data, estimating the parameters of the Copula function. Parameter estimation methods can include maximum likelihood estimation or semi-parametric methods. After fitting all candidate Copula functions, the optimal pairwise Copula functions for each resource pair are selected according to preset screening criteria. Screening criteria can include the Akaike information criterion, the Bayesian information criterion, or the likelihood ratio test. By comparing the information criterion values ​​of different Copula functions, the Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for the resource pair.

[0080] S22. Taking wind speed resource as the root node, the optimal pairwise Copula functions of wind speed and wave height resource pairs and the optimal pairwise Copula functions of wind speed and solar radiation resource pairs are used as the first layer connection structure, and the optimal pairwise Copula functions of wave height and solar radiation resource pairs are used as the second layer connection structure to construct a three-dimensional C-Vine Copula joint probability distribution model.

[0081] In one embodiment, wind speed resource is used as the root node, and the optimal pairwise Copula functions for wind speed and wave height resource pairs and the optimal pairwise Copula functions for wind speed and solar radiation resource pairs are used as the first layer of connection structure. In the C-Vine Copula structure, the first layer contains binary connections with the root node as the core, and the root node and each of the other variables constitute a binary Copula. In this step, the root node is selected as wind speed resource, so the first layer includes binary Copula functions for wind speed and wave height, and binary Copula functions for wind speed and solar radiation. The optimal pairwise Copula functions for wave height and solar radiation resource pairs are used as the second layer of connection structure. In the second layer of the C-Vine Copula, based on the conditions processed in the first layer, a conditional Copula function for wave height and solar radiation under given wind speed resource conditions is constructed. The edge inputs of this conditional Copula function are the conditional distributions of wave height resource and solar radiation resource under given wind speed. The optimal pairwise Copula functions for wave height and solar radiation resource pairs are used to describe this conditional dependency. Multiplying the binary Copula functions of the first and second layers according to the tree structure of C-Vine, and then multiplying them by the three marginal distribution functions, yields the complete three-dimensional C-Vine Copula joint probability distribution model. The joint probability density function of this model can be expressed as the product of the three marginal probability density functions multiplied by the density functions of the two binary Copula functions of the first layer, and then multiplied by the density function of the single binary Copula function of the second layer.

[0082] In this embodiment, the optimal binary Copula function is selected hierarchically and organized according to a vine structure rooted in wind speed to achieve refined modeling of the dependencies between the three resources.

[0083] Furthermore, in this embodiment of the application, step S21 includes:

[0084] S211. For wind speed and wave height resource pairs, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the Gumbel Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for wind speed and wave height resource pairs after comparing the information criterion values.

[0085] S212. For the wind speed and solar radiation resource pair, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for the wind speed and solar radiation resource pair after comparing the information criterion values.

[0086] S213. For the pair of wave height and solar radiation resources, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for the pair of wave height and solar radiation resources.

[0087] It should be noted that in step S211, the wind speed and wave height resource pairs are fitted using the Gumbel Copula function. The Gumbel Copula function belongs to the Archimedes family of Copulas, and its generating function has a specific mathematical form. A key characteristic of the Gumbel Copula function is that the upper tail correlation coefficient is non-zero while the lower tail correlation coefficient is zero, meaning it can capture the dependency relationship when both variables simultaneously take large values, but is insensitive to the dependency relationship when both take small values. The wind speed resource data is transformed into a uniform distribution sequence using the cumulative distribution function of the optimal marginal distribution model for wind speed, and the wave height resource data is transformed into a uniform distribution sequence using the cumulative distribution function of the optimal marginal distribution model for wave height. Then, the parameters of the Gumbel Copula function are estimated. The parameter estimation method uses maximum likelihood estimation to obtain parameter values ​​describing the strength of the dependency between wind speed and wave height. After fitting, the Gumbel Copula function outperforms other candidate Copula functions under the preset screening criteria, and is therefore selected as the optimal pairwise Copula function for the wind speed and wave height resource pairs. This choice is consistent with the actual laws of the marine environment in a physical sense: strong winds are often accompanied by large waves, while wave height may not be low under weak wind conditions due to the influence of distant swells. Therefore, the dependence between wind speed and wave height is mainly manifested as upper tail correlation, and the Gumbel Copula function is just right to characterize this asymmetric tail dependency structure.

[0088] It should be noted that in step S212, the wind speed and solar radiation resource pairs are fitted using the t-Copula function. The t-Copula function belongs to the elliptic family of Copulas and is derived from a multivariate t-distribution. The t-Copula function has a symmetrical tail structure, meaning that the upper and lower tails have the same tail correlation coefficient. The t-Copula function contains two parameters: a correlation matrix parameter controlling the overall correlation between variables, and a degree of freedom parameter controlling the thickness of the tail. The smaller the degree of freedom, the thicker the tail, and the higher the probability of extreme values ​​occurring simultaneously. After transforming the wind speed resource data and solar radiation resource data into uniform distribution sequences, the parameters are estimated using the t-Copula function to obtain the correlation matrix parameter and degree of freedom parameter. The t-Copula function can simultaneously capture the positive or negative correlation between wind speed and solar radiation, and it exhibits symmetry in its dependence on extreme values ​​at both ends. After comparison using screening criteria, the goodness of fit of the t-Copula function is superior to other Copulas functions such as Gumbel, Clayton, and Frank, and therefore it was selected as the optimal pairwise Copula function.

[0089] It should be noted that in step S213, the t-Copula function is also used as the optimal pairwise Copula function for the wave height and solar radiation resource pairs. The dependency between wave height and solar radiation can be complex. For example, when solar radiation is strong in summer, sea surface warming may affect atmospheric stability and thus indirectly affect waves, but this relationship is not a simple one-tailed extreme dependency. The symmetric tail characteristic of the t-Copula function is suitable for describing this type of relationship. After transforming the wave height resource data and solar radiation resource data into uniformly distributed sequences, the t-Copula function is used for parameter estimation to obtain the corresponding correlation matrix parameters and degrees of freedom parameters. Through comparison using screening criteria, the t-Copula function has the best performance and is therefore selected as the optimal pairwise Copula function for the wave height and solar radiation resource pairs.

[0090] In this embodiment, Copula functions with corresponding tail distribution characteristics are selected for different resources to their respective dependence characteristics, so that the three-dimensional C-Vine Copula joint probability distribution model can accurately capture the upper tail correlation between wind speed and wave height, as well as the symmetrical tail correlation between wind speed and solar radiation, and between wave height and solar radiation.

[0091] S3. Based on the three-dimensional C-Vine Copula joint probability distribution model, extract resource combination feature data under the target joint probability condition and generate resource combination feature vector;

[0092] In one embodiment, the target joint probability condition can be set to a specific joint probability value, such as a certain equiprobability surface of the joint probability distribution. On this equiprobability surface, the combinations of wind speed, wave height, and solar radiation satisfy the same cumulative joint probability. Resource combination feature data for each spatial grid point on the aforementioned equiprobability surface are extracted. That is, for each spatial grid point within the target sea area, based on the edge distribution parameters of the three resources at that grid point and the dependency structure determined by the C-Vine Copula model, the specific values ​​of wind speed, wave height, and solar radiation corresponding to when the joint probability equals a preset value are calculated. These values ​​are combined to form a multidimensional vector, called the resource combination feature vector.

[0093] Furthermore, in this embodiment of the application, step S3 includes:

[0094] S31. Based on the three-dimensional C-Vine Copula joint probability distribution model, calculate the resource combination feature data on the equiprobability surface corresponding to the preset joint probability value.

[0095] The three-dimensional C-Vine Copula joint probability distribution model defines the joint cumulative distribution function of wind speed resource data, wave height resource data, and solar radiation resource data. For a given spatial grid point, the three resource data at that grid point each follow their respective optimal marginal distributions, and the dependencies are determined by the C-Vine Copula structure.

[0096] The preset joint probability value is a value between zero and one. For example, a moderate probability level is chosen to represent the resource combination under medium probability.

[0097] Here, the equiprobability surface refers to all points that satisfy the joint cumulative distribution function being equal to the preset value.

[0098] It should be noted that on an equiprobability surface, the combination of wind speed, wave height, and solar radiation values ​​is not uniquely determined, and further reasonable selection criteria need to be added. As a possible implementation, the point with the largest joint probability density function, i.e., the mode point on the equiprobability surface, can be selected; alternatively, points where the quantile levels of each resource are the same can be selected. In practice, numerical methods can be used to solve this problem. For example, a search can be conducted on a three-dimensional grid to find points that satisfy the condition that the joint cumulative distribution function equals a preset value and has the largest joint probability density, yielding a set of wind speed, wave height, and solar radiation values, which can then be used as the resource combination characteristic data on the equiprobability surface.

[0099] S32. Extract the resource combination feature data of each spatial grid point on the equiprobability surface, and combine them to form a resource combination feature vector.

[0100] It should be noted that for each spatial grid point within the target sea area, the parameters of the optimal edge distribution models for wind speed, wave height, and solar radiation differ, and the dependency parameters of C-VineCopula may also vary with spatial location. Therefore, each grid point corresponds to different equiprobability surface resource combination feature data. In one embodiment, step S31 is calculated independently for each grid point to obtain a set of triplet data, namely, the wind speed value, wave height value, and solar radiation value at that grid point. These values ​​are arranged in a preset order, such as wind speed value, wave height value, and solar radiation value, to form a three-dimensional vector, called the resource combination feature vector. The resource combination feature vectors of all spatial grid points within the target sea area constitute a vector set, with each grid point corresponding to a unique feature vector.

[0101] In this embodiment, resource combination features are extracted by fixing the joint probability level, eliminating the interference of differences in the marginal distribution of different spatial grid points on feature comparability, and generating a unified feature vector suitable for cluster analysis.

[0102] S4. Using the resource combination feature vector as input, the K-Means clustering algorithm is used to perform clustering analysis on each spatial grid point of the target sea area to generate resource combination partitioning results. The resource combination partitioning results are used to classify different resource combination types.

[0103] Furthermore, in this embodiment of the application, step S4 includes:

[0104] S41. Obtain the resource combination feature vector of each spatial grid point to form a feature vector set;

[0105] It should be noted that each spatial grid point in the feature vector set corresponds to a multi-dimensional vector, and the dimension of the vector is equal to the number of resource types. The resource combination feature vectors of all spatial grid points within the target sea area are arranged in spatial order or any other order to form a complete set.

[0106] S42. Randomly select a preset number of feature vectors from the feature vector set as initial cluster centers;

[0107] It should be noted that the preset number refers to the expected number of clusters. The random selection method involves randomly drawing a preset number of feature vectors without replacement from the feature vector set, with each selected feature vector serving as an initial cluster center. The selection of initial cluster centers affects the convergence speed and final result of the clustering. Stability can be improved by performing multiple random initializations and selecting the optimal result.

[0108] S43. Calculate the distance between the feature vector of each spatial grid point and each initial cluster center, and assign each spatial grid point to the cluster to which the initial cluster center with the smallest distance belongs;

[0109] It should be noted that, for each vector in the feature vector set, the distance between the vector and each cluster center is calculated, these distance values ​​are compared, and the cluster center with the smallest distance is found. The spatial grid point is then assigned to the cluster represented by that cluster center. After step S43, all spatial grid points are initially divided into multiple clusters, and each cluster corresponds to a cluster center.

[0110] S44. Calculate the mean of all feature vectors within each cluster, and update the mean to the new cluster center of the corresponding cluster;

[0111] It should be noted that, for a given cluster, the feature vectors of all spatial grid points within that cluster are taken, and the arithmetic mean is calculated for each dimension to obtain a mean vector. This mean vector serves as the new cluster center for that cluster. The new cluster center may differ from the initial cluster center randomly selected in step S42, reflecting the central location of all samples within the current cluster.

[0112] S45. Repeat S43 and S44 until the cluster centers of each cluster no longer change. Output the cluster label to which each spatial grid point belongs and generate the resource combination partitioning result.

[0113] It should be noted that the specific process of repeated execution is as follows: replace the original cluster centers with the new cluster centers obtained in step S44, return to step S43, recalculate the distance between the feature vector of each spatial grid point and each new cluster center, and reallocate each spatial grid point to the cluster with the smallest distance; then execute step S44 again to calculate the intra-cluster mean for a new round and update the cluster centers. Repeat the above process until the new cluster centers of all clusters are exactly the same as the cluster centers of the previous iteration, or the change is less than a preset threshold, then the clustering is determined to have converged, and the iteration stops. After the clustering converges, the cluster label to which each spatial grid point belongs is output, where the cluster label is used to represent the cluster to which each spatial grid point finally belongs, and the cluster labels of all spatial grid points together constitute the resource combination partitioning result.

[0114] In this embodiment, the goal of the K-Means clustering algorithm is to minimize the sum of squared distances from all samples within a cluster to their cluster centers. Through alternating allocation operations in step S43 and update operations in step S44, the algorithm gradually reduces the objective function value, eventually converging to a local optimum. Within the same cluster, the resource combination feature vectors of spatial grid points are close to each other, meaning that these grid points have similar wind speed, wave height, and solar radiation configurations under the same joint probability level. Between different clusters, the feature vectors differ significantly. Therefore, the resource combination partitioning results output in step S45 can accurately reflect the spatial differentiation patterns of resource combination patterns within the target sea area, providing a reasonable basis for subsequent regional division assessment of development potential.

[0115] For example, calculate the distance between the feature vector of each spatial grid point and each initial cluster center. This distance can be Euclidean distance or Manhattan distance.

[0116] For example, the resource combination zoning results are used to classify different resource combination types. Some zonings are characterized by high wind speed, relatively high wave height, and moderate solar radiation, while other zonings are characterized by moderate wind speed, relatively low wave height, and high solar radiation.

[0117] In this embodiment, the resource combination feature vectors form different clusters in the multidimensional space. The K-Means algorithm finds the natural boundaries of these clusters by minimizing the sum of squares within the cluster, so that the spatial grid points within the same partition have similar resource combination patterns, and the pattern differences between different partitions are significant.

[0118] In one embodiment, randomly selecting a preset number of feature vectors in S42 further includes: setting the number of clusters to four clusters, randomly selecting four feature vectors as initial cluster centers, and after executing S43 to S45, outputting the cluster label of one of the four clusters to which each spatial grid point belongs, thereby generating a resource combination partitioning result containing four resource combination clusters.

[0119] In step S42, the number of clusters is set to four. The setting of four clusters is based on the typical classification requirements of multi-energy complementary resource combinations at sea. For example, the four clusters can correspond to wind speed-dominated, wave height-dominated, solar radiation-dominated, and balanced types, or to high-energy, medium-energy, low-energy, and special combination types. According to this setting, four feature vectors are randomly selected from the feature vector set as initial cluster centers. During random selection, it is ensured that the four initial cluster centers are distinct and represent samples at different locations in the feature vector space.

[0120] Perform steps S43 to S45, following the iterative allocation and update process described above for cluster analysis. In each iteration, calculate the distance between the feature vector of each spatial grid point and the four cluster centers, and assign the spatial grid point to the cluster belonging to the cluster center with the smallest distance. Repeat until the four cluster centers no longer change. After convergence, each spatial grid point obtains a cluster label, which is one of four preset identifiers, used for the four corresponding clusters.

[0121] Output the cluster labels corresponding to each spatial grid point to obtain the resource combination partitioning result containing four resource combination clusters. Based on the cluster labels, the target sea area is divided into four regions, each region corresponding to a resource combination cluster. The resource combination characteristics of each spatial grid point within the same cluster are highly consistent, while the resource combination characteristics between different clusters show significant differences.

[0122] In the above embodiments, by pre-setting four clusters to output four types of resource combination spatial partitions, the partition granularity can distinguish the main resource combination modes without being too fragmented, which facilitates the subsequent formulation of differentiated multi-functional complementary development strategies for different partitions.

[0123] S5. Obtain the resource combination partitioning results and the water depth data corresponding to each partition. Based on the water depth data, determine the multi-energy complementary development potential of each partition and generate the final resource partitioning results.

[0124] Furthermore, in this embodiment of the application, step S5 includes:

[0125] S51. Obtain the resource combination partitioning results and the water depth data for each spatial grid point;

[0126] It should be noted that each spatial grid point in the resource combination partitioning results is assigned a cluster label, and spatial grid points with the same cluster label belong to the same partition. Water depth data comes from water depth measurements of the target sea area or global water depth models. Each spatial grid point corresponds to a water depth value, usually in meters. Positive values ​​represent water depth, while negative values ​​represent elevation or land. For spatial grid points within a sea area, water depth data is generally taken as positive values, with larger values ​​indicating deeper water.

[0127] S52. For each partition in the resource combination partitioning result, statistically analyze the water depth data distribution characteristics of all spatial grid points within the partition.

[0128] For example, the distribution characteristics of water depth data include, but are not limited to, minimum, maximum, mean, median, mode, and frequency histograms or cumulative frequency curves of water depth values.

[0129] It should be noted that the above statistical method is as follows: For a given zone, water depth values ​​of all spatial grid points within that zone are collected to form a water depth dataset. The arithmetic mean of this dataset is calculated to obtain the average water depth. The water depth values ​​are sorted in ascending order, and the median value is obtained. The proportion of grid points in each water depth interval to the total number of grid points is calculated to obtain the water depth frequency distribution. These distribution characteristics are used to quantify the overall water depth conditions and water depth variation range of the zone.

[0130] S53. Calculate the proportion of the number of spatial grid points in each partition whose water depth data is within the preset exploitable water depth range to the total number of spatial grid points in the partition, and mark the partitions with a proportion greater than the preset proportion threshold as priority development areas.

[0131] For example, the tagging method could be to add an attribute field, such as a priority identifier or a boolean-type exploitable tag.

[0132] S54. Output the final resource partitioning results, including priority development area markers.

[0133] It should be noted that the final resource zoning result is based on the resource combination zoning result, with the addition of a marker indicating whether each zone belongs to a priority development area. The output format can be a Geographic Information System (GIS) layer, with each zone polygon bearing a zone number, a resource combination type description, and a priority development marker. The final resource zoning result can be directly used to guide the site selection and planning of offshore multi-energy complementary projects.

[0134] As one possible approach, even if a zone has a combination of wind speed, wave height, and solar radiation that is highly conducive to multi-energy complementarity, if the water depth is too shallow or too deep to allow for equipment installation, then the zone does not have practical development value.

[0135] This application provides a method for marine multi-energy complementary resource zoning based on C-Vine Copula and K-Means clustering. First, wind speed, wave height, and solar radiation data are acquired and preprocessed to fit marginal distributions, generating three optimal marginal distribution models. A three-dimensional C-Vine Copula joint distribution model with wind speed as the root node is constructed to quantify the three resource dependencies. Resource combination features under joint probability are extracted to generate feature vectors for each spatial grid point. K-Means clustering is used to partition each grid point, generating resource combination zoning results. The development potential of each zone is evaluated using water depth data, and the final zoning result is output. This application uses C-Vine Copula to characterize the nonlinear dependencies of multiple resources, combines it with equiprobability surfaces to extract features uniformly, and introduces water depth engineering constraints to achieve scientific zoning, improving the accuracy and practicality of multi-energy complementary development potential assessment.

[0136] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0137] To implement the above embodiments, this application also proposes a terminal device.

[0138] Figure 2 This is a schematic diagram of the structure of a terminal device according to an embodiment of this application.

[0139] like Figure 2 As shown, the terminal device 200 includes:

[0140] The system includes a memory 210 and at least one processor 220, and a bus 230 connecting different components (including the memory 210 and the processor 220). The memory 210 stores a computer program, which, when executed by the processor 220, implements the marine multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering as described in the embodiments of this application.

[0141] Bus 230 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MCA) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0142] Terminal device 200 typically includes various electronically readable media. These media can be any available media that can be accessed by terminal device 200, including volatile and non-volatile media, removable and non-removable media.

[0143] Memory 210 may also include computer system readable media in the form of volatile memory, such as RAM 240 and / or cache 250. Terminal device 200 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 260 may be used to read and write non-removable, non-volatile magnetic media (… Figure 2 Not shown; usually referred to as a "hard drive"). Although Figure 2 As not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 230 via one or more data media interfaces. Memory 210 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application.

[0144] A program / utility 280 having a set (at least one) of program modules 270 may be stored in, for example, memory 210. Such program modules 270 include—but are not limited to—an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 270 typically perform the functions and / or methods described in the embodiments of this application.

[0145] Terminal device 200 can also communicate with one or more external devices 290 (e.g., keyboard, pointing device, display 291, etc.), and with one or more devices that enable a user to interact with terminal device 200, and / or with any device that enables terminal device 200 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 292. Furthermore, terminal device 200 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 293. As shown, network adapter 293 communicates with other modules of terminal device 200 via bus 230. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with terminal device 200, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0146] The processor 220 performs various functional applications and data processing by running programs stored in the memory 210.

[0147] It should be noted that the implementation process and technical principles of the terminal device in this embodiment are explained in the foregoing description of a marine multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering in this application embodiment, and will not be repeated here.

[0148] Corresponding to the above embodiment, a method for partitioning marine multi-energy complementary resources based on C-Vine Copula and K-Means clustering, Figure 3 This paper illustrates a structural block diagram of a marine multi-energy complementary resource zoning system based on C-Vine Copula and K-Means clustering, according to an embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown.

[0149] Reference Figure 3 The system 300 includes:

[0150] The data acquisition and edge distribution fitting module 301 is used to acquire three types of resource data of the target sea area, including wind speed resource data, wave height resource data and solar radiation resource data. The module preprocesses the three types of resource data and fits the edge distribution of the three types of resource data respectively. It estimates the model parameters of the probability distribution model corresponding to each resource data and generates the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height and the optimal edge distribution model of solar radiation.

[0151] The 3D C-Vine Copula joint modeling module 302 is used to construct a 3D C-Vine Copula joint probability distribution model with wind speed resource as the root node based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation. The 3D C-Vine Copula joint probability distribution model is used to quantify the dependency relationship between the three resources.

[0152] The resource combination feature extraction module 303 is used to extract resource combination feature data under the target joint probability condition based on the three-dimensional C-Vine Copula joint probability distribution model, and generate resource combination feature vector;

[0153] The K-Means clustering partitioning module 304 is used to perform clustering analysis on each spatial grid point of the target sea area using the K-Means clustering algorithm as input, and generate resource combination partitioning results. The resource combination partitioning results are used to classify different resource combination types.

[0154] The partition potential assessment module 305 is used to obtain the resource combination partitioning results and the water depth data corresponding to each partition, determine the multi-energy complementary development potential of each partition based on the water depth data, and generate the final resource partitioning results.

[0155] In practical use, the marine multi-energy complementary resource partitioning system based on C-Vine Copula and K-Means clustering provided in this application embodiment can be configured in any terminal device to execute the aforementioned marine multi-energy complementary resource partitioning method based on C-Vine Copula and K-Means clustering.

[0156] This application provides a marine multi-energy complementary resource zoning system based on C-Vine Copula and K-Means clustering. First, wind speed, wave height, and solar radiation data are acquired and preprocessed to fit marginal distributions, generating three optimal marginal distribution models. A three-dimensional C-Vine Copula joint distribution model with wind speed as the root node is constructed to quantify the three resource dependencies. Resource combination features under joint probability are extracted, generating feature vectors for each spatial grid point. K-Means clustering is used to partition each grid point, generating resource combination zoning results. The development potential of each zone is evaluated using water depth data, and the final zoning result is output. This application uses C-Vine Copula to characterize the nonlinear dependencies of multiple resources, combines it with equiprobability surfaces to extract features uniformly, and introduces water depth engineering constraints to achieve scientific zoning, improving the accuracy and practicality of multi-energy complementary development potential assessment.

[0157] It should be noted that the information interaction and execution process between the above systems / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0158] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0159] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps described in the various method embodiments above.

[0160] This application provides a computer program product that, when run on a terminal device, causes the terminal device to execute the steps described in the various method embodiments above.

[0161] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some regions, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0162] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0163] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can implement the described functions using different methods for each specific application, but such implementation should not be considered beyond the scope of this application.

[0164] In the embodiments provided in this application, it should be understood that the disclosed systems / terminal devices and methods can be implemented in other ways. For example, the system / terminal device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0165] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0166] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for partitioning marine multi-energy complementary resources based on C-Vine Copula and K-Means clustering, characterized in that, include: S1. Obtain three types of resource data for the target sea area, including wind speed resource data, wave height resource data, and solar radiation resource data. Preprocess the three types of resource data and fit the marginal distribution of each type of resource data respectively. Estimate the model parameters of the probability distribution model corresponding to each resource data and generate the optimal marginal distribution model of wind speed, the optimal marginal distribution model of wave height, and the optimal marginal distribution model of solar radiation. S2. Based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, a three-dimensional C-Vine Copula joint probability distribution model with wind speed resource as the root node is constructed. The three-dimensional C-Vine Copula joint probability distribution model is used to quantify the dependency relationship between the three resources. S3. Based on the three-dimensional C-Vine Copula joint probability distribution model, extract resource combination feature data under the target joint probability condition and generate resource combination feature vector; S4. Using the resource combination feature vector as input, the K-Means clustering algorithm is used to perform clustering analysis on each spatial grid point of the target sea area to generate resource combination partitioning results. The resource combination partitioning results are used to classify different resource combination types. S5. Obtain the resource combination partitioning results and the water depth data corresponding to each partition. Based on the water depth data, determine the multi-energy complementary development potential of each partition and generate the final resource partitioning results.

2. The method according to claim 1, characterized in that, S1 involves fitting the marginal distributions of the three types of resource data, estimating the model parameters of the probability distribution model corresponding to each resource data, and generating the optimal marginal distribution model for wind speed, wave height, and solar radiation, including: For each type of resource data, multiple candidate marginal distribution models are used for fitting, and the model parameters of each candidate marginal distribution model are estimated. The goodness-of-fit value of each candidate marginal distribution model is calculated. The candidate marginal distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data, thus obtaining the optimal marginal distribution model for wind speed, wave height, and solar radiation.

3. The method according to claim 2, characterized in that, For each type of resource data, multiple candidate marginal distribution models are used for fitting, and the goodness-of-fit value of each candidate marginal distribution model is calculated. The candidate marginal distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for the corresponding resource data, resulting in the optimal marginal distribution model for wind speed, wave height, and solar radiation, including: Based on wind speed time series data in wind speed resource data, the Weibull distribution model is used for parameter fitting, the goodness-of-fit value of the Weibull distribution model is calculated, and the Weibull distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model for wind speed. Based on the wave height time series data in the wave height resource data, the Gamma distribution model is used for parameter fitting, the goodness of fit value of the Gamma distribution model is calculated, and the Gamma distribution model with the smallest goodness of fit value is selected as the optimal marginal distribution model for wave height. Based on solar radiation time series data in solar radiation resource data, a nonlinear mixed-weight zero-inflation Beta distribution model is used for parameter fitting. The goodness-of-fit value of the nonlinear mixed-weight zero-inflation Beta distribution model is calculated, and the nonlinear mixed-weight zero-inflation Beta distribution model with the smallest goodness-of-fit value is selected as the optimal marginal distribution model of solar radiation.

4. The method according to claim 3, characterized in that, S2, based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, constructs a three-dimensional C-VineCopula joint probability distribution model with wind speed resource as the root node, including: S21. Based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, various Copula functions are used to fit the three resource pairs of wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation respectively. The optimal pairwise Copula functions are selected for each resource pair through the preset screening criteria. S22. Taking wind speed resource as the root node, the optimal pairwise Copula functions of wind speed and wave height resource pairs and the optimal pairwise Copula functions of wind speed and solar radiation resource pairs are used as the first layer connection structure, and the optimal pairwise Copula functions of wave height and solar radiation resource pairs are used as the second layer connection structure to construct a three-dimensional C-Vine Copula joint probability distribution model.

5. The method according to claim 4, characterized in that, S21, based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation, fits the three resource pairs of wind speed and wave height, wind speed and solar radiation, and wave height and solar radiation using various Copula functions. The optimal pairwise Copula functions for each resource pair are selected according to a preset screening criterion, including: For wind speed and wave height resource pairs, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the Gumbel Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for wind speed and wave height resource pairs after comparing the information criterion values. For wind speed and solar radiation resource pairs, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for wind speed and solar radiation resource pairs after comparing the information criterion values. For the pair of wave height and solar radiation resources, various Copula functions are used for fitting, the information criterion value of each Copula function is calculated, and the t-Copula function with the smallest information criterion value is selected as the optimal pairwise Copula function for the pair of wave height and solar radiation resources.

6. The method according to claim 5, characterized in that, S3, based on the three-dimensional C-Vine Copula joint probability distribution model, extracts resource combination feature data under the target joint probability condition and generates a resource combination feature vector, including: S31. Based on the three-dimensional C-Vine Copula joint probability distribution model, calculate the resource combination feature data on the equiprobability surface corresponding to the preset joint probability value. S32. Extract the resource combination feature data of each spatial grid point on the equiprobability surface, and combine them to form a resource combination feature vector.

7. The method according to claim 6, characterized in that, S4, using the resource combination feature vector as input, employs the K-Means clustering algorithm to perform clustering analysis on each spatial grid point of the target sea area, generating resource combination partitioning results, including: S41. Obtain the resource combination feature vector of each spatial grid point to form a feature vector set; S42. Randomly select a preset number of feature vectors from the feature vector set as initial cluster centers; S43. Calculate the distance between the feature vector of each spatial grid point and each initial cluster center, and assign each spatial grid point to the cluster to which the initial cluster center with the smallest distance belongs; S44. Calculate the mean of all feature vectors within each cluster, and update the mean to the new cluster center of the corresponding cluster; S45. Repeat S43 and S44 until the cluster centers of each cluster no longer change. Output the cluster label to which each spatial grid point belongs and generate the resource combination partitioning result.

8. The method according to claim 7, characterized in that, S5 involves obtaining the resource combination partitioning results and the corresponding water depth data for each partition, determining the multi-energy complementary development potential of each partition based on the water depth data, and generating the final resource partitioning results, including: S51. Obtain the resource combination partitioning results and the water depth data for each spatial grid point; S52. For each partition in the resource combination partitioning result, statistically analyze the water depth data distribution characteristics of all spatial grid points within the partition. S53. Calculate the proportion of the number of spatial grid points in each partition whose water depth data is within the preset exploitable water depth range to the total number of spatial grid points in the partition, and mark the partitions with a proportion greater than the preset proportion threshold as priority development areas. S54. Output the final resource partitioning results, including priority development area markers.

9. A marine multi-energy complementary resource zoning system based on C-Vine Copula and K-Means clustering, applied to the method described in any one of claims 1-8, characterized in that, include: The data acquisition and edge distribution fitting module is used to acquire three types of resource data for the target sea area, including wind speed resource data, wave height resource data, and solar radiation resource data. The module preprocesses the three types of resource data and fits the edge distribution of each type of resource data. It estimates the model parameters of the probability distribution model corresponding to each resource data and generates the optimal edge distribution model for wind speed, wave height, and solar radiation. The 3D C-Vine Copula joint modeling module is used to construct a 3D C-Vine Copula joint probability distribution model with wind speed resource as the root node, based on the optimal edge distribution model of wind speed, the optimal edge distribution model of wave height, and the optimal edge distribution model of solar radiation. The 3D C-Vine Copula joint probability distribution model is used to quantify the dependency relationship between the three resources. The resource combination feature extraction module is used to extract resource combination feature data under the target joint probability condition based on the three-dimensional C-Vine Copula joint probability distribution model, and generate resource combination feature vectors. The K-Means clustering partitioning module is used to perform clustering analysis on each spatial grid point of the target sea area using the K-Means clustering algorithm as input, and generate resource combination partitioning results. The resource combination partitioning results are used to classify different resource combination types. The partition potential assessment module is used to obtain the resource combination partitioning results and the water depth data corresponding to each partition. Based on the water depth data, the multi-energy complementary development potential of each partition is determined, and the final resource partitioning results are generated.

10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

Images