Offshore wind power decoupling evaluation method based on improved data envelope analysis

Abstract

The invention discloses an offshore wind power decoupling evaluation method based on improved data envelope analysis, relates to the technical field of offshore wind power delivery, and is used for solving the problem that existing evaluation is not objective, and the method comprises the following steps: constructing an input-output index system for evaluating an offshore wind power delivery scheme; historical operation data of the offshore wind power sending-out system are obtained, and according to the historical operation data, the coupling relation between the indexes in the index system is quantified through a mutual information distance coupling coefficient; according to the mutual information distance coupling coefficient, weighting the index system by adopting a network analysis method to obtain a weight vector reflecting a coupling relationship between indexes; converting the weight vector into a weight constraint, and constructing an improved data envelope analysis model; and comprehensively evaluating the to-be-evaluated scheme by using the improved data envelope analysis model. The invention provides an objective evaluation method by combining efficiency evaluation of data envelope analysis with a mutual information distance coupling coefficient.

Description

Technical Field

[0001] This invention relates to the field of offshore wind power transmission technology, and more particularly to an offshore wind power decoupling evaluation method based on improved data envelopment analysis. Background Technology

[0002] Currently, offshore wind power, with its significant advantages such as stable wind energy resources, huge reserves, and high utilization hours, has become a key pillar in the energy system for ensuring power supply security. Among these, the system evaluation of offshore wind power transmission schemes is a critical engineering decision-making issue, directly related to the safety, economy, reliability, and power quality of the power grid.

[0003] At the technical level, the system evaluation of offshore wind power transmission schemes needs to balance the applicability of transmission technology, transmission efficiency, and system operational reliability. This involves assessing the annual average availability of the entire transmission system, the fault ride-through capability of key equipment, the system's operational complexity, and the risk of broadband oscillations to ensure long-term stable operation. Furthermore, it requires evaluating the entire transmission system's adaptability to weak grids at onshore access points, its active / reactive power regulation capabilities, and its control flexibility and response speed under system faults or black-start scenarios to ensure overall grid stability.

[0004] Furthermore, from an economic perspective, the key to the assessment lies in a meticulous trade-off and optimization of the entire lifecycle cost of the transmission scheme to achieve optimal economic benefits. This requires assessing the total amount of upfront fixed asset investment, including high-voltage equipment, submarine cable laying, converter / substation construction, and land acquisition. It also includes the recurring expenses required for daily maintenance, emergency repairs, insurance, and personnel allocation throughout the system's operational lifecycle. Finally, it quantifies the energy losses and corresponding economic value losses caused by transmission losses during long-term operation of the transmission system.

[0005] The complex nonlinear coupling and interrelationships among the indicators in the system evaluation of offshore wind power transmission schemes make it difficult to objectively quantify the coupling strength among multiple indicators using traditional evaluation methods such as the analytic hierarchy process (AHP) and entropy weighting. This leads to high subjectivity in weight allocation and fails to truly reflect the transmission of system risks and synergistic benefits. Therefore, it is necessary to propose a mechanism that accurately captures the information distance coupling relationship between indicators to overcome the subjective defects of traditional weighting and achieve data-driven objective weighting evaluation. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, one of the objectives of this invention is to provide an improved data envelopment analysis-based decoupling evaluation method for offshore wind power. This method quantifies the nonlinear correlation between indicators through mutual information distance coupling coefficients, and uses this to constrain the data envelopment analysis model, thereby achieving an accurate evaluation of the coupling relationship of offshore wind power transmission schemes.

[0007] One of the objectives of this invention is achieved through the following technical solution:

[0008] A decoupled evaluation method for offshore wind power based on improved data envelopment analysis includes the following steps: Construct an input-output indicator system for evaluating offshore wind power transmission schemes; Historical operating data of offshore wind power transmission systems are acquired, and the coupling relationship between indicators in the indicator system is quantified based on the historical operating data using the mutual information distance coupling coefficient. Based on the mutual information distance coupling coefficient, the index system is weighted using network analysis to obtain a weight vector reflecting the coupling relationship between the indicators; The weight vector is transformed into weight constraints to construct an improved data envelopment analysis model; the improved data envelopment analysis model is then used to comprehensively evaluate the offshore wind power transmission scheme to be evaluated.

[0009] Furthermore, in order to make the evaluation conclusions more comprehensive, the indicators in the indicator system shall include at least input indicators and output indicators; The input-type indicators include at least the total life cycle cost indicators, and the cost indicators include at least the initial investment cost, operation and maintenance cost, operating loss cost, and equipment retirement cost; The output-type indicators include at least: Transmission capacity and efficiency indicators, including rated transmission capacity and active power loss per unit distance; Control flexibility indicators include active / reactive power regulation capability, black start capability, and adaptability to weak power grids; Operational reliability metrics include annual average availability, operational complexity, fault ride-through capability, and broadband oscillation risk; Carbon emission reduction benefit indicators include carbon dioxide emission reduction, contribution rate of power grid cleanliness, and sensitivity of system carbon emission correlation.

[0010] To ensure that heterogeneous data from different sources and with different dimensions can be processed uniformly and in a standardized manner, and to provide a reliable data foundation for subsequent high-precision coupling relationship calculations and model solving, historical operational data of offshore wind power transmission systems are acquired, including: Build with Sample The original indicator data matrix of each evaluation indicator ; The original index data matrix is ​​then normalized.

[0011] Furthermore, in order to more comprehensively and robustly quantify the complex correlation strength between any two indicators and overcome the limitations of a single correlation coefficient, the coupling relationship between the indicators in the indicator system is quantified by the mutual information distance coupling coefficient, including: For any two index vectors, calculate their mutual information coefficient, which satisfies the following: ,in, Indicators and Mutual information coefficient between them Indicates the number of evaluation indicators. and Indicators or indicators The marginal probability distribution; Calculate the distance correlation coefficient between the two indicator vectors, the calculation of which satisfies: ,in, Represents a bicentered matrix , The correlation coefficient between the distances, The squared variance of the distance to the bicentered matrix. Represents a bicentered matrix , The covariance matrix between them , As an indicator and The distance matrix is ​​processed by bicentering to obtain the bicentered matrix; The mutual information coefficient and the distance correlation coefficient are weighted and fused to obtain the mutual information distance coupling coefficient, which is calculated to satisfy the following: ,in, As an indicator and Mutual information distance coupling coefficient between them For balancing parameters, As an indicator and The normalized value of the mutual information coefficient.

[0012] Furthermore, for any two index vectors, their mutual information coefficient is calculated, including: Transform continuous index data into discrete probability distributions; Calculate the joint probability distribution and the individual marginal probability distributions of the two indicators; Based on the joint probability distribution and the marginal probability distribution, the information entropy of the two indicators is calculated, and the calculation satisfies: , in, Indicators Information entropy Indicators Information entropy; The mutual information value is calculated based on the information entropy and joint probability distribution, and then normalized. The normalization calculation satisfies the following: ,in, Indicators and The minimum value of its information entropy.

[0013] This gives the mutual information value a clear upper bound, facilitating interpretation and comparison; and improves the stability of the calculation and the comparability between different indicator pairs. This ensures the quality of input data for subsequent weighting steps and is the foundation for achieving objective and scientific weighting.

[0014] Furthermore, to eliminate the influence of the data mean and isolate the inherent dependence between distances, the distance correlation coefficient between the two indicator vectors is calculated, including: Calculation indicators and index vector and The Euclidean distance between any two points in the middle forms an index. and Distance matrix and ; For distance matrix and Perform dual-center processing to generate a dual-center matrix. and This eliminates the dependency between mean effect and isolation distance.

[0015] To accurately reflect the inherent and objective coupling relationships within the indicator system, and to eliminate the subjective element of expert scoring in traditional network analysis, this paper employs network analysis to weight the indicator system based on the mutual information distance coupling coefficient, obtaining a weight vector reflecting the coupling relationships between indicators. This process includes the following steps: The mutual information distance coupling coefficient between each pair of indicators is used as the objective influence coefficient. The objective influence coefficient is normalized column-wise to construct the objective influence matrix. Solve for all eigenvectors of the objective influence matrix and construct a weight matrix composed of the eigenvectors. ; Based on the weight matrix The objective influence matrix is ​​weighted to construct a weighted supermatrix;

[0016] in, Represents the first in the objective influence matrix Line number The weighted element of the column, Represents the weight matrix of the first element. Line number The weighted element of the column; The weighted hypermatrix is ​​subjected to power iterations, and the weight vector is obtained after convergence.

[0017] Furthermore, the weight vector is transformed into weight constraints, which satisfy the following: , in, Indicates the first The weight of each input indicator Indicates the first The weight of input indicators Indicates the first Weights of output indicators Indicates the first The weight of output indicators and These represent the lower limits of the input-type weighting ratio and the output-type weighting ratio, respectively. and They represent the first Item and the The extreme weight of input-type indicators and They represent the first Item and the The extreme weights of output-type indicators This indicates a very small amount of relaxation.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention effectively solves the problems of single-dimensionality and inaccurate ranking in traditional evaluation methods by combining the efficiency evaluation of data envelopment analysis with the mutual information distance coupling coefficient, thereby achieving comprehensive evaluation and accurate ranking of typical offshore wind power transmission schemes. This invention also solves the technical problems of the difficulty in objectively quantifying the nonlinear coupling relationship between indicators in the evaluation of offshore wind power transmission schemes and the strong subjectivity of traditional methods in weighting, providing a new objective, accurate and interpretable evaluation framework for engineering decision-making. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the present invention; Figure 2 This is a flowchart of the offshore wind power decoupling evaluation method based on improved data envelopment analysis in Example 1; Figure 3 This is an example diagram of weighted coupling sensitivity analysis in Example 2; Figure 4 This is an example diagram comparing the evaluation results of Embodiment 2 with those of the prior art. Detailed Implementation

[0020] The present invention will now be described in more detail with reference to the accompanying drawings. It should be noted that the following description of the present invention with reference to the accompanying drawings is merely illustrative and not restrictive. Various embodiments can be combined with each other to form other embodiments not shown in the following description.

[0021] Example 1 Example 1 provides a decoupled evaluation method for offshore wind power based on improved data envelopment analysis. It aims to solve the problems of difficulty in quantifying the complex coupling relationship between indicators in the evaluation of offshore wind power transmission schemes and the strong subjectivity of traditional evaluation. It uses data-driven mutual information distance coupling coefficient to objectively quantify the nonlinear correlation between indicators, and uses this as a constraint to improve the data envelopment analysis model, and finally realizes the comprehensive evaluation and accurate ranking of typical offshore wind power transmission schemes.

[0022] The evaluation of offshore wind power transmission schemes is a complex systems engineering project involving multiple inputs, multiple outputs, and strong correlations, and the evaluation system is the foundation of the project. However, traditional methods such as the analytic hierarchy process (AHP) and entropy weighting have the following main drawbacks: on the one hand, weighting relies heavily on expert experience, leading to high subjectivity in the evaluation results; on the other hand, these methods struggle to objectively quantify the complex nonlinear couplings and interrelationships between various indicators. These shortcomings prevent the evaluation results from truly reflecting the transmission mechanism of system risks and synergistic benefits, lacking comprehensiveness and accuracy, and failing to meet the scientific requirements of major engineering decisions.

[0023] This embodiment innovatively introduces a mutual information distance coupling coefficient to achieve precise quantification and evaluation of the coupling relationship between indicators. Simultaneously, this method uses the mutual information distance coupling coefficient as an objective influence coefficient, replacing the subjective judgment in traditional network analysis methods, and constructs a coupling weighting mechanism for network analysis methods with mutual information distance coupling coefficient correction. Finally, this coupling weight is transformed into weight constraints in an improved data envelopment analysis model, ensuring that the evaluation results not only reflect the efficiency of the solution itself but also incorporate the complex coupling information and associated risks between indicators, providing a reliable basis for scientific operation and management decisions in engineering projects.

[0024] Please refer to Figure 1 The flowchart shown and Figure 2 The overall flowchart is shown. This embodiment provides a decoupling evaluation method for offshore wind power based on improved data envelopment analysis, including the following steps: S1. Construct an input-output indicator system for evaluating offshore wind power transmission schemes; The indicators in the S1 indicator system include at least input indicators and output indicators; The input-type indicators include at least the total life cycle cost indicators, and the cost indicators include at least the initial investment cost, operation and maintenance cost, operating loss cost, and equipment retirement cost; The output-type indicators include at least: Transmission capacity and efficiency indicators, including rated transmission capacity and active power loss per unit distance; Control flexibility indicators include active / reactive power regulation capability, black start capability, and adaptability to weak power grids; Operational reliability metrics include annual average availability, operational complexity, fault ride-through capability, and broadband oscillation risk; Carbon emission reduction benefit indicators include carbon dioxide emission reduction, contribution rate of power grid cleanliness, and sensitivity of system carbon emission correlation.

[0025] It can also be divided into a three-level indicator system. The first-level indicator is a decoupled evaluation method for offshore wind power based on improved data envelopment analysis; the second-level indicators include: life cycle cost input indicators, transmission capacity and efficiency output indicators, control flexibility output indicators, operational reliability output indicators, and carbon emission reduction output indicators; the third-level indicators include: initial investment cost, operation and maintenance cost, operation loss cost, and equipment decommissioning cost under the above-mentioned life cycle cost input indicators; rated transmission capacity and active power loss per unit distance under the transmission capacity and efficiency output indicators; active / reactive power regulation capability, black start capability, and weak grid adaptability under the control flexibility output indicators; annual average availability, operation and maintenance complexity, fault ride-through capability, and broadband oscillation risk under the reliability and stability output indicators; and carbon emission reduction output indicators include carbon dioxide emission reduction, grid clean-up contribution rate, and system carbon emission correlation sensitivity. The specific indicators required can be set according to the actual situation and are not limited to the indicators and indicator system described in this embodiment. It is only necessary to ensure the distinction between input indicators and output indicators to facilitate subsequent calculation steps.

[0026] For indicators with clear dimensions and engineering observation values, such as rated transmission capacity and annual average availability, direct data for these indicators are obtained through project financial records, equipment manufacturer specifications, and historical operation and maintenance reports. For comprehensive or probabilistic indicators such as active power loss per unit distance, adaptability to weak grids, broadband oscillation risk, contribution rate of grid clean energy conversion, and sensitivity of system carbon emissions, indirect values ​​of the indicators are obtained through aggregation calculations using multiple basic variables, including but not limited to environmental parameters, fault frequency, recovery time, power flow data, and grid power supply structure. This calculation is a conventional calculation method and will not be elaborated on in this embodiment.

[0027] S2. Obtain historical operating data of the offshore wind power transmission system, and quantify the coupling relationship between the indicators in the indicator system based on the historical operating data and the mutual information distance coupling coefficient. S2 acquires historical operational data of offshore wind power transmission systems, including: Build with Sample The original indicator data matrix of each evaluation indicator ; , in, This represents input-type indicator data; This represents output-type indicator data; This represents the total number of samples. The total number of evaluation indicators; This is the original indicator data matrix.

[0028] To eliminate the differences in dimensions and orders of magnitude among the indicators, the original indicator data matrix is ​​normalized.

[0029] For input-type indicators: , in, For the first The first in the sample The original values ​​of each input-type indicator; , Let each represent the i-th sample in all samples. The maximum and minimum values ​​of the indicator; This is the normalized value.

[0030] For output-type indicators: .

[0031] in, For the first The first in the sample The original values ​​of each output-type indicator; , Let each represent the i-th sample in all samples. The maximum and minimum values ​​of the indicator; This is the normalized value.

[0032] S2 quantifies the coupling relationship between indicators in the indicator system using mutual information distance coupling coefficients, including: For any two index vectors, calculate their mutual information coefficient, which satisfies the following: ,in, Indicators and Mutual information coefficient between them Indicates the number of evaluation indicators. and Indicators or indicators The marginal probability distribution; Calculate the distance correlation coefficient between the two indicator vectors, the calculation of which satisfies: ,in, Represents a bicentered matrix , The correlation coefficient between the distances, The squared variance of the distance to the bicentered matrix. Represents a bicentered matrix , The covariance matrix between them , As an indicator and The distance matrix is ​​processed by bicentering to obtain the bicentered matrix; The mutual information coefficient and the distance correlation coefficient are weighted and fused to obtain the mutual information distance coupling coefficient, which is calculated to satisfy the following: .in, As an indicator and Mutual information distance coupling coefficient between them The balancing parameter has a value range of [0, 1]. As an indicator and The normalized value of the mutual information coefficient.

[0033] For any two index vectors, their mutual information coefficient is calculated, including: Transform continuous index data into discrete probability distributions; Calculate the joint probability distribution and the individual marginal probability distributions of the two indicators; Specifically, for those with Sample Original evaluation data matrix of each evaluation indicator , , As one of the indicators and indicators The original data vector will be converted into a continuous evaluation index vector. and Transform it into a discrete probability distribution; Will and The value range is divided into several and non-overlapping intervals and ; Calculate the joint probability distribution and marginal probability distribution and : , , , in, Indicators and The joint probability distribution of ; Indicate data pairs Joint fall The intervals and The Frequency of each interval; Indicators The marginal probability distribution; Indicators The marginal probability distribution.

[0034] Based on the joint probability distribution and the marginal probability distribution, the information entropy of the two indicators is calculated, and the calculation satisfies: , in, Indicators Information entropy Indicators Information entropy; Calculation indicators and The mutual information coefficient measures the reduction in uncertainty about one indicator when information about another is known. , in, Indicators and Mutual information coefficient between them.

[0035] The mutual information value is calculated based on the information entropy and joint probability distribution, and normalization is performed using the minimum marginal entropy method. and The mutual information coefficients between them are converted into a coupling strength index with a defined upper limit, and the normalized calculation satisfies: ,in, Indicators and The minimum value of its information entropy.

[0036] The above calculation of the distance correlation coefficient between the two indicator vectors includes: Calculation indicators and index vector and The Euclidean distance between any two points in the middle forms an index. and Distance matrix and ;

[0037] Wherein, the distance matrix for The square formation, Represents the distance matrix in which the first... Line number The elements of the column are also vectors. In The component and the first The absolute value of the difference between the components.

[0038] , Similarly, vectors can be obtained. Distance matrix ;

[0039] .

[0040] For distance matrix and Perform dual-center processing to generate a dual-center matrix. and To eliminate the dependency between mean effect and isolation distance, its calculation satisfies: , in, For a dual-centered matrix The Middle Line number Column elements; Distance matrix The Middle The mean of the rows; Distance matrix The Middle The mean of the column; Distance matrix The total average of all elements in the dataset.

[0041] Similarly, the distance matrix can be obtained. Bi-centered matrix , its first Line number Column elements Obtained from the following formula; , in, Bi-centered distance matrix The Middle Line number Column elements; Distance matrix The Middle The mean of the rows; Distance matrix The Middle The mean of the column; Distance matrix The total average of all elements in the dataset.

[0042] Calculate the bicentered matrix , The covariance matrix between them:

[0043] in, Represents a bicentered matrix , The covariance matrix between them.

[0044] S3. Based on the mutual information distance coupling coefficient, the index system is weighted using network analysis to obtain a weight vector reflecting the coupling relationship between the indicators; S3 specifically includes the following steps: The mutual information distance coupling coefficient between each pair of indicators As objective impact coefficients, the objective impact coefficients are normalized column-wise to construct an objective impact matrix; Specifically, the objective influence matrix is ​​constructed after column normalization. ,in: , In the formula: Indicating the objective indirect influence matrix, the first... Line number Column elements; Indicates the first The first indicator and the first Mutual information distance coupling coefficient between indicators; Indicates the first The sum of the column mutual information distance coupling coefficients.

[0045] Solve for all eigenvectors of the objective influence matrix and construct a weight matrix composed of the eigenvectors. ; Based on the weight matrix We weight the objective influence matrix to construct a weighted hypermatrix: , in, Represents the first in the objective influence matrix Line number The weighted element of the column, Represents the weight matrix of the first element. Line number The weighted element of the column.

[0046] The weighted hypermatrix is ​​iterated through to a power level, and the weight vector is obtained after convergence. Its calculation satisfies the following: ,in, The extreme weight vector representing the indicator system already incorporates the coupling information between indicators. express Power of 1.

[0047] By transforming the aforementioned weight vector into weight constraints of the improved data envelopment model, the final evaluation correction superefficiency of the sending scheme can be solved.

[0048] S4. Transform the weight vector into weight constraints and construct an improved data envelopment analysis model; use the improved data envelopment analysis model to comprehensively evaluate the offshore wind power transmission scheme to be evaluated.

[0049] The weight constraint of S4 forces the improved data envelopment model to adhere to the importance levels of the indicators determined by the weighting method described above when searching for the optimal efficiency solution. If the weighting considers the indicators... The importance of indicators The virtual weight index obtained by the improved data envelopment model is twice that of the original model. The importance of indicators Twice as much.

[0050] To eliminate the problem of excessive weight freedom in the improved data envelopment model, the evaluation results are ensured to reflect not only the input-output ratio of the scheme itself, but also the complex coupling relationship and associated risks between indicators, thereby improving the scientific nature of the evaluation results.

[0051] The above weight constraints satisfy: , in, Indicates the first The weight of each input indicator Indicates the first The weight of input indicators Indicates the first Weights of output indicators Indicates the first The weight of output indicators and These represent the lower limits of the input-type weighting ratio and the output-type weighting ratio, respectively. and They represent the first Item and the The ultimate weights of input-type indicators, i.e., the weight vector obtained after iterative convergence. and They represent the first Item and the The extreme weights of output-type indicators This represents the minimum relaxation amount, which is approximately zero.

[0052] The improved data envelopment analysis model satisfies: , in, This represents the relative efficiency value of the proposed solution to be evaluated. Indicates the first The weight of input indicators Indicates the first The weight of output indicators and These represent the total number of input indicators and output indicators, respectively. This indicates the total number of submitted proposals to be evaluated. and They represent the first The first evaluation scheme Input-type indicators and the first The value of an output-type indicator, and They represent the first The first evaluation scheme Input-type indicators and the first The value of an output-type indicator, Represented as the first The weight constraints in each evaluation scheme, Represented as an infinitesimal; The improved data envelopment analysis model is solved to obtain the relative efficiency value of each sending scheme. ; according to Sort all sending schemes by size, and The scheme with the largest value is selected as the optimal sending scheme.

[0053] Example 2 Example 2 is a simulation experiment conducted based on Example 1 to illustrate the effectiveness of this method.

[0054] Figure 3 This paper presents a comparison of the sensitivity of the proposed coupling weighting method with traditional weighting methods when facing changes in indicators. As shown in the figure, for the two strongly coupled indicators I5 and I8, the proposed method exhibits a significant sensitivity response, with weight changes reaching approximately +0.06 and +0.05, respectively; while the traditional method shows a relatively sluggish response, only +0.02 and +0.03. This demonstrates that the proposed method successfully identifies and quantifies the nonlinear close relationship between indicators using the mutual information distance coupling coefficient. The higher sensitivity of the proposed method means that when an indicator in the system fluctuates, the model can objectively transmit this influence through the coupling coefficient and automatically adjust the weights of related indicators. This overcomes the isolated evaluation defect caused by the traditional method ignoring the nonlinear correlation between indicators, and can more realistically reflect the risk transmission mechanism in offshore wind power transmission systems where a change in one indicator affects the entire system. Compared to traditional weighting methods, the method proposed in this patent objectively quantifies the nonlinear correlation between indicators through the mutual information distance coupling coefficient and integrates the coupling information into the hypermatrix iteration of network analysis, thus identifying indicators as having a strong coupling relationship. Based on the weight differentiation allocation of coupled information, the traditional weighting method overcomes the shortcomings of high subjectivity and inability to capture the complex relationship between indicators, and realizes the scientific allocation of weights in the evaluation system, ensuring that the final evaluation results can accurately reflect the complex coupling and dependency relationship between indicators of offshore wind power transmission scheme. Figure 4 This diagram illustrates the evaluation results according to an embodiment of the present invention. Under the evaluation of the traditional DEA model, the efficiency values ​​of schemes DMU-A, DMU-C, and DMU-D are all 1.0, at the same level, making it difficult for decision-makers to distinguish the merits of these three schemes, i.e., the problem of multiple effective units being unable to be ranked arises. The method of the present invention, by introducing coupling weight constraints, breaks through the limitations of the traditional model and further distinguishes the effective schemes. The results show that DMU-D (efficiency value > 1.2) is significantly better than DMU-A (approximately 1.15) and DMU-C (approximately 1.05). The evaluation results of the present invention not only have high discrimination, but also, because the model incorporates complex coupling relationships between indicators, the resulting super-efficiency value more realistically reflects the comprehensive performance of each scheme when facing complex risks and related constraints in actual engineering. DMU-D is identified as the optimal scheme, providing decision-makers with a clear and scientific basis. By combining coupled weighting with a super-efficiency model, this method overcomes the shortcomings of traditional data envelopment models in terms of low scheme differentiation, and ensures that the evaluation results incorporate the complex coupling relationships and associated risks between indicators, thereby providing a reliable basis for scientific decision-making on offshore wind power transmission schemes.

[0055] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this invention.

Claims

1. An offshore wind power decoupling evaluation method based on improved data envelopment analysis, characterized in that, The method comprises the following steps: An input-output index system for evaluating offshore wind power transmission schemes is constructed; Historical operation data of the offshore wind power transmission system are obtained, and the coupling relationship between the indexes in the index system is quantified by using a mutual information distance coupling coefficient according to the historical operation data; The index system is weighted by using a network analysis method according to the mutual information distance coupling coefficient, and a weight vector reflecting the coupling relationship between the indexes is obtained; The weight vector is converted into a weight constraint, and an improved data envelopment analysis model is constructed; and the offshore wind power transmission scheme to be evaluated is comprehensively evaluated by using the improved data envelopment analysis model.

2. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1, wherein, The indexes in the index system at least include input indexes and output indexes; The input indexes at least include a whole life cycle cost index, and the cost index at least includes an initial investment cost, an operation and maintenance cost, an operation loss cost and a device decommissioning cost; The output indexes at least include: a power transmission capacity and efficiency index including a rated power transmission capacity and a unit distance active loss; a control flexibility index including an active / reactive power regulation capacity, a black start capacity and a weak grid adaptability; an operation reliability index including an annual average availability, an operation and maintenance complexity, a fault ride-through capacity and a wide frequency oscillation risk; a carbon emission reduction benefit index including a carbon dioxide emission reduction amount, a power grid clean contribution rate and a system carbon emission correlation sensitivity.

3. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1 or 2, characterized in that, The historical operation data of the offshore wind power transmission system are obtained, including: Constructing a raw index data matrix with one sample one evaluation index ; The original index data matrix is normalized.

4. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1, wherein, The coupling relationship between the indexes in the index system is quantified by using a mutual information distance coupling coefficient, including: For any two index vectors, a mutual information coefficient is calculated, and the calculation satisfies: wherein, denotes the index and the mutual information coefficient between denotes the number of evaluation indices, and denotes the index or the marginal probability distribution of the index . A distance correlation coefficient between the two index vectors is calculated, and the calculation satisfies: wherein denotes a double centered matrix , the distance correlation coefficient between is the squared distance variance of the double centered matrix denotes a double centered matrix , the covariance matrix between , is the distance matrix of the indicators and after double centering; The mutual information coefficient and the distance correlation coefficient are weighted and fused to obtain a mutual information distance coupling coefficient, and the calculation satisfies: wherein, is an index and is a mutual information distance coupling coefficient between is a balancing parameter, is an index and is a normalized value of the mutual information coefficient.

5. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 4, characterized in that, For any two index vectors, a mutual information coefficient is calculated, including: Continuous index data are converted into discrete probability distribution; Joint probability distribution and marginal probability distribution of the two indexes are calculated; Based on the joint probability distribution and the marginal probability distribution, information entropy of the two indexes is calculated, and the calculation satisfies: ， wherein represents an index of the information entropy, represents an index of the information entropy; A mutual information value is calculated according to the information entropy and the joint probability distribution, and normalized, and the normalization calculation satisfies: wherein, denotes an indicator and the minimum value in its information entropy.

6. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 4, wherein, A distance correlation coefficient between the two index vectors is calculated, including: Computing the indicators and Indicator vectors and Euclidean distance between any two points in and Distance matrix and ; The distance matrix and is double centered to produce a double centered matrix and which eliminates the mean effect and isolates the dependence between distances.

7. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1 or 4, characterized in that, The index system is weighted by using a network analysis method according to the mutual information distance coupling coefficient, and a weight vector reflecting the coupling relationship between the indexes is obtained, including the following steps: The mutual information distance coupling coefficient between every two indexes is taken as an objective influence coefficient, the objective influence coefficient is normalized by column, and an objective influence matrix is constructed; Solving all eigenvectors of the objective influence matrix, constructing a weight matrix composed of the eigenvectors ; Based on the weight matrix The objective influence matrix is weighted to construct a weighted super matrix; The weighted super matrix is iterated by power, and a weight vector is obtained after convergence.

8. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 7, characterized in that, Based on the weight matrix The objective influence matrix is weighted to construct a weighted super matrix, and the calculation satisfies: Constructing weighted supermatrices , wherein, denotes a weight element of the objective influence matrix in the i-th row and the j-th column, denotes a weight element of the objective influence matrix in the i-th row and the j-th column, denotes a weight element of the objective influence matrix in the i-th row and the j-th column, denotes a weight element of the objective influence matrix in the i-th row and the j-th column, denotes a weight element of the objective influence matrix in the i-th row and the j-th column, denotes 9. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1, wherein, The weight vector is converted into a weight constraint, and the weight constraint satisfies: ， in, Indicates the first The weight of each input indicator Indicates the first The weight of input indicators Indicates the first Weights of output indicators Indicates the first The weight of output indicators and These represent the lower limits of the input-type weighting ratio and the output-type weighting ratio, respectively. and They represent the first Item and the The extreme weight of input-type indicators and They represent the first Item and the The extreme weights of output-type indicators This indicates a very small amount of relaxation.

10. The offshore wind power decoupling evaluation method based on improved data envelopment analysis according to claim 1, wherein, The offshore wind power transmission scheme to be evaluated is comprehensively evaluated by using the improved data envelopment analysis model, and the improved data envelopment analysis model satisfies: ， wherein, represents the relative efficiency value of the delivery scheme to be evaluated, represents the value of the i-th input-type index in the j-th evaluation scheme, represents the weight of the i-th input-type index, represents the value of the i-th output-type index in the j-th evaluation scheme, represents the weight of the i-th output-type index, and respectively represent the total number of input-type indexes and output-type indexes, represents the total number of delivery schemes to be evaluated, and respectively represent the value of the i-th input-type index and the value of the i-th output-type index in the j-th evaluation scheme, represents the value of the i-th input-type index in the j-th evaluation scheme, represents the value of the i-th output-type index in the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents the weight constraint for the j-th evaluation scheme, represents an infinitesimal quantity; Solving the improved data envelopment analysis model to obtain a relative efficiency value of each delivery scheme ; According to the size of all delivery schemes, the scheme with the largest value is sorted as the optimal delivery scheme.

Images